Niedermeyer’s

Electroencephalography
Basic Principles, Clinical Applications, and Related Fields

SIXTH EDITION
Niedermeyer’s
Electroencephalography
Basic Principles, Clinical Applications, and Related Fields

SIXTH EDITION

Donald L. Schomer, MA, MD

Professor of Neurology
Harvard Medical School
Director, Laboratory for Clinical Neurophysiology
Chief, Comprehensive Epilepsy Program
Beth Israel Deaconess Medical Center
Boston, Massachusetts

Fernando H. Lopes da Silva, MD, PhD

Emeritus Professor
Swammerdam Institute for Life Sciences
Center of Neurosciences
Faculty of Science
University of Amsterdam
Amsterdam, The Netherlands
Acquisitions Editor: Frances Destefano
Product Manager: Tom Gibbons
Vendor Manager: Bridgett Dougherty
Senior Manufacturing Manager: Benjamin Rivera
Marketing Manager: Brian Freiland
Design Coordinator: Stephen Druding
Production Service: MPS, a Macmillan Company

© 2011 by LIPPINCOTT WILLIAMS & WILKINS, a WOLTERS KLUWER business.

Two Commerce Square

2001 Market Street
Philadelphia, PA 19103 USA
LWW.com

Printed in China

All rights reserved. This book is protected by copyright. No part of this book may be reproduced in
any form by any means, including photocopying, or utilized by any information storage and retrieval
system without written permission from the copyright owner, except for brief quotations embodied
in critical articles and reviews. Materials appearing in this book prepared by individuals as part of
their official duties as U.S. government employees are not covered by the above-mentioned copyright.

Library of Congress Cataloging-in-Publication Data

Niedermeyer’s electroencephalography: basic principles, clinical applications, and related fields/

[edited by] Donald L. Schomer, Fernando H. Lopes da Silva.—6th ed.
p. ; cm.
Electroencephalography
Rev. ed. of: Electroencephalography/[edited by] Ernst Niedermeyer, Fernando H. Lopes da Silva.
5th ed. c2005.
Includes bibliographical references and index.
ISBN-13: 978-0-7817-8942-4 (hardback: alk. paper)
ISBN-10: 0-7817-8942-7 (hardback: alk. paper) 1.
Electroencephalography. I. Niedermeyer, Ernst, 1920– II. Schomer, Donald L. III. Lopes da Silva,
F. H., 1935– IV. Title: Electroencephalography.
[DNLM: 1. Electroencephalography. 2. Central Nervous System Diseases—diagnosis. WL 150]
RC386.6.E43N54 2011
616.8’047547—dc22
2010037480

Care has been taken to confirm the accuracy of the information presented and to describe generally
accepted practices. However, the authors, editors, and publisher are not responsible for errors or
omissions or for any consequences from application of the information in this book and make no
warranty, expressed or implied, with respect to the currency, completeness, or accuracy of the
contents of the publication. Application of the information in a particular situation remains the
professional responsibility of the practitioner.
The authors, editors, and publisher have exerted every effort to ensure that drug selection and
dosage set forth in this text are in accordance with current recommendations and practice at the time
of publication. However, in view of ongoing research, changes in government regulations, and the
constant flow of information relating to drug therapy and drug reactions, the reader is urged to check
the package insert for each drug for any change in indications and dosage and for added warnings and
precautions. This is particularly important when the recommended agent is a new or infrequently
employed drug.
Some drugs and medical devices presented in the publication have Food and Drug Administration
(FDA) clearance for limited use in restricted research settings. It is the responsibility of the health care
provider to ascertain the FDA status of each drug or device planned for use in their clinical practice.
To purchase additional copies of this book, call our customer service department at (800)
638-3030 or fax orders to (301) 223-2320. International customers should call (301) 223-2300.
Visit Lippincott Williams & Wilkins on the Internet: at LWW.com. Lippincott Williams &
Wilkins customer service representatives are available from 8:30 am to 6 pm, EST.
10 9 8 7 6 5 4 3 2 1
 Dedicated to Ernst Niedermeyer, M.D.
for all of the help and inspiration offered to us and
to countless clinical neurophysiologists worldwide
over his truly remarkable career.
Preface to the First Edition
The history of clinical electroencephalography (EEG) has just
passed the 50-year mark. The age of the pioneers was followed
by the stage of expansion. What began in a few prestigious cen-
ters gradually became a tool of all academic medical institu-
tions, and eventually of all major hospitals. In more recent
years, EEG even invaded the private offices of practicing neurol-
ogists and other specialists interested in central nervous system
(CNS) disease.
From this perspective, the history of clinical EEG looks like
the “via triumphalis” of a buoyantly dynamic new subspecialty.
The elucidation of the electrophysiologic process underlying
epileptic seizure disorders and a variety of other CNS dysfunc-
tions was indeed a unique achievement made possible by the
new method. The original intention of the founder of clinical
EEG, Hans Berger, had been the exploration of mental and psy-
chological processes, and even in this domain the yield has been
substantial. Moreover, electroencephalographers have not con-
fined themselves to the spontaneous wave patterns of the brain;
forms of EEG data analysis with the aid of computers were
introduced in order to demonstrate evoked and event-related
potentials and to investigate the wealth of frequencies that con-
stitute the EEG. In the search for the sources of EEG generation,
the brain tissue became the target of exploration.
Depth electrodes became one of the most important tools of
experimental neurophysiologists, who also investigated the sin-
gle neuron using microelectrodes. The implantation of depth
electrodes in the human brain has aided in the evaluation of
chronic epileptics considered candidates for seizure surgery.
This impressive progress, however, has been counterbal-
anced by signs of pessimism, fatigue, and resignation. A certain
malaise has inched its way into the hearts of thousands of elec-
troencephalographers who have started to feel the grip of stag-
nation. Generation of EEG potentials has proved to be
extremely complex and difficult to understand; the feeling of
doing pragmatically useful work with an ill-understood
method has been depressing to many workers in the field. The
pragmatists have further suffered from the limitations of EEG
as a method of localization of cerebral lesions. These feelings
have been nourished by the phenomenal achievements of non-
invasive radionuclear and radiologic scanning methods; this
progress of new methods in the field of structural diagnosis has
been a matter of concern for many electroencephalographers.
A more real danger, perhaps, is presented by the poorly trained
colleagues who are tarnishing the image of EEG.
In reality, however, these challenges present a stimulus for
the electrophysiologic field. The function-oriented aspects of
neurologic sciences will always be of paramount significance.
The loss of function-oriented neurology would foreshadow the
death of neurology. With all due respect for the structural
vi
aspects of lesions and tissue changes, neurology would be
shallow and barren without awareness of the constant fluctua-
tion of functional states in the CNS. Another important sti-
mulus has been the establishment of standards of quality for
electroencephalographers and EEG laboratories.
In these times of challenge, a review of the state of affairs in
EEG seems to be appropriate. Such thoughts have prompted the
genesis of this large one-volume textbook, which, by its mere
size, sets itself apart from the group of smaller introductory
textbooks and from the huge multivolume handbook. The one-
volume character of the book symbolizes the spirit of unity that
should reign among clinical electroencephalographers, basic
science researchers, and the workers in the field of compu -
terized data analysis.
One author can hardly undertake such a task alone. For this
reason, we have reached out across the ocean for collaboration.
It became clear, however, that a two-man effort would not
suffice to cover the entire field in the relatively short working
period of 2 years. We solicited for assistance and found a
wonderful group of contributors in various special areas. Above
all, coverage of the fields of neurophysiology and neurophar-
macology has depended most heavily on the assistance of
prestigious specialists.
Attempts at synthesis have not been the goal of this book.
Instead the reader will find a more individualistic approach
from which the personal basic philosophy of each author can be
derived. No effort has been made to achieve strict standardiza-
tion of symbols and terminology; as an example, frequencies
are described in various terms (10 cps, 10/sec, 10 Hz, etc.).
There is also some overlap between certain chapters; we feel
that the reader will benefit from the presentation of a given
topic as seen from two somewhat different viewpoints.
A piece of technical information might be worthwhile.
Unless the filter setting is specifically indicated in the illus-
trations of EEG tracings, a time constant between 0.15 and
0.5 seconds was used (above 0.4 seconds when slow frequencies
played a major role). The use of “muscle filters” was avoided.
We have tried to combine didactic and academic elements in
this book. Hence, trainees as well as seasoned professionals in
the field will, we hope, find what they are searching for. This
dual approach does mean that some parts of the book require
greater sophistication of the reader than do others.
Acknowledgments for invaluable help in this undertaking
come from the depth of our hearts. Mr. Braxton Dallam
Mitchell, President of Urban & Schwarzenberg in Baltimore,
Maryland, deserves the honor of having been the initiator of
this book. His encouragement has been very much appreciated.
Detlev Moos has coordinated the production of this book with
care and efficiency; Suzanne Lohmeyer has copyedited and

indexed it well; and Nan Tyler, Carola Sautter, and Victoria
Doherty at the publisher’s Baltimore office assisted with their
experience. In the Johns Hopkins Hospital in Baltimore,
Maryland, the technical staff of the laboratory deserves great
praise: Mrs. Judy Nastalski, R. EEG T. and chief technologist,
Mr. Eric DeShields, Mrs. Debbie Reichenbach, R. EEG T.,
Miss Sharon Vaughan, Mrs. Kathleen Daniecki, R. EEG T.,
Mrs. Cindy Haywood, and Miss Kim Rimel. How deeply the
electroencephalographer depends on the quality of the record-
ings and the dedication of the technologists! Truly invaluable
was the secretarial assistance of Mrs. Catherine Bonolis. The
operation of the laboratory was further aided by the experience
of Mrs. Marie Simpson. Important contributions came from
Mr. Joe Dieter, who is responsible for the pictorial artwork,
Mr. Ron Garret (lettering of tracings), and Mr. Zuhair Kareem
and Mrs. Lillian Reich, the staff of Medical Photography.
Assistance and advice in the clinical EEG field was given
most freely by Dr. Gisela Freund, visiting assistant professor at
the John Hopkins Hospital EEG Laboratory (1980/81), of the
Department of Clinical Neurophysiology, Free University
Berlin (Klinikum Westend). E.N.’s principal teacher in the field
of EEG, Dr. John R. Knott (presently of Boston, Massachusetts),
and the great master of neurosurgery, epileptology, and neuro-
Preface vii
physiology, Dr. A Earl Walker (presently of Albuquerque, New
Mexico), deserve special gratitude.
Gratitude is expressed (by F.L.S.) to professor Dr. W. Storm
von Leeuwen, Dr. G. Wieneke, and Dr. K. Van Hulten (Utrecht
University Hospital), Mr. N.J.I. Mars (Twente University of
Technology), and Mr. A. Van Rotterdam (Institute of Medical
Physics, Utrecht) for their advice and encouragement. The high
professional competence of Mrs. Ada Van Schaik and Mr. Nico
Haagen (Institute of Medical Physics, Utrecht) in their fields of
secretarial work and artwork was of invaluable help.
Heartfelt thanks are also expressed to the contributing authors
of the book; they have naturally become a part of this undertak-
ing. The response of these splendid coworkers was exemplary.
Joseph J. Tecce and Lynn Cattanach, the authors of the article on
contingent negative variation, substituted for a colleague who
had to step down from his obligations at the last minute. They
made possible the almost impossible when they declared their
willingness to join the team of coworkers. To them, and to all the
contributing authors of this volume, our deepest thanks.
Ernst Niedermeyer
Fernando H. Lopes da Silva
Spring 1981
Preface to the Sixth Edition
We would like to express our most sincere thanks to Dr. Ernst
Niedermeyer for his leadership and direction over the past several
years as we have moved forth to produce this sixth edition of the
textbook that he had first introduced back in 1982. Additionally,
we would like to honor him as he is so richly deserving by renam-
ing this book Niedermeyer’s Electroencephalography: Basic
Principles, Clinical Applications, and Related Fields.
Dr. Niedermeyer was born in Vienna, Austria, in 1920. He
attended medical school in Austria but, during his third year in
medical school, he was witness to the rise of Nazism in
Germany and Austria and was inducted into the German army
just prior to the onset of World War II. He was stationed in
Berlin and was then literally drafted into the medical service
even though he had not yet graduated from medical school. He
said that he was basically treating wounded soldiers coming
back from the eastern front during the first several years of the
German invasion of the Soviet Union. He remained there until
Nazi officials learned that his maternal grandmother was Jewish
which, according to German law, made him an “undesirable.”
However, because of his medical background, he was not
interned but rather was sent to Normandy. He mentioned that
he was, as were many of the soldiers in Normandy, well aware of
the coming invasion by the Allied forces and he looked upon
the day of the invasion as one of his “happiest.” He said he was
wounded during the invasion and was taken as a POW by the
American forces taking part in the invasion. He was sent back as
a POW to Kansas where he remained and served out the rest of
the war as the prison camp doctor. He returned to Austria fol-
lowing cessation of fighting and finished his medical school and
stayed on in Austria, interested in and learning neurology/psy-
chiatry. He was introduced into clinical neurophysiology
through some of the historical works in neurophysiology that
had been conducted in Europe prior to World War II, including
the work of Hans Berger. He became a clinical neurophysiology
addict and was offered a training position at the University of
Iowa with Dr. John Knott. He took that position and stayed
with Dr. Knott for a number of years but was then offered a
position to work with Dr. A. Earl Walker at the Johns Hopkins
Hospital, Baltimore. He continued to do his clinical and basic
science research at Hopkins, where his career blossomed and he
became well known in the field. He has been an inspiration to
many trainees and colleagues, both at Johns Hopkins and
worldwide. Many of our colleagues have come to know him
through his kindness and his willingness to personally take an
interest in their research and to make recommendations and
suggestions about directions for further study. His knowledge
of the field is legendary and his personality is truly “old world”
in the best sense of the term. It is our most sincere desire that
viii
this classic volume continue for the years to come and evolve
with the field as we have seen Dr. Niedermeyer do so success-
fully over the last 25 years as its senior editor.
This edition has several new features, reflective of the changes
that have occurred in our field over the past 5 years since the fifth
edition. More and more, the field of digital recording has
expanded; however, in order to understand some of the short-
comings and pitfalls of digital EEG, people need to still address
the issues of basic analog recording principles. With an increased
use of digital recording, laboratories have collected new and dif-
ferent “technical artifacts.” We present here an attempt to start a
database for such artifacts in a hope that future editions will
continue to expand upon this and offer a fairly complete library
for beginners interested in our field. As noted in the fifth edition,
epilepsy monitoring units (EMUs) have continued to mush-
room. Similar growth has occurred in the use of EEG monitor-
ing in newborn, cardiac, trauma, and postoperative intensive
care units. With the significant advances in wireless communica-
tion and easy access to the Internet, such recordings can also be
viewed and transmitted locally virtually instantaneously and can
allow well-trained clinical neurophysiologists to see and opine
about patients’ conditions on a very time-relevant basis.
Hopefully, as future generations may show, this ability will sig-
nificantly influence our patients’ outcomes. Similarly, the field of
intraoperative clinical neurophysiology for spinal cord function,
cranial nerve function, and cranial vascular therapies has con-
tinued to evolve along with the wireless and Internet communi-
cations. This has allowed for close monitoring of neurologic
function during critical periods of operations, again with a time
course that allows for corrective actions to be taken on a mean-
ingful time frame.
We have reorganized the text regarding normal EEG and
epilepsy to more closely follow the normal aging patterns. We
have continued to present some of the more classical chapters
that have evolved over years on evoked potentials and routine
electroencephalography, including EEG in common neuro-
logic, metabolic, and heredito-degenerative diseases. We have
added and updated chapters and text regarding automated and
specialized mathematics-based analysis techniques to try to
keep up with this rapidly expanding field. We will continue to
update these techniques in future editions. We have included a
chapter on linking clinical neurophysiology to other investiga-
tive techniques such as functional MRI. We have updated the
chapter on magnetoencephalography to, again, reflect the sig-
nificant changes that have taken place, both from a technical
and a clinical perspective, in this field. The last several chapters
of this textbook attempt to present an overview of clinical neu-
rophysiology research that intersects with many other aspects of

the realm of brain sciences such as consciousness and cognitive
processing.
We would like to encourage the readers of these words and
this text to please let us know if they are aware of other areas of
interest that could or should be involved in future editions of
this textbook that reflect either oversights on our part or are
harbingers of future development. Finally, we thank all of the
authors and coauthors of the chapters of this book for their
tremendous effort in keeping this textbook relevant to our field.
Special thanks go to Fran Destefano and Franny Murphy who
Preface ix
have offered tremendous help and support in the technical
aspects of getting the chapters organized and together and for
Tom Gibbons from Lippincott and all those who have helped
again in getting the textbook prepared for publication. Again, in
the words of Dr. Niedermeyer, “many, many thanks.”
Donald L. Schomer
Fernando H. Lopes da Silva
ix
Contributors

Florin Amzica, Ph.D. Sudhansu Chokroverty, M.D., F.R.C.P.

Associate Professor of Stomatology Professor of Neuroscience
Montreal University Seton Hall University
Montreal, Canada South Orange, NJ
Clinical Professor of Neurology
Mary Repole Andriola, M.D.
Robert Wood Johnson Medical School
Professor of Neurology and Pediatrics
New Brunswick, NJ
Director, Divisions of Child Neurology, Clinical Neurophysiology
Professor and Co-chair of Neurology
and Pediatric Epilepsy
Program Director, Sleep Medicine and Clinical Neurophysiology
SUNY at Stony Brook
New Jersey Neuroscience Institute at JFK Medical Center
Stony Brook, NY
Edison, NJ
Helen Barkan, M.D., Ph.D.
Nathan E. Crone, M.D.
Assistant Professor of Neurology
Associate Professor of Neurology
Upstate Medical University Hospital
Johns Hopkins University
SUNY Upstate Medical University
Attending
Syracuse, NY
Johns Hopkins Hospital
Gerhard Bauer, M.D. Baltimore, MD
Professor of Neurology
Edgar dePeralta, M.D.
Medical University Innsbruck
Fellow in Neurology and Clinical Neurophysiology
Innsbruck, Austria
State University of New York Upstate
Richard Bauer, M.D., M.Sc. Syracuse, NY
Oberarzt, Department of Neurosurgery
Frank W. Drislane, M.D.
Medical University Innsbruck
Professor of Neurology
Innsbruck, Austria
Harvard Medical School
Robert L. Beach, M.D., Ph.D. Neurologist
Professor of Neurology Comprehensive Epilepsy Center
Upstate Medical University Beth Israel Deaconess Medical Center
Director, Epilepsy Program and EEG Laboratories Boston, MA
Upstate Medical University Hospital
Barbara Ann Dworetzky, M.D.
Syracuse, NY
Assistant Professor of Neurology
Steve Bild, R. EEG/ EP T., C.N.I.M. Harvard Medical School
Manager of Clinical Services, EEG/Evoked Potentials Lab Chief, Division of Epilepsy, EEG, and Sleep Neurology
Rush University Medical Center Brigham and Women’s Hospital
Chicago, IL Boston, MA
Gastone G. Celesia, M.D. Günter Edlinger, M.Sc., Ph.D.
Professor of Neurology (Retired) CEO
Loyola University of Chicago g.tec Medical Engineering GmbH
Chicago, IL Schiedlberg, Austria
Bernard S. Chang, M.D., M.M.Sc. Jonathan Charles Edwards, M.D.
Assistant Professor of Neurology Associate Professor of Neurosciences
Harvard Medical School Director, Comprehensive Epilepsy Center and Clinical
Comprehensive Epilepsy Center Neurophysiology Laboratories
Beth Israel Deaconess Medical Center Medical University of South Carolina
Boston, MA Charleston, SC
Keith H. Chiappa, M.D. Christian E. Elger, M.D., F.R.C.P.
Associate Professor of Neurology Professor of Epileptology
Harvard Medical School University of Bonn
Director, Electroencephalography Laboratory Head, Department of Epileptology
Massachusetts General Hospital University of Bonn Medical Centre
Boston, MA Bonn, Germany

x
 Contributors xi

Ronald G. Emerson, M.D. Richard A. Hrachovy, M.D.

Professor of Clinical Neurology Professor of Neurology
Columbia University College of Physicians and Surgeons Baylor College of Medicine
Attending Neurologist Medical Director of Neurophysiology
New York Presbyterian Hospital St. Luke’s Episcopal Hospital
Columbia University Medical Center Houston, TX
New York, NY
Aatif M. Husain, M.D.
Charles M. Epstein, M.D. Department of Neurology
Professor of Neurology Duke University Medical Center and Neurodiagnostic Center
Emory University School of Medicine Veterans Affairs Medical Center
Emory Healthcare Durham, NC
Atlanta, GA
Kai Kaila, Ph.D.
Bruce J. Fisch, M.D. Professor of Biological and Environmental Sciences
Professor of Neurology Neuroscience Center
University of New Mexico University of Helsinki
Director, Neurodiagnostic Laboratories and MEG Center Helsinki, Finland
University of New Mexico Hospital
Stiliyan Kalitzin, M.D.
Albuquerque, NM
Image Sciences Institute
John N. Gaitanis, M.D. University Medical Center Utrecht
Assistant Professor of Neurology and Pediatrics, Clinical Utrecht, The Netherlands
Warren Alpert School of Medicine at Brown University
Anton Kamp
Director of Pediatric Epilepsy
Biological Centre
Rhode Island Hospital/Hasbro Children’s Hospital
University of Amsterdam
Providence, RI
Amsterdam, The Netherlands
Ali Gorji, M.D.
Andres M. Kanner, M.D.
Institute for Physiology
Professor of Neurological Sciences and Psychiatry
Münster University
Rush Medical College at Rush University
Münster, Germany
Director, Laboratories of EEG and Video-EEG-Telemetry
Jean Gotman, Ph.D. Associate Director, Epilepsy Section
Professor Rush University Medical Center
Montreal Neurological Institute Chicago, IL
McGill University
Peter W. Kaplan, M.B., F.R.C.P.
Montreal, Québec, Canada
Department of Neurology
Riitta Hari, M.D., Ph.D. Johns Hopkins Bayview Medical Center
Academy Professor Baltimore, MD
Brain Research Unit, Low Temperature Laboratory
Andrew D. Krystal, M.D., M.S.
Aalto School of Science and Technology
Professor of Psychiatry and Behavioral Sciences
Espoo, Finland
Director, Quantitative EEG Laboratory
Consultant
Director, Insomnia and Sleep Research Laboratory
Department of Clinical Neurophysiology
Duke University School of Medicine
HUSLAB, Helsinki University Central Hospital
Durham, NC
Helsinki, Finland
Ekrem Kutluay, M.D.
Adam L. Hartman, M.D.
Associate Professor of Neurology
Assistant Professor of Neurology and Pediatrics
Medical College of Wisconsin
Johns Hopkins University School of Medicine
Milwaukee, WI
Attending Physician
Johns Hopkins Hospital Emma Laureta, M.D.
Baltimore, MD Assistant Professor of Neurology
Albert Einstein College of Medicine
Bin He
Attending
Department of Biomedical Engineering and Center for
Montefiore Medical Center
Neuroengineering
New York, NY
University of Minnesota
Minneapolis, MN Alan D. Legatt, M.D., Ph.D.
Professor of Neurology
Susan T. Herman, M.D.
Albert Einstein College of Medicine
Assistant Professor of Neurology
Director of Intraoperative Neurophysiology
Harvard Medical School
Montefiore Medical Center
Beth Israel Deaconess Medical Center
New York, NY
Boston, MA
xii Contributors

Ronald P. Lesser, M.D. Ernst Niedermeyer, M.D.

Professor of Neurology and Neurosurgery Professor Emeritus of Neurology and Neurological Surgery
Johns Hopkins University Johns Hopkins University School of Medicine
Johns Hopkins Medical Institutions Baltimore, MD
Baltimore, MD
Douglas R. Nordli, Jr., M.D.
Fernando H. Lopes da Silva, M.D., Ph.D. Professor of Pediatrics
Emeritus Professor Northwestern University Feinberg School of Medicine
Swammerdam Institute for Life Sciences, Center of Neurosciences Lorna S. and James P. Langdon Chair of Pediatric Epilepsy
Faculty of Science Children’s Memorial Hospital
University of Amsterdam Chicago, IL
Amsterdam, The Netherlands
Marc R. Nuwer, M.D., Ph.D.
François Mauguière, M.D. Professor of Neurology
Head, Department of Functional Neurology and Epileptology UCLA School of Medicine
Lyon University and Lyon Federative Institute of Neuroscience Department Head
Neurological Hospital Clinical Neurophysiology
Lyon, France Ronald Reagan UCLA Medical Center
Los Angeles, CA
Douglas Maus, M.D., Ph.D.
Assistant Professor of Neurology Trudy D. Pang, M.D., M.M.Sc.
SUNY Downstate Medical Center Instructor in Neurology
State University of New York at Brooklyn Harvard University
New York, NY Staff Neurologist
Beth Israel Deaconess Medical Center
Christoph M. Michel, M.D.
Boston, MA
Functional Brain Mapping Laboratory
Neurology Clinic and Department of Basic Neuroscience Alvaro Pascual-Leone, M.D., Ph.D.
University Hospital Professor of Neurology
University of Geneva Harvard Medical School
Geneva, Switzerland Director, Department of Neurology
Berenson-Allen Center for Noninvasive Brain Stimulation
Eli M. Mizrahi, M.D.
Beth Israel Deaconess Medical Center
Chair, Department of Neurology
Boston, MA
Professor of Neurology and Pediatrics
Peter Kellaway Section of Neurophysiology, Department of Neal S. Peachey, M.D., Ph.D.
Neurology Professor of Ophthalmology
Section of Pediatric Neurology, Department of Pediatrics Cleveland Clinic Lerner College of Medicine
Baylor College of Medicine Associate Chief of Staff for Research
Houston, TX Cleveland VA Medical Center
Cleveland, OH
Solomon L. Moshé, M.D.
Professor of Neurology Pediatrics, and Neuroscience Gert Pfurtscheller, Ph.D.
Albert Einstein College of Medicine Emeritus Professor
Director, Pediatric Neurology and Clinical Neurophysiology Laboratory of Brain–Computer Interfaces (BCI-Lab)
Montefiore Medical Center Institute for Knowledge Discovery
New York, NY Graz University of Technology
Graz, Austria
Janet M. Mullington, Ph.D.
Associate Professor of Neurology Rodney A. Radtke, M.D.
Harvard University Professor of Neurology
Beth Israel Deaconess Medical Center Duke University
Boston, MA Medical Director
Duke Hospital Sleep Disorder Center
Christa Neuper, Ph.D.
Duke University Hospital
Professor
Durham, NC
Institute of Psychology
University of Graz James J. Riviello, Jr., M.D.
Head of Institute George Peterkin Endowed Chair in Pediatrics
Laboratory of Brain–Computer Interfaces (BCI-Lab) Professor of Pediatrics and Neurology
Institute for Knowledge Discovery Baylor College of Medicine
Graz University of Technology Director, Epilepsy and Neurophysiology Program
Graz, Austria Director, Neurocritical Service, Section of Neurology
Chief of Neurophysiology
Texas Children’s Hospital
Houston, TX
 Contributors xiii

Alexander Rotenberg, M.D., Ph.D. William O. Tatum IV, D.O.

Assistant Professor of Neurology Professor of Neurology
Harvard Medical School Mayo School of Medicine
Attending Staff Physician, Neurology Director, Epilepsy Monitoring Unit
Children’s Hospital Mayo Clinic and Hospital
Boston, MA Jacksonville, FL
Erik Rumpl, M.D. Barbara Tettenborn, M.D., Ph.D.
Professor of Neurology Chair, Department of Neurology
Landeskrankenhaus Kantonsspital St. Gallen
Klagenfurt, Austria St. Gallen, Switzerland
Johannes Gutenberg University
Donald L. Schomer, M.A., M.D.
Mainz, Germany
Professor of Neurology
Harvard Medical School Ab van Rotterdam, Ph.D.
Director, Laboratory for Clinical Neurophysiology Senior Researcher (Retired)
Chief, Comprehensive Epilepsy Program Radiobiological Laboratory
Beth Israel Deaconess Medical Center Radiobiological Institute TNO
Boston, MA Rijswijk, The Netherlands
Margitta Seeck, M.D. Sampsa Vanhatalo, M.D., Ph.D.
Professor Department of Clinical Neurophysiology
Director of the EEG and Epilepsy Unit Children’s Castle
University Hospital of Geneva Helsinki University Central Hospital
Geneva, Switzerland Helsinki, Finland
Megan Selvitelli, M.D. Juha Voipio, Ph.D.
Maine Medical Partners Neurology Professor of Biological and Environmental Sciences
Scarborough, Maine University of Helsinki
Helsinki, Finland
Erw in-Josef Speckmann, M.D.
Professor Emeritus Thoru Yamada, M.D.
Institute of Physiology (Neurophysiology) Professor of Neurology
University of Münster Roy J. and Lucille A. Carver College of Medicine
Münster, Germany University of Iowa
Director, Division of Clinical Electrophysiology
Cornelis Jan Stam, M.D.
University of Iowa Hospitals and Clinics
Department of Clinical Neurophysiology
Iowa City, IA
VU University Medical Center
Amsterdam, The Netherlands Malcolm Yeh, M.D.
Associate Clinical Professor of Neurology
Travis R. Stoub, Ph.D.
Division of Clinical Electrophysiology
Assistant Professor of Neurological Sciences
Roy J. and Lucille A. Carver College of Medicine
Affiliated Scientist
University of Iowa
Rush University Medical Center
Staff Neurologist
Chicago, IL
University of Iowa Hospitals and Clinics
Takeo Takahashi, M.D. Iowa City, IA
Yaotome Clinic
Sendai, Japan
Contents
Preface to the First Edition vi
Preface to the Sixth Edition viii
Contributors x
Part I Basic Principles
1 Historical Aspects of EEG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Ernst Niedermeyer and Donald L. Schomer
2 Neurophysiologic Basis of EEG and DC Potentials . . . . . . . . . . 17
Erwin-Josef Speckmann, Christian E. Elger, and Ali Gorji
3 Cellular Substrates of Brain Rhythms . . . . . . . . . . . . . . . . . . . . . 33
Florin Amzica and Fernando H. Lopes da Silva
4 Dynamics of EEGs as Signals of Neuronal Populations:
Models and Theoretical Considerations . . . . . . . . . . . . . . . . . . . 65
Fernando H. Lopes da Silva
5 Biophysical Aspects of EEG and Magnetoencephalogram
Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Fernando H. Lopes da Silva With an Appendix by Ab Van Rotterdam
6 Analog Signal Recording Principles . . . . . . . . . . . . . . . . . . . . . . 111
Charles M. Epstein
7 Digital EEG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Douglas Maus, Charles M. Epstein, and Susan T. Herman
8 Polarity and Field Determinations . . . . . . . . . . . . . . . . . . . . . . . 143
Bruce J. Fisch
Part II Normal EEG
9 Normal EEG and Sleep: Preterm and Term Neonates . . . . . . . 153
Eli M. Mizrahi, Solomon L. Moshé, and Richard A. Hrachovy
10 Normal EEG and Sleep: Infants to Adolescents . . . . . . . . . . . . 163
James J. Riviello, Jr., Douglas R. Nordli, Jr., and Ernst Niedermeyer
11 Normal EEG and Sleep: Adults and Elderly . . . . . . . . . . . . . . . 183
Bernard S. Chang, Donald L. Schomer, and Ernst Niedermeyer
12 Activation Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Takeo Takahashi and Keith H. Chiappa
13 Artifacts of Recording . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
Barbara Dworetzky, Susan Herman, and William O. Tatum IV
Part III Clinical EEG: General
14 Patterns of Unclear Significance . . . . . . . . . . . . . . . . . . . . . . . . . 267
Jonathan Charles Edwards and Ekrem Kutluay
15 EEG of Degenerative Disorders of the Central
Nervous System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
John Gaitanis
16 The EEG in Congenital Malformations of Cortical
Development, Neurocutaneous Disorders, Cerebral Palsy,
Autism/Mental Retardation, and ADHD/Learning
Disabilities of Childhood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
Mary Repole Andriola
17 Brain Tumors and Other Space-Occupying Lesions . . . . . . . . 321
Adam L. Hartman and Ronald P. Lesser
18 The EEG in Inflammatory CNS Conditions . . . . . . . . . . . . . . . 331
Robert L. Beach, Helen Barkan, and Edgar DePeralta
xiv
19 Cerebrovascular Diseases and EEG . . . . . . . . . . . . . . . . . . . . . . 351
Barbara Tettenborn, Ernst Niedermeyer, and Donald L. Schomer
20 Dementia and EEG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Cornelis Jan Stam
21 Metabolic Disorders and EEG . . . . . . . . . . . . . . . . . . . . . . . . . . 395
Trudy Pang, Megan Selvitelli, Donald L. Schomer, and Ernst Niedermeyer
22 EEG and Craniocerebral Trauma . . . . . . . . . . . . . . . . . . . . . . . . 411
Erik Rumpl
23 Anoxia, Coma, and Brain Death . . . . . . . . . . . . . . . . . . . . . . . . 435
Peter W. Kaplan and Gerhard Bauer
24 The EEG in Patients with Migraine and Other
Forms of Headache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
Ernst Niedermeyer and Donald L. Schomer
Part IV Clinical EEG in Epilepsy
and Related Disorders
25 Seizures and Epilepsies in the Preterm and Term Neonate ... 465
Emma Laureta, Eli M. Mizrahi, and Solomon L. Moshé
26 Seizures and Epilepsy in Infants to Adolescents . . . . . . . . . . . . 479
Douglas R. Nordli, Jr., James J. Riviello, Jr., and Ernst Niedermeyer
27 Epilepsy in Adults and the Elderly . . . . . . . . . . . . . . . . . . . . . . . 541
Bernard S. Chang, Donald L. Schomer, and Ernst Niedermeyer
28 Convulsive Status Epilepticus . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
Frank W. Drislane, Susan T. Herman, and Peter W. Kaplan
29 Nonconvulsive Status Epilepticus . . . . . . . . . . . . . . . . . . . . . . . 595
Frank W. Drislane, Peter W. Kaplan, and Susan T. Herman
30 Anticipating Seizures Based on EEG . . . . . . . . . . . . . . . . . . . . . 645
Fernando H. Lopes da Silva and Stiliyan Kalitzin
31 Nonepileptic Attacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
Megan Selvitelli, Trudy Pang, Donald L. Schomer, and Ernst Niedermeyer
Part V Complementary and Special Techniques
32 Nasopharyngeal, Anterotemporal,
and Sphenoidal Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
Andres M. Kanner, Travis Stoub, and Steve Bild
33 Intracranial Monitoring: Depth, Subdural,
and Foramen Ovale Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . 677
Margitta Seeck, Donald L. Schomer, and Ernst Niedermeyer
34 Electrocorticography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715
Marc R. Nuwer
35 Principles and Techniques for Long-Term EEG Recording
(EMU, ICU, Ambulatory) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Jean Gotman, Marc Nuwer, and Ronald G. Emerson
36 Infraslow EEG Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741
Sampsa Vanhatalo, Juha Voipio, and Kai Kaila
37 High-Frequency EEG Activity . . . . . . . . . . . . . . . . . . . . . . . . . . 749
Jean Gotman and Nathan E. Crone
38 Intraoperative Evoked Potential Monitoring . . . . . . . . . . . . . . 767
Alan D. Legatt
39 Monitoring EEG during Carotid Surgery . . . . . . . . . . . . . . . . . 787
Thoru Yamada and Malcolm Yeh
 Contents xv

40 Polygraphy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809 50 Neurocognitive Processes and the EEG/MEG . . . . . . . . . . . . . 1083

Anton Kamp, Gert Pfurtscheller, Günter Edlinger, Fernando H. Lopes da Silva
and Fernando H. Lopes da Silva

41 Polysomnography: Technical and Clinical Aspects . . . . . . . . . 817 Part VIII EEG and Neurocognitive Functions
Sudhansu Chokroverty, Rodney Radtke, and Janet Mullington
51 Psychiatric Disorders and EEG . . . . . . . . . . . . . . . . . . . . . . . . . 1113
42 Magnetoencephalography: Methods and Applications . . . . . . 865 Andrew D. Krystal
Riitta Hari
52 Technical Aspects of Transcranial Magnetic
Part VI EEG: Neuropharmacology and Anesthesia and Electrical Stimulation of the Brain . . . . . . . . . . . . . . . . . . 1129
Alan D. Legatt, Alvaro Pascual-Leone, and Alexander Rotenberg
43 EEG, Drug Effects, and Central Nervous System Poisoning . . 901 53 Transcranial Magnetic Stimulation (TMS):
Gerhard Bauer and Richard Bauer
Clinical Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
Alexander Rotenberg and Alvaro Pascual-Leone
Part VII Evoked Potentials and Event-Related EEG
Phenomena Part IX Computer-Assisted EEG Analysis
44 Event-Related Potentials: General Aspects of 54 EEG Analysis: Theory and Practice . . . . . . . . . . . . . . . . . . . . . 1147
Methodology and Quantification . . . . . . . . . . . . . . . . . . . . . . . 923 Fernando H. Lopes da Silva
Fernando H. Lopes da Silva
55 EEG Mapping and Source Imaging . . . . . . . . . . . . . . . . . . . . . 1179
45 EEG Event-Related Desynchronization (ERD) Christoph M. Michel and Bin He
and Event-Related Synchronization (ERS) . . . . . . . . . . . . . . . . 935 56 Computer-Assisted EEG Pattern Recognition
Gert Pfurtscheller and Fernando H. Lopes da Silva
and Diagnostic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1203
46 Visual Evoked Potentials and Electroretinograms . . . . . . . . . . 949 Fernando H. Lopes da Silva
Gastone G. Celesia and Neal S. Peachey
57 EEG-Based Brain–Computer Interfaces . . . . . . . . . . . . . . . . . 1227
47 Brainstem Auditory Evoked Potentials (BAEPs) Gert Pfurtscheller and Christa Neuper
and Other Auditory Evoked Potentials . . . . . . . . . . . . . . . . . . . 975 58 Multimodal Monitoring of EEG and Evoked Potentials . . . . 1237
Gastone G. Celesia Gert Pfurtscheller
48 Somatosensory-Evoked Potentials: Normal Responses,
Abnormal Waveforms, and Clinical Applications
in Neurologic Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1003 Index 1243
François Mauguière

49 Evoked Potentials in Children and Infants . . . . . . . . . . . . . . . 1057

Aatif M. Husain

Historical Aspects of EEG
ERNST NIEDERMEYER AND DONALD L. SCHOMER
DISCOVERY OF ELECTRICAL PHENOMENA
Thales from Miletos has been credited with the discovery of
static electricity produced by friction (rubbing fur or glass with
silk). He was one of the pre-Socratic “natural philosophers” of
Greece (around 620–550 BC) and considered water the origin
of all things. Thus, friction was recognized as the generator of a
phenomenon that derived its name from the Greek work “elec-
tron,” which stands for amber. This discovery fell into a dor-
mant stage for more than two millennia.
Around 1600, William Gilbert began to study the electrical
properties of various substances, and Otto von Guericke
(1602–1686) invented the friction machine to create electrical
fields. This machine eventually found its way into doctors’
offices and even university hospitals. Its electrical field would
make a patient’s hair stand up, creating a strong impression on
a psychologically gullible patient. These friction machines now
ornament high school laboratories and technical museums. In
the 17th and 18th centuries, the friction machine taught invalu-
able lessons on attraction and repulsion of charged bodies, on
conductors and nonconductors, and on the rather questionable
dualism of positive and negative electricity.
A new and very important piece of electrical equipment
entered the scene in 1746 when the Leyden jar was introduced
by Pieter van Musschenbroek (following the earlier work of
Ewald von Kleist). This invention resulted in the storage of elec-
tricity, and its upshot, the condenser or capacitor, turned into
an indispensable part of modern electronics. Benjamin
Franklin’s bold experiment caught electrical discharges of a
thunderstorm in a Leyden jar.
What the friction machine could generate, the Leyden jar
could store. Its sudden discharge was used in many experiments
(O’Leary and Goldring, 1976).
The role of static electricity in medicine appeared to be for-
gotten for about 150 years and became resurrected with the
introduction of the defibrillating cardioversion by William B.
Kouwenhoven and his coworkers in the 1950s and 1960s; this
approach may hold promises for cerebral applications
(Niedermeyer, 2003a).
A serious scientific controversy developed in Italy between
Luigi Galvani (1737–1798), a professor at the University of
Bologna, and Alessandro Volta (1745–1832) in the wake of
Galvani’s discovery of frog leg contractions within an electrical
circuit and especially in the presence of a thunderstorm (1780).
Volta doubted the biologic nature of the contraction (animal
Part I Basic Principles
CHAPTER
1
electricity) and placed the emphasis on physics—on his “pile,”
the first battery (around 1800). This bimetallic pile was a gen-
erator capable of producing a steady flow of electricity. Volta’s
view more or less prevailed in this hotly debated argument. The
laws governing flowing electricity were soon discovered by
Georg Ohm in 1827.
Nevertheless, Galvani’s belief in “animal electricity” was not
lost with other discarded false ideas. There still remained the
nagging question of an active electrical contribution of animal
muscle tissue.
BEGINNINGS OF ELECTROPHYSIOLOGY
The introduction of the galvanometer has been associated
chiefly with the name of Nobili in Florence; this instrument was
refined in 1858 by William Thompson (Lord Kelvin) in
England (O’Leary and Goldring, 1976). These galvanometers
would faithfully demonstrate continuous electrical currents
and their variations in intensity but failed in the detection of
instantaneous electrical phenomena.
Carlo Matteucci (1811–1868) in Bologna and Emil Du Bois-
Reymond (1818–1896) in Berlin became the major proponents of
an electrophysiologically based physiology of the nervous system.
(The French name of Du Bois-Reymond indicates the Huguenot
origin of this Prussian investigator.) Du Bois-Reymond coined the
term negative variation for a phenomenon occurring during mus-
cle contraction when the galvanometer indicated an unexpected
decrease in current intensity (O’Leary and Goldring, 1976). This
term was later resurrected in earliest electroencephalogram (EEG)
research (Caton, 1875) and with the discovery of the “contingent
negative variation” (Walter, 1964).
Hermann von Helmholtz (1821–1894) accurately measured
the velocity of nerve conduction, which had been vastly overes-
timated up to that time. The electrodes used in physiologic
research were improved and made nonpolarizable (Du Bois-
Reymond). The concept of “action current” was introduced by
L. Hermann (1834–1919) and thus clarified Du Bois-Reymond’s
negative variations found during muscle contraction. Julius
Bernstein (1839–1917) proposed a membrane theory of nerve
tissue, which ultimately was elucidated as late as 1939 and the
following years by A. L. Hodgkin and A. F. Huxley in England.
Against this background of strongly evolving electrophysiology
of the nervous system, the first observation of EEG-like electri-
cal brain activity took place.
1
2 Part I ■ Basic Principles
CATON: THE FIRST ATTEMPT AT THE
ELECTRICAL ACTIVITY OF THE BRAIN
Richard Caton (1842–1926) (Fig. 1.1) was a physician practic-
ing in Liverpool who became deeply interested in electrophysi-
ologic phenomena and eventually received a grant from the
British Medical Association to explore electrical phenomena of
the exposed cerebral hemispheres of rabbits and monkeys.
According to Brazier (1961), Caton presented his findings to the
association on August 24, 1875, and a very short report of 20
lines subsequently appeared in the British Medical Journal. A
more detailed report was presented in the same journal in 1877
on experiments of more than 40 rabbits, cats, and monkeys, the
rabbit having been principally employed.
Caton used a galvanometer. A beam of light was thrown on
the mirror of the galvanometer and reflected on a large scale
placed on the wall. With this type of visualization, Caton found
that “feeble currents of varying direction pass through the mul-
tiplier when the electrodes are placed on two points of the
external surface, or one electrode on the grey matter, and one
on the surface of the skull.” This sentence is regarded as indicat-
ing the birth of the electrophysiologram because one can
assume that EEG phenomena made the needle move from one
direction to the other. (The suffix “gram” naturally is out of
place since “graphein” means “to write” and there was no writ-
ten recording.) Even though artifacts could have played a major
role, Caton deserves credit for the discovery of the fluctuating
potentials that constitute the EEG.
Figure 1.1 Richard Caton at the time of his work on the electrical
activity of the brain. (From Brazier MAB. A History of the Electrical
Activity of the Brain. The First Half-Century. London: Pitman; 1961, with
permission from Macmillan.)
Caton also described a few more interesting observations.
He noted that the external surface of the gray matter was posi-
tive in relation to deep structures of the cerebrum. He also
noted that the electric currents of the cerebrum appeared to
have a relation to underlying function: “When any part of the
grey matter is in a state of functional activity, its electric current
usually exhibits negative variation.” Thus, Caton has also been
credited with pioneer work on evoked potential. Furthermore,
the difference in polarity found between cortical surface and
deeper areas could be interpreted as the discovery of the “steady
potential” (“DC potential”), but it might be wise to refrain from
such statements that cannot be fully supported by the evidence.
With regard to the fluctuations, Geddes (1987) pointed out that
Caton’s galvanometer had a very limited frequency response
range from 0 to 6 Hz.
Caton found some measure of success and recognition with
this work and held the chair of physiology at the University
College of Liverpool from 1884 to 1891, when he resigned from
this post. Later he became dean of the medical faculty and, in
1907, mayor (Lord Mayor) of Liverpool. The electrical activity of
the brain did not occupy a predominant position in his further
endeavors. Even though Caton became an EEG research dropout,
his bold work will always remain a milestone in the history of the
electrical activity of the brain. (More information on Caton’s life
and work is found in Mary Brazier’s (1961) fine account.)
EASTERN EUROPEAN STUDIES OF
ELECTRICAL BRAIN ACTIVITY
The time was ripe for further studies of electrical phenomena of
the cerebrum. Concurrent with Caton’s epochal work of 1875,
physiologists of Eastern Europe began to demonstrate their
independent observations and discoveries concerning the brain
and its electrical activity. Another discovery of the 1870s had
an incomparably greater impact on the neuroscientific world
than Caton’s demonstration of electrical activity of the brain.
The capability of the human cerebral cortex to be electrically
stimulated was discovered by G. Fritsch (1838–1927) and Julius
Eduard Hitzig (1838–1907) in a joint study in 1870. According
to O’Leary and Goldring (1976), an unusual observation had
prompted Fritsch in his work: he had observed contralateral
muscle contractions during dressing of an open brain wound in
the Prussian–Danish War of 1864. The work of Fritsch and
Hitzig was furthered by D. Ferrier and G. F. Yeo in 1880, who
performed electrical stimulations of the cerebrum in apes and
also in a patient who was operated on for a brain tumor. The
repercussions of the stimulation studies were considerable since
many investigators of that time held the view that the entire
cerebrum is a homogeneous organ that harbors mental functions.
The response of the cortex to electrical stimulation probably
was a special incentive for the study of its spontaneous electri-
cal phenomena. This incentive was particularly strong in
Eastern Europe, that is, in laboratories of Russian and Polish
universities. (In spite of the important historical ethnic and
national differences, the fact cannot be ignored that most of
Poland was part of the Czarist Russian Empire throughout the
19th century.)

Vasili Yakovlevich Danilevsky (1852–1939) was only 25 years
old when he finished his thesis entitled “Investigations into the
Physiology of the Brain” (Danilevsky, 1877), written at the
University of Kharkov. This work was based on electrical stim-
ulation as well as on spontaneous electrical activity in the
brains of animals. Thus, Danilevsky walked in Caton’s foot-
steps; in 1891, he gave full credit to Caton’s priority. Mary
Brazier (1961) comments on the disappointment of Danilevsky
who saw his high hopes unfulfilled as far as the spontaneous
electrical activity of the brain was concerned; he had expected
better correlation with psychic and emotional processes. He
remained deeply involved in brain physiology and published an
extensive textbook of human physiology in 1915. He was not
the only EEG researcher with shattered hopes in the field of
psychophysiology.
The life and work of Adolf Beck (1863–1939) have been
described in great detail by Brazier (1961). Beck worked in
Kraków as well as in Lwow (the Polish province of Galicia, at
that time a part of the Austrian–Hungarian monarchy). With
nonpolarizable electrodes, Beck investigated the spontaneous
electrical activity of the brain in rabbits and dogs. He observed
the disappearance of rhythmical oscillations when the eyes were
stimulated with light and thus became a forebear of Berger’s
discovery of alpha blocking. His work became widely known
due to its publication in the Centralblatt.
To present a chronologic account of the events, let us leave
Eastern Europe for a moment, but not Vienna. In 1883, Ernst
Fleischl von Marxow (1846–1891) deposited a sealed letter at
the Imperial Academy of Sciences in Vienna that contained
observations on cerebral electrical activity recorded over the
visual cortex in various species of animals. He did not observe
oscillatory activity. He claimed priority when Beck in 1890 pub-
lished his data but was not aware of earlier work done by Caton
and Danilevsky. The oddity of this episode is underscored by
more recent historical accounts (Brazier, 1961; O’Leary and
Goldring, 1976) that indicate that Fleischl von Marxow’s work
was not of first-rate quality. This does not detract from the ren-
aissance-man versatility of this Austrian physiologist, who was
well versed in linguistics (even Sanskrit), swimming, hunting,
and mountain climbing.
Exciting new studies were being conducted in Eastern
European universities. Napoleon Cybulski (1854–1919), who
was Beck’s teacher in Kraków and an internationally renowned
leader in general physiology, presented experimental electroen-
cephalographic studies in graphical form by using a gal-
vanometer with a photographic attachment. He provided EEG
evidence of an epileptic seizure in the dog caused by electrical
stimulation.
Two Russian physiologists made further studies along these
lines: Pavel Yurevich Kaufman (1877–1951) and Vladimir
Vladimirovich Pravdich-Neminsky (1879–1952). Prior to the
discussion of their work, a few words must be said about tech-
nological developments. The d’Arsonval galvanometer featured
a mirror mounted on a movable coil; light focused on the mir-
ror was deflected when a current passed the coil. The capillary
electrometer was introduced by G. Lippmann and H. J. Marey
(for further details, see O’Leary and Goldring (1976)). One of
Chapter 1 ■ Historical Aspects of EEG 3
the most important advances was Willem Einthoven’s introduc-
tion of the string galvanometer in 1903, a very sensitive instru-
ment that required photographic recording and became the
standard instrumentation for electrocardiography at the turn of
the century.
Kaufman’s work and life are portrayed in Brazier’s (1961)
historical account. Kaufman expressed the view that an epilep-
tic attack would have to be associated with abnormal electrical
discharges, and he studied the effects of cortical electrical stim-
ulation. With World War I, he took the name of Rostoutsev and
worked mainly at the University of Baku.
Pravdich-Neminsky began recording electrical brain activity
of animals in 1912 with the string galvanometer. As Brazier
(1961) has pointed out, his recordings, published in 1912, were
the first pictorial demonstration of EEG and appeared 2 years
earlier than Cybulski’s tracings. Pravdich-Neminsky recorded
the EEG from the brain, the dura, or the intact skull of the dog.
He described a 12 to 14/sec rhythm under normal conditions
and marked slowing under asphyxia. Furthermore, he coined
the term electrocerebrogram (Fig. 1.2).
The achievements of the Eastern European neuroscientists
during those 50 years preceding the outbreak of World War I fill
us with awe and clearly demonstrate their special talent for elec-
trophysiologic neurophysiology. Limiting discussion to EEG
history can show only the tip of the iceberg. To assess the true
strength of their neuroscientific institutions, one must mention
investigators in somewhat related electrophysiologic areas. Ivan
Michailovich Sechenov (1829–1905) appears to be founder of
this powerful school of eminent neurophysiologists. He studied
the electrical activity of the spinal cord and oblongata in the
frog and was a predecessor of Pavlovian thought. Nikolai
Yevgenevich Wedensky followed Sechenov as the chair and pro-
fessor of physiology at St. Petersburg (known for the concept of
Wedensky inhibition). Vladimir Efimovich Larionov, also
working in St. Petersburg, conducted beautiful studies of the
auditory cortex in the dog.
The greatest Russian neuroscientist was also the most eminent
clinical neurologist of his country: Vladimir Mikhailovich
Bechterev (also “Bekhterev”) (1857–1927). He occupied the chair
of psychiatry in St. Petersburg, which included the field of clini-
cal neurology. He was a disciple of Du Bois-Reymond, Paul Emil
Flechsig, and Wilhelm Wundt, and also worked at Charcot’s clinic
in Paris. The influence of Wundt prompted Bechterev’s associa-
tive reflexology (I treasure his work “Allgemeine Grundlagen der
Reflexologie des Menschen,” Deuticke, Leipzig, 1926, even
though I can hardly agree with his “objective study of personal-
ity”). He combined his clinical work and private practice with
tireless psychophysiologic methods. A photograph with one of
his two assistants shows Bechterev wearing a thick winter coat in
his institute in Leningrad, giving testimony to the icy cold in the
unheated laboratory. “Functional anatomy of the brain, experi-
mental psychology and clinical neurology were the three fields in
which Bekhterev carved out a place for himself” (Yakovlev, 1953).
The Soviet regime had to choose between Bechterev and
Ivan Petrovich Pavlov (1849–1936), a physiologist who had won
the Nobel Prize in 1904 for his early work on conditioned
reflexes. Pavlov was the choice, and the magic of conditioned
4 Part I ■ Basic Principles

Figure 1.2 The first photographs to be published of electroencephalograms. In the upper record Neminsky shows (in the
third trace) the brain potentials of a curarized dog with the pulsations from an artery in the brain recorded above them.
In the lower record the sciatic nerve is being stimulated from time to time, and the decrease in activity noted by Neminsky
can be seen. The record reads from right to left, line I being a time marker in one fifth of a second, line III the galvanometer
string, and line V the signal for stimulation. (From Pravdich-Neminsky VV. Ein Versuch der Registrierung der elektrischen
Gehirnerscheinungen. Zbl. Physiol. 1913;27:951–960, with permission from Dr. Mary Brazier and Macmillan.)

reflex overshadowed all Soviet neurophysiology by highest
decree (even though Pavlov himself was highly critical of the
regime). The Pavlovian concept was closer to the ideology of
dialectic materialism, and this maxim with all its intolerant
dogmatism outlasted Pavlov’s death by two decades. This
ideopolitically governed form of neuroscience stifled all
progress of customary neurophysiology. The world leadership
in EEG and related fields quickly crumbled, accompanied by a
terrifying decline of a dogmatically governed neurophysiology.
DEVELOPMENTS IN WESTERN
AND CENTRAL EUROPE
Electroencephalographic research was in a dormant state in
Western and Central Europe while it was flourishing in Eastern
European countries. This is quite amazing because neurophys-
iology in general was healthy and well outside Russia and her
neighbors, but the ancestral lineage from Galvani to Du Bois-
Reymond and Caton broke off and the neurophysiological
field was watered by rivulets of different orientation. Thus, the
work of Fleischl von Marxow lies like an erratic patch in the
vast field.
The Western neurophysiologists followed attentively the
work of their neuroanatomical confreres and the great contro-
versy between network theories (the nerve cells forming a felt-
like net) and the neuron theory (the neuron representing a
unit). The net theory was supported by Joseph von Gerlach and
by Camillo Golgi, the discoverer of the silver chromate staining
method, but the neuron theory with its great principal propo-
nent Santiago Ramón y Cajal (1852–1934) proved to be victo-
rious in this struggle despite further attempts of new
generations of “reticularist” believers in a continuous network
of nerve cells.
In a similar manner, cerebral localizationists struggled
against antilocalizationists. In Germany, Friedrich Leopold
Goltz (1834–1902) removed the cerebral hemispheres in the
“dog without cerebrum” living in a state of extreme lethargy
and mental inertia (Goltz, 1888). H. Rothmann (1923) per-
formed similar investigations. This type of research was aimed
at the working of the cerebral hemispheres as a whole and de-
emphasized the aspects of cortical localization. As it was
pointed out previously, the cerebral stimulation studies of
Fritsch, Hitzig, Ferrier, and Yeo initiated a new era of interest in
cortical localization.
 Chapter 1 ■ Historical Aspects of EEG 5

In this period—the last third of the 19th and the dawning of

the 20th century—the work of Charles Scott Sherrington
(1857–1952), performed in Liverpool and Oxford, became
most influential in the development of a modern Western type
of reflexology. The Integrative Action of the Nervous System
(Sherrington, 1906) was based on a series of lectures held at Yale
University in 1904. The scope of this work reaches from reflex-
ology to decerebrate rigidity, from motor cortex to sensory
function, while the issue of mental function is being skirted
with the modesty of a truly great neuroscientist. Inhibition is
one of the great Sherringtonian discoveries. Even Ramón y
Cajal’s net of independent synaptically connected neurons
stood solely in the service of neural excitation. (Incidentally, the
term synapse was introduced by Sherrington.)
It is unfortunate that this greatest master of neurophysiology
stood miles away from electrophysiologic thought. His work
was based chiefly on ablation techniques. He may hardly ever
have given a thought to EEG methods even though he lived a
full active life. His disciples—to name only Edward Liddell and
Derek Ernest Denny-Brown—held similar views, while in
Cambridge electrically oriented neurophysiology found its
greatest proponent in Edgar Douglas Adrian (1889–1977),
whom we discuss later in his relationship to Hans Berger’s
work.

HANS BERGER AND THE HUMAN
ELECTROENCEPHALOGRAM
Hans Berger (1873–1941) (Fig. 1.3), the discoverer of the
human EEG, was a neuropsychiatrist. What neuropsychiatry
really meant in those years is poorly understood nowadays.
Neurology and psychiatry formed one specialty, one discipline,
in Germany, Austria, and a considerable number of other coun-
tries. Neuropsychiatric departments at university hospitals and
other institutions consisted of neurologic and psychiatric
floors; medical specialty training meant rotation from one dis-
cipline to the other, and a the head of one department was sup-
posed to be a master of both disciplines. Pure neurology was
just beginning to emerge as a special discipline in the German-
speaking countries (with the work of Wilhelm Erb
(1840–1921), Max Nonne (1861–1959), and the incomparable
Otfrid Foerster who, like Hans Berger, lived from 1873 to 1941).
Berger was not a leader, neither in neurology nor in psychi-
atry. Without his pioneering EEG work, his name would have
been forgotten. Biographic sketches (especially Kolle, 1956)
portray Berger as an extremely meticulous and conscientious
person, somewhat aloof in his contact with his patients, a very
strict and authoritarian department head, and an “anima can-
dida” (a pure soul), a hard-working professor without any
interest in faculty schemes and diatribes. He hardly ever
attended the annual meetings of the German neuropsychiatric
society. His electroencephalographic work was carried out in a
small and very primitive laboratory. His first scientific interest
aimed at the cerebral circulation; plethysmographic methods
were used in patients with skull defects. From 1902 to 1910, he
studied the electrical activity of the cerebrum in the dog with a
Figure 1.3 Hans Berger. (From Kolle K. Hans Berger. In: Kolle K, ed.
Grosse NervenŠrzte. Vol. 1. Stuttgart: Thieme; 1956: 1–6.)
capillary electrometer after Lippmann, but the results were dis-
appointing. Naturally, Berger was aware of the scanty pertinent
literature from Caton to Cybulski and Pravdieh-Neminsky. His
studies of the human EEG started in 1920; the introduction to
Gloor’s authoritative translation of Berger’s work contains a
plethora of interesting details (Gloor, 1969).
Every electroencephalographer should be familiar with
Berger’s work, an undertaking that has been greatly facilitated
by Gloor’s English translation. It is true that the original
German text is cumbersome and does not make easy reading,
which is probably a reason for the very slow acceptance of
Berger’s work. Those 14 reports bear the same title: “On the
Electroencephalogram of Man.” Surely, a more attractive title
would have helped somewhat. Berger’s humanistic educational
background becomes obvious in the rejection of the term elec-
trocerebrogram of Pravdich-Neminsky for strictly linguistic
reasons: the “ugly” mixture of Greek (“electro,” “gram”) and
Latin (“cerebro”) fragments. What Berger proposed in German
was the term Elektrenkephalogram (sic) since the root
enkephalo from the Greek is linguistically more correct than
encephalo.
Before Berger’s work as such is brought into focus, we must
discuss his electrophysiologic instrumentation. He used a string
galvanometer starting in 1910—first with the Einthoven type,
later with the smaller Edelmann model, and after 1924 with
the larger Edelmann model. In 1926, Berger started to use the
more powerful Siemens double-coil galvanometer (attaining a
6 Part I ■ Basic Principles
sensitivity of 130 V/cm; Grass, 1984). With this instrument
and the use of nonpolarizable pad electrodes, Berger recorded
the human EEG tracings shown in his first report of 1929.
The records were made on photographic paper with record-
ings from 1 to 3 minutes’ duration. Berger used a bipolar record-
ing technique with fronto-occipital leads for his one-channel
EEG tracings along with simultaneous electrocardiogram
(ECG) recording and a time marker. In 1932, he received an
oscillograph from Siemens but was unable to obtain further
amplifiers with oscillographs in order to obtain multichannel
recordings.
Studies of the human EEG began in 1924. Appointments
were made for several patients with large skull bone defects
(there was no scarcity of such patients in post-World War I
Germany). On July 6, 1924, the small Edelmann string gal-
vanometer showed oscillations presumably coming from the
underlying brain. In 1925, Berger recognized that skull defects
were not necessarily advantageous to obtaining a recording
because of thickening of dura, postoperative adhesions, etc.,
and he found that recordings could be made just as well (or
even better) through the intact skull and scalp. Between 1926
and 1929 Berger obtained good records with alpha waves; the
double-coil galvanometer was crucial for these observations.
The data were often uncertain and, in 1928, Berger was beset
with doubts concerning the authenticity of his observations
(according to his diary entries, very impressively demonstrated
by Jung, 1963).
The first report of 1929 features the alpha rhythm and the
alpha blocking response (naturally along with a description of
the smaller beta waves). Chlorinated silver needle electrodes,
platinum wires, and zinc-plated steel needles were used in those
years (Fig. 1.4). The bold first report of 1929 produced no
“waves” until a confirming report came from Adrian in
Cambridge (Adrian and Matthews, 1934). Throughout the
1930s, Berger’s reports on the human EEG contained veritable
gems: studies of fluctuation of consciousness, first EEG record-
ings of sleep (the first recording of spindles), the effect of
hypoxia on the human brain, a variety of diffuse and localized
brain disorders, and even an inkling of epileptic discharges.
Eventually, Berger was invited to an international congress of
psychologists in Paris in 1937 and to Bologna where the bicen-
tennial of Galvani’s birthday was celebrated (also 1937). His
relationship to the Nazi regime was not good, and Berger was
most unceremoniously made a professor emeritus at earliest
convenience, in 1938. This was indeed a hard blow to his plans
for further electroencephalographic studies and, in the wake of
a flu-like disease, he evidently developed a severe endogenous
Figure 1.4 The first recorded electroencephalogram of a human.
The lower line is a 10 cycles/ sec sine wave for use as a time marker.
The upper line is the recording from Berger’s young son made in
1925. (From Berger H. Über das Elektrenkephalogramm des
Menschen. 1st report. Arch Psychiat Nervenkr. 1929;87:527–570,
with permission from Dr. Mary Brazier and Macmillan.)
depression, which remained undiagnosed. He ended his life by
suicide on June 1, 1941 at the age of 68. External factors may
have contributed to his depression: in addition to his forced
retirement at 65 (a few additional “years of honor” were granted
to most retiring directors of university institutions), he also felt
challenged by a group of independent EEG workers at the
Institute of Brain Research at Berlin-Buch. This group was led
by A. E. Kornmüller and produced excellent experimental EEG
work, which is discussed later. Kornmüller might have had the
better connections to the government institutions in Berlin,
and the highly sensitive and often insecure Berger was afraid
that his discovery was being taken away from him by his more
aggressive colleagues in Berlin-Buch.
Berger was a very complex person and investigator. It was
pointed out previously that he did not excel clinically, neither
as psychiatrist nor as neurologist (even though he was very
interested in cerebral localization and particularly in the local-
ization of brain tumors). Berger also developed very unscien-
tific ideas about the nature of the EEG, even though he was a
meticulous scientist in his EEG work. The driving force in all
his research work was the quest for the nature of the all-pow-
erful force of mental energy (“psychische Energie”). An early
personal experience convinced him that such a mental
energy—even capable of transmitting thoughts and emotions
from person to person—does exist. According to Berger’s con-
cept, influenced by the Danish physiologist Alfred Lehmann in
1901, mental energy is thought to be a partial product of meta-
bolic energies (warmth and electricity being the other two
products). This concept gives the EEG waves the eerie charac-
ter of messengers within the mental activities, even as messen-
gers from person to person.
Friedrich Rueckert (1788–1866), a great German poet of the
Romantic period, was Berger’s maternal grandfather. Behind
the strict directorial facade of Berger was the gentle soul of a
highly vulnerable man. It was the psychophysiologist Berger
who searched for the correlate of mental energy and, on this
voyage, he found the human EEG. It was one of those
“Columbus syndromes”: that a discovery is made as a by-
product of a search for a different goal. Jung (1963) noted that
Berger pursued his goal with the extremely powerful energy of
a dilettante who had found a concept. Specialists, like the excel-
lent physiologist (and electrophysiologist) Wilhelm
Biedermann, in Jena were convinced that Berger’s dilettantism
would lead nowhere, but it was the dilettante and not the sea-
soned specialist who emerged victorious. Even though the EEG
is not exactly what Berger assumed, his contribution was the
greatest in the history of electroencephalography.

BERGER’S CONTEMPORARIES
The Berliner Group
The Institute of Brain Research (Hirnforschungs-Institut) in
Berlin-Bush harbored a group of ambitious and energetic
investigators in various neurosciences. Oskar Vogt
(1870–1959), one of the great neuroanatomists and neu-
ropathologists of his time and a remarkably independent
thinker with a wide intellectual horizon, was the director of the
institute. In 1936 he lost his “directorship for lifetime” when the
Nazi government became aware of Vogt’s activities at a similar
institute in Moscow and his reluctance to get rid of Jewish
coworkers (Hassler, 1959).
The Berliner Institute (a section of the Kaiser-Wilhelm
Institutes) was composed of a variety of departments. In this
context, the Department of Physiology under M. H. Fischer and
the Department of Electrophysiology under A. E. Kornmüller
must be singled out. These departments enjoyed the collabora-
tion of an outstanding physicist and electronic engineer, J. F.
Toennies (1902–1970), a personal friend of Oskar Vogt (both
coming from the town of Husum in Holstein).
Toennies built the first ink-writing biologic amplifier for the
recording of brain potentials. While in New York as a fellow of
the Rockefeller Foundation in 1932, he designed the differential
amplifier—the still all-important principle of EEG amplifica-
tion—but he shares this achievement with Brian Matthews,
Adrian’s ingenious coworker whose work is discussed in the
next section.
The collaboration with Toennies gave the Berlin group a
much better tool for EEG research in comparison with Berger’s
instrumentation. Kornmüller quickly recognized the impor-
tance of recordings from a greater number of electrodes. His
EEG studies in the humans placed particular emphasis on the
differences between given regions of the cerebrum
(“Hirnrindenfelder”) (Kornmüller, 1932, 1933, 1935, 1937). His
studies of the clinical significance of EEG (Kornmüller, 1944)
appear to be somewhat pale when compared with the impor-
tance of his earliest experimental EEG work carried out with
Fischer and also with H. Löwenbach, who later came to the
United States, where he became one of the earliest EEG pioneers.
In those early studies, the EEG was obtained from the cortex of
animals following poisoning with convulsive substances. This is
the first EEG work focusing on epileptic manifestations and the
first demonstration of epileptiform spikes (Fischer, 1933;
Fischer and Löwenbach, 1934a,b; Kornmüller, 1935).
Oskar Vogt (his wife Cécile also being known as a great neu-
ropathologist) developed a concept of strict cerebral compart-
mentalization in sharply separated areas. He showed indeed
extremely impressive boundaries between healthy and dis-
eased areas in the hippocampus (Vogt and Vogt, 1937) and
conceived the cortex as divided into about 200 regions with
precise demarcation from field to field. This concept became a
powerful leitmotiv for Vogt’s coworkers, and Kornmüller’s EEG
work clearly shows the marks of his authoritarian boss.
Richard Jung joined this group in 1937 when Vogt was
already fired. (The tycoon Alfred Krupp provided the Vogts
with a privately built institute in the Black Forest in his expression
Chapter 1 ■ Historical Aspects of EEG 7
of personal gratitude. Decades earlier, Vogt successfully treated
an ailing Krupp—with hypnosis. Indeed, Vogt also used to be
a master of this method!) Hugo Spatz replaced Vogt and con-
siderably changed the goals of research. Toennies stayed with
H. S. Gasser at the Rockefeller Institute in New York, where he
constructed the first cathode follower to record from high-
resistance electrodes. This was the birth of microelectrode
recording, which developed into an enormously powerful sci-
entific tool in the 1950s and 1960s.
Kornmüller’s work declined after World War II, when he
appeared to be obsessed with a totally unproven theory of glia as
the generator of slow brain potentials. Jung, however, developed
into one of the greatest electroneurophysiologists of his time.
Developments in Great Britain
Edgar Douglas Adrian (Baron of Cambridge and, as such, Lord
Adrian) (1889–1977) was not only one of the greatest electrophys-
iologic neurophysiologists of the 20th century; his name is also
intimately associated with the discovery of the EEG because of his
confirmation of Berger’s observations (Adrian and Matthews,
1934). He showed his colleagues his own beautiful alpha rhythm
and the blocking effect due to eye opening, but, alas, his great
electronic wizard, Brian Matthews, apparently had a low-voltage
EEG with no alpha rhythm. Due to some strange quirk, Adrian’s
recording from the head ganglion of a water beetle happened to
be indistinguishable from Adrian’s alpha rhythm and was blocked
in the same manner, namely by light falling on the beetle’s eyes.
The Adrian-water beetle similarity and the Adrian-Matthews EEG
dissimilarity must have been terribly confusing to the onlookers
(also see Adrian, 1936).
Adrian was already a neurophysiologist of great prestige
when he confirmed Berger’s data. He had been credited with
the demonstration of single sensory nerve fiber potential and
the analysis of unit activity, which resulted in the Adrian–Bronk
law (Adrian and Bronk, 1929). Incidentally, his collaborator
Detlef Bronk became president of Johns Hopkins University in
later years. Prior to Matthews, Keith Lucas had been Adrian’s
brilliant electronic engineer and experimental coworker.
W. Grey Walter became the pioneer of clinical electroen-
cephalography in England, and his discovery of foci of slow
activity (delta waves, named by Walter) generated enormous
clinical interest in the new method. Grey Walter, however, was a
Ph.D. basic scientist. The fact that he was principally an aca-
demic scholar, not a clinician, could have laid the foundation for
the aversion of England’s great neurologists toward the method
of EEG, which in the following years was either ignored or left to
Ph.D. electroencephalographers in the laboratory. We will find
Grey Walter again in later decades; let it be said that he was one
of the most brilliant minds in all neurosciences—an independ-
ent thinker, a powerful writer, and quite often a man nearly con-
sumed by the flame of his own brilliance. He founded a small
but very effective school in Bristol at the Burden Institute.
Developments in France and Belgium
Thus far, France seems to be unduly neglected in this historical
overview. It had its own proud neurophysiologic schools in the
19th century, and the names of François Magendie (1783–1855)
8 Part I ■ Basic Principles
and, above all, Claude Bernard (1813–1878) belong in the pan-
theon of neuroscience.
A fine school of early electroencephalographers developed in
Paris in the 1930s. A. Fessard at the Collège de France must be
singled out as the most towering figure. Together with G.
Durup, he also confirmed the results of Berger. Durup and
Fessard even used EEG in the study of conditioned reflexes.
Clinical electroencephalography started in France under the
aegis of A. Baudouin and G. Fischgold. Fischgold had come
from Rumania, developed into a leading clinical electroen-
cephalographer, and—what an unusual combination—became
a leader in neuroradiology. Baudouin was the key figure in
the invitation of Berger to Paris in 1937.
Neighboring Belgium was the home of a giant in electro-
physiologic neurophysiology: Frederic Bremer (1892–1982) from
the Université Libre of Brussels. Bremer quickly recognized the
usefulness of EEG methods in the experimental investigation of
the brain. He recognized the influence of afferent signals on the
state of vigilance and compared his feline preparation called
“cerveau isolé” (with midbrain transection) with the “encéphale
isolé” resulting from transection at the boundary between the
medulla oblongata and cervical cord. The former preparation
would produce permanent coma; the latter would cause a vari-
able state of vigilance, with waking and sleeping demonstrated
on the EEG recording. In other words, trigeminal-sensory, audi-
tory, visual, and probably also olfactory influences would help
to keep the (artificially ventilated) encéphale isolé preparation in
a waking–sleeping rhythm (Bremer, 1935).
The greatness of this investigator must be reemphasized,
especially in today’s era of short memory. Whoever reads his
study entitled “Cerebral and Cerebellar Potentials” (Bremer,
1958) will roughly understand the dimensions of this
neurophysiologist.
Developments in Other European Countries
Italy was one of the first countries where the EEG found fertile
soil. Mario Gozzano, for many years professor of clinical neu-
rology in Bologna (later in Rome), published his experiences
with the new method as early as 1935. Gozzano personifies the
(not too common) example of a leading neurologist assuming
leadership in clinical electroencephalography (Mazza et al.,
2002). All too often, eminent clinical neurologists spurned the
new method.
A. Gemelli came from the diametrically opposite area of
neurosciences. This great scholar was a monk, psychologist,
philosopher, polyhistor, and president of the Catholic
University in Milan. In 1937, he reported his first studies of the
human EEG. Gemelli hence represents the psychological wing
of EEG research, which subsequently spawned a number of
outstanding Ph.D. electroencephalographers. (Others would
come from the ranks of experimental neurophysiologists.)
The Austrian psychologist Hubert Rohracher falls into the
Gemelli category. He held the chair of psychology for many
years at the University of Vienna, but, in his early academic
work, he fell under the spell of the alpha rhythm and even made
a “pilgrimage” to Hans Berger in Jena (in the 1930s). His early
EEG studies can be dated back to 1938.
America Enters the Scene
Around 1935, the center of gravity in the still modest bulk of
EEG work started to shift from Europe to North America.
Fascinating new reports came from the United States. European
investigators started to travel across the Atlantic, and even Hans
Berger was about to accept an invitation to the United States in
1939, when the beginning of World War II thwarted his plans.
In the pre-Berger development of experimental EEG studies,
America had not played any role. Schwab (1951) reports that, in
1918, a medical student of Harvard Medical School, Donald
McPherson, worked under the eminent physiologist Alexander
Forbes. When McPherson placed two electrodes on the exposed
brain of a cat and ran the output into a string galvanometer, he
saw rhythmical 10/sec EEG activity. This finding was rejected as
an artifact by Forbes. Was Forbes completely unaware of the
work from Caton to Pravdich-Neminsky?
The rise of American EEG work to international fame is cus-
tomarily associated with the work of Hallowell Davis, Frederic
A. Gibbs, and Erna Gibbs at Harvard and also with Herbert
Jasper’s work at Brown University in Providence, Rhode Island.
According to O’Leary and Goldring (1976), A. J. Derbyshire, a
graduate student of Hallowell Davis, brought Berger’s paper of
1929 to Davis’s attention. Derbyshire, Pauline Davis, and H. N.
Simpson then tried in vain to demonstrate their own alpha
rhythms. There were finally shouts of joy when Hallowell Davis
himself was found to have a good alpha rhythm. Otherwise, the
first human EEG study in America would have been a negative
one. This work was done in 1934, just before human EEG stud-
ies started to mushroom in the United States.
The EEG, however, had been used for animal experiments
for some years in the United States, starting with Bartley and
Newman (1930, 1931) and Bartley (1932), who produced EEG
tracings in the dog. Howard Bartley did his work at Washington
University in St. Louis, a place that had already developed into
a hotbed of neurophysiology due to the magnificent work of
Herbert S. Gasser, Joseph Erlanger, and George Bishop—a
group that made excellent use of Braun’s cathode ray oscillo-
scope (oscillograph) in the study of peripheral nerve potentials.
This outstanding group was joined later by James L. O’Leary, a
prominent neurophysiologist, electroencephalographer, and
neurologic clinician. Early experimental EEG work was done by
Davis and Saul (1931), Travis and Dorsey (1932), Travis and
Herren (1931), Bishop and Bartley (1932), Bartley (1932), and
Gerard et al. (1933) (after Grass, 1984). The work of Ralph W.
Gerard (1900–1974) is linked with the introduction of a con-
centric needle electrode for the stereotaxic exploration of the
brain in experimental animals. Gerard joined forces (in 1934)
with Franklin Offner, one of the leading electronic engineers in
the development of EEG and related equipment.
American EEG work in the human started, as it was pointed
out before, at Harvard in Boston (Hallowell and Pauline Davis,
Frederic and Erna Gibbs, and William G. Lennox), at Brown in
Providence (Herbert H. Jasper), but also at the University of
Iowa in Iowa City where Lee Travis worked, an experimental
psychologist who became the founder of a powerful school
(Herbert Jasper, Donald Lindsley, John R. Knott, and Charles
Henry).

The great international breakthrough in clinical electroen-
cephalography came in 1934 with studies of epileptic patients.
Frederic Gibbs had come from Johns Hopkins University in
Baltimore to join the Harvard group. He sought out William G.
Lennox, who had already become a widely known epileptologist.
It might be interesting to point out that Lennox had started
studies of the cerebral circulation by measuring the O2 and CO2
content of the jugular veins (Lennox, 1930, 1931; Lennox and
Gibbs, 1932). E. L. Gibbs was Erna L. Gibbs, originally the tech-
nical coworker of Lennox but who became the wife of Frederic
Gibbs and one of the world’s first EEG technicians and the coau-
thor of numerous papers. She had come to Boston as an immi-
grant from Germany. The pre-EEG work of Lennox and Gibbs
on the cerebral blood flow was a milestone in this field. (One of
the great present-day masters of cerebral blood flow, Louis
Sokoloff from the National Institutes of Health, expressed to me
in a personal communication his profound admiration for this
pioneering work.) EEG simply exerted a greater degree of fasci-
nation to W. G. Lennox than did cerebral blood flow.
Twelve children with petit mal epilepsy were the clinical sub-
jects for the petit mal epilepsy study of Gibbs and Davis (1935)
and Gibbs et al. (1935, 1937). This work remains evergreen in
the entire EEG literature; hardly any EEG finding has left such
an indelible impression as the association of petit mal absences
and 3/sec spike-wave complexes. (Of course, it was found out
later that spike waves could occur without petit mal.) While
Berger was gripped by the rhythms, Frederic Gibbs came under
the fascination of paroxysmal patterns such as spike waves.
Shortly afterward, the EEG patterns of grand mal and psy-
chomotor seizures were reported by the same team (Gibbs,
Lennox, and Gibbs), but the stretches of fast spikes (in grand
mal) and the rhythmical activity in 4 or 6/sec frequency (in psy-
chomotor seizures) were no match in popularity for the 3/sec
spike waves of petit mal.
The technical quality of the EEG tracings shown in these
studies left much to be desired. Dr. and Mrs. Gibbs traveled to
Germany in the summer of 1935, paid a visit to Hans Berger,
spent some time at the Berlin-Buch Institute, and studied the
“polyneurograph” instrument of Jan F. Toennies; they also saw
the instrumentation of Matthews in England. Frederic Gibbs
then contracted Albert Grass (then at the Massachusetts
Institute of Technology) to build a three-channel preamplifier.
In 1935, the Grass Model I went into use; it had three channels
and an ink writer that recorded on rolls of paper (the folded
paper not yet being in use).
The Gibbs–Gibbs–Lennox era of the 1930s proved to be per-
haps the most exciting period in the history of EEG. In those
years, EEG found the domain of greatest effectiveness: the
realm of the epileptic seizure disorders. Epileptology can
be divided historically into two periods: before and after the
advent of EEG. Insights into the nature of the epileptic mecha-
nisms deepened, not in a subtle manner but with a huge leap.
What Fischer had started in 1931 with his experimental studies
on picrotoxin and its effect on the cortical EEG in animals, the
Gibbses and Lennox applied to human epileptology, and a wide
door was flung open for the work of future decades. It is true
that Berger in his seventh report (Berger, 1933) had shown a
Chapter 1 ■ Historical Aspects of EEG 9
few examples of paroxysmal EEG discharges in a case of pre-
sumed petit mal attacks and also during a focal motor seizure
in a patient with general paresis. These observations were just
mentioned in passing and the opportunity of a major break-
through was missed.
As to the other great pioneers of electroencephalography in
North America, Hallowell and Pauline Davis produced fine
work on the normal EEG and its variants. They were also
among the earliest investigators of the human sleep EEG. In the
domain of sleep, A. L. Loomis and his coworkers E. N. Harvey
and G. A. Hobart were the first who methodically studied the
human sleep EEG patterns and the stages of sleep. This research
was done off the academic track in Tuxedo Park, New Jersey
(Loomis et al., 1935, their first study). The Davises eventually
turned to audiology and moved to St. Louis. At Brown
University in Providence, Rhode Island, Jasper studied the EEG
of behavior disorders in children before he found his niche in
basic and clinical epileptology at McGill University in Montreal
in his epochal collaboration with Wilder Penfield (discussed
later). Lee Travis gradually disappeared from the scene but his
foremost disciples, John R. Knott and Charles E. Henry (Ph.D.
electroencephalographers with strong clinical inclinations)
were bound to assume a very important role in America’s EEG
work. Their ultimate skill and supreme dedication turned them
into the “conscience of EEG,” steering developments into the
right direction and correcting the course when there was dan-
ger of going astray. D. Lindsley became one of the pioneers in
the investigation of maturational EEG aspects; at the University
of California at Los Angeles, he directed excellent neurophysio-
logic EEG research.
This was the first wave of American EEG pioneers and their
immediate disciples and followers. It is impossible for the his-
torian to do justice to the second wave, which started before the
decade of the 1930s was over. There was Robert Schwab at
Harvard and at the Massachusetts General Hospital in Boston,
in whom the mastery of EEG was combined with great clinical
neurologic talents (especially in the field of myasthenia gravis,
parkinsonism, and epilepsy). Across the Charles River, at the
Massachusetts Institute of Technology, there was Warren
McCulloch, a fiery genius like Grey Walter in Bristol and a pro-
found thinker. His scope would range light years beyond the
limits of EEG and neurophysiology. (One must read his
Embodiments of Mind to fathom his greatness, even though one
may be inclined to disagree in many points.) He and Grey
Walter lived in the world of brain machines, but there was
still a niche for a psyche (when one tries to read between the
lines). Earlier at Northwestern University in Evanston, Illinois,
outside Chicago, McCulloch had been involved with Dusser de
Barenne in “neuronographic” work, an import from Utrecht,
Netherlands; this work was based on topical strychnine poison-
ing of the cortex and exploration of transmitted spiking to other
regions.
Clinical EEG research already started to conquer certain
fields outside epileptology. Grey Walter’s discovery of the delta
focus (Walter, 1936) located over hemispheric brain tumors
had opened the search for further relationships between brain
lesions and focal EEG correlates; metabolic disturbances and
10 Part I ■ Basic Principles
especially hypoglycemia were explored with EEG. (The work of
H. Hoagland and his coworkers dates back to 1937.)
With the end of the 1930s, North America found itself in a
leading position in the domain of EEG, whereas progress in
Europe was quite limited.
World War II and the 1940s
During World War II, from 1939 to 1945, research and clinical
EEG activities were not flourishing, particularly in Europe.
There were some neurologic units where the EEG was used in
the localization of traumatic brain lesions and epileptogenic
foci. After World War II, the gap between North America and
Europe was bigger than ever before, and European EEG
research found itself at a low point.
After the war, new activities started in England and France,
while the situation in Germany looked desperate.
W. Grey Walter with his associates V. J. Dovey and H. Shipton
(a brilliant electronic engineer who later moved to Iowa City
and then St. Louis) at the Burden Institute in Bristol discovered
the paroxysmal response to flickering light at critical frequen-
cies between 10 and 20/sec. Further work on epileptic photo-
sensitivity immediately shifted from Bristol to Marseille,
France, where a young and incredibly talented Henri Gastaut
used this method, in combination with intravenous dosages of
pentylenetetrazol, to determine the “seuil épileptique,” that is,
the individual threshold for paroxysmal responses (Gastaut,
1949; Gastaut et al., 1948).
In 1947, the American EEG Society was founded and the
First International EEG Congress was held in London; a second
one followed in 1949 in Paris (in association with clinical neu-
rology and other neurologic disciplines). EEG activities in
Germany were still minimal; Japan, however, gained attention
by the work of K. Motokawa, a researcher of EEG rhythms.
Switzerland started to develop its own profile; the neurophysi-
ologist Marcel Monnier was instrumental in this regard. W. R.
Hess, however, a Nobelist, had gained great prestige by the func-
tional mapping of thalamus and hypothalamus with regard to
autonomic responses to electrical stimulation.
The American scene was bustling with activities. Frederic A.
Gibbs with his coworkers Erna Gibbs and B. Fuster from
Uruguay produced another epochal study on the interictal
anterior temporal spike or sharp wave discharge in the inter-
seizure interval in patients with psychomotor seizures. This was
an important step in the elucidation of temporal lobe epilepsy,
a work with far-reaching consequences for the entire develop-
ment of EEG laboratories and their routine work. It was found
(Gibbs et al., 1948) that the anterior temporal discharges were
often limited to the state of sleep. This observation meant that
a tracing without a sleep portion could be insufficient, uninfor-
mative, and even misleading. Thus, EEG laboratories would
include sleep in most (if not all) of their EEG evaluations. This
required pasted electrodes (rather than rubber bands or caps),
a much longer recording time, and a much smaller numerical
output of recordings per technician (incidentally, an evolving
profession, which is discussed later).
Transatlantic communication was poor at that time, and
it was at this point when the routine work in American
(or Canadian) EEG laboratories started to become more
sophisticated than that of their European counterparts because
of the inclusion of sleep.
Frederic Gibbs enjoyed enormous international prestige at
that time as the world’s leader in clinical electroencephalogra-
phy. Nevertheless, his position at Harvard was much less presti-
gious; he held the academic rank of an instructor (below the
professorial ranks), even though a visit to his laboratory was the
goal of European colleagues (who could afford the trip). This
disproportion drastically shows the negative attitude toward
EEG in neurologic departments (not universally, of course).
Robert Schwab did not fare much better at Harvard in spite of
his fine clinical neurologic talents.
Toward the end of the 1940s, Herbert H. Jasper turned into
a strong competitor of Frederic Gibbs. Jasper had moved to the
Neurologic Institute of McGill University in Montreal, joining
forces with Wilder Penfield, a neurosurgeon with a profound
neuroscientific background. We discuss the rise of the Montreal
group in section Developments in the 1950s.
Two new developments started in the late 1940s. The EEG
technique started to become invasive and, with the use of spe-
cial depth electrodes, the exploration of deep intracerebral
regions began. This is discussed in section Developments in the
1950s. Automatic frequency analysis also started in the 1940s,
but this development reached loftier heights in the 1960s.
A discussion of the 1940s would be incomplete without a
brief glance at the work of the neurophysiologists. A large seg-
ment of neurophysiologic work was dominated by the use of
EEG. One of the most fascinating results of these researchers
was the demonstration of thalamocortical relationships, thus
far explored solely with anatomical methods (e.g., the study of
the thalamus by A. Earl Walker in 1938, which propelled this
young neuroscientist to great fame for decades to come). The
work of Morison and Dempsey (1942) on the recruiting
response had great impact on the neuroscientific world with the
demonstration of cortical responses to relatively slow stimula-
tion of the intralaminar structures of the thalamus in the cat.
This work emphasized the role of the thalamus in the cortical
electrogenesis and broke the ground for the concept of a “cen-
trencephalic epilepsy,” a concept promoted by Penfield and
Jasper in Montreal (somewhat naively understood as a concept
of the thalamic origin of primary generalized epilepsy).
Even greater was the impact of the work of Horace W.
Magoun, who had studied the effects of descending and mostly
inhibitory influences of the brainstem reticular formation dur-
ing his work at Northwestern University. Together with G.
Moruzzi (a fine neurophysiologist from Pisa, Italy, and investi-
gator of basic epileptic mechanisms), Magoun subsequently
studied the ascending system of the brainstem reticular forma-
tion (chiefly in the midbrain level) and the effect of high-
frequency electrical stimulation, consisting of EEG
desynchronization and behavioral arousal, on cortical function.
Magoun, who had moved to the University of California at Los
Angeles, subsequently investigated the effects of acute lesions
made in the midbrain-level reticular formation in cats. These
cats remained in a comatose state with EEG slowing in spite
of electrical brainstem stimulation because of the destruction

of the all-important ascending portion of the brainstem reticu-
lar formation (Lindsley et al., 1949). It is no exaggeration when
one describes the effect of these studies on the world of neuro-
science as a “bombshell.” For the ensuing 10 to 15 years, the
association of consciousness with reticular formation and
Magoun’s name was so strong that it even had considerable
influence on the Pavlovian dogmatism of the Eastern Bloc
countries. Nowadays, however, even talented young neuroscien-
tists react to Magoun’s name with a blank expression—sic transit
gloria mundi!
The reason for discussing this experimental work in a histor-
ical overview is to demonstrate the incredibly powerful role of
EEG in the neurophysiology of the 1940s. This was a high
watermark. Subsequently, experimental EEG work started to
concentrate on single neurons while the “macro-EEG” gradu-
ally declined.
Developments in the 1950s
This is the last decade presented for historical analysis. Our story
is gradually approaching the present, and a historical outline
must shy away from events that occurred over the last 40+ years.
It does not behoove the historian to place living and active col-
leagues into the focus of discussion (with few exceptions).
The 1950s was the decade when EEG became a household
word. During the early stretch of the decade, almost every uni-
versity (teaching) hospital had at least one EEG machine. At the
end of the decade, EEG apparatuses had found their way into a
large number of other hospitals and even into private practice.
At university hospitals, central as well as departmental EEG lab-
oratories emerged. The latter were usually limited to children or
adults, and pediatric EEG units evolved (while specialized neona-
tologic EEG units followed suit about 10 years later). Some psy-
chiatric departments took particular pride in their clinical and
research-oriented EEG work. It is absolutely true that psychiatry
was always “nice” to the electroencephalographer. Psychiatry’s
domain was in need of organic or neurophysiologic substrata of
disorders and dysfunctions of psychiatric–psychological nature.
What could the electroencephalographer give in return? It was
very little, but the psychiatrists did not seem to mind. On the
other hand, there was so much to give to neurology. At that
time, it had become clear that the majority of diseases affecting
the central nervous system (CNS) had more or less impressive
EEG correlates, but the majority of neurologists remained
either reserved or hostile to EEG. Neurosurgeons were inter-
ested as long as EEG could contribute to the determination of
focal cerebral lesions (and before EEG became overpowered by
noninvasive neuroimaging techniques). Some epilepsy-ori-
ented neurosurgeons like W. Penfield or A. Earl Walker
remained interested in EEG and its use in the depth of the cere-
brum or on the cortex.
The epileptologic EEG work of Herbert Jasper in collabora-
tion with Wilder Penfield reached new heights, and Montreal
reigned supreme as the place for neurosurgical treatment of
focal epilepsies. Penfield was far more than a neurosurgeon. His
operations for the removal of epileptogenic foci and, in a later
phase, large portions of affected lobes were associated with elec-
trical stimulation and a systematic study of the behavioral
Chapter 1 ■ Historical Aspects of EEG 11
effects. At that time, local anesthesia was still widely used in
neurosurgery. Jasper was chiefly a neuroscientist and not
merely an electroencephalographer. The book entitled Epilepsy
and the Functional Anatomy of the Human Brain (Penfield and
Jasper, 1954) was a result of this fruitful collaboration.
Very controversial, however, was a concept of the primary
generalized form of epilepsy characterized by generalized syn-
chronous paroxysmal EEG discharges and exemplified by the
3/sec spike waves of petit mal absences. Penfield and Jasper
listed these epilepsies as “centrencephalic” with the concept of
“center of the encephalon” (i.e., “thalamic midline structures”)
serving as the starting point of the bilateral discharges. Henri
Gastaut from Marseille would follow the lead and so did many
others, but Frederic Gibbs and a host of other electroen-
cephalographers and neuroscientists became detractors of the
centrencephalic concept. It wasn’t until the late 1960s that it
became clear that the centrencephalic concept stood on a very
shaky ground and was ripe for being dismantled. Montreal’s
own Pierre Gloor helped to do this in a cautious and diplomatic
manner; others buried the centrencephalic concept more
bluntly.
Frederic Gibbs had moved to the University of Illinois
School of Medicine in Chicago (where full professorship was
given to him instantly after Harvard had denied him any pro-
motion for more than a decade). Chicago—especially the
University of Chicago but also Northwestern University and the
University of Illinois—had become a world leader in neurologic
sciences over the past 20 years. Percival Bailey, Paul Bucy, Roy
Grinker, A. Earl Walker, Gerhardt Von Bonin, C. J. Herrick,
Frederic Gibbs, and many others give testimony to the glory of
neurologic science in Chicago at the middle of that century. In
the field of EEG, the Chicago group under the Gibbses and the
Montreal group under Jasper and Gloor were strong rivals
throughout the 1950s, especially with respect to leadership in
epileptologic electroencephalography. One of the greatest mas-
ters from the Chicago school, A. Earl Walker, came to Johns
Hopkins in Baltimore in 1947, introducing depth EEG, electro-
corticography, epilepsy surgery, and a scientifically oriented
epileptology to his new place. Walker, who in 1972 had moved
to the University of New Mexico and died in 1995, will always
be remembered as one of the great scholars of neurosurgery
and epileptology.
The 1950s saw a strong comeback of the Europeans. Henri
Gastaut’s intellectual brilliancy was hard to match. In Marseille,
disciples of great stature flocked around him, especially Robert
Naquet, Joseph Roger, and Annette Beaumanoir, to mention
only the earliest nucleus of this group. At the great world cen-
ters of neurology, Salpêtrière Hospital in Paris and National
Hospital, Queen Square, London, Antoine Remond and
William Cobb, respectively, represented the EEG, but unfortu-
nately too much in the shadow of the leading neurologists.
Remond later turned into a protagonist of computerization of
EEG data.
The star neurologists of both Queen Square and Salpêtrière
lived in the world of classical neurology, which gave them so
much satisfaction and happiness that one could hardly expect
their openness for the world of brain potentials. The Queen
12 Part I ■ Basic Principles
Square guard appeared to be more detached from EEG than
their Parisian confrères, perhaps due to the fact that Gastaut,
the man from Marseille, came from the neurologic ranks to
achieve instant stardom with his EEG achievements. Probably
no other famous neurologist has expressed his opinion about
EEG more scathingly than Francis M. R. Walshe has done. He is
the brilliant Queen Square star who apparently knew every-
thing about neurology except EEG. To Sir Francis, the electrical
activity of the brain was, “a bloodless dance of action potentials
… hurrying to and fro of its molecules” (after Critchley, 1990).
Let us assume, for everyone’s benefit, that Sir Francis had meant
it to be a joke.
Fine schools of EEG developed in the Netherlands with O.
Magnus (son of a Nobel Prize winning physiologist) in
Wassenaar and Storm van Leeuwen in Utrecht. In Switzerland,
Rudolf Hess (also son of a Nobelist physiologist) created an
important school of electroencephalographers at the Zurich
University Hospital. Giuseppe Pampiglione, from Italy, initiated
pediatric electroencephalography at the Hospital for Sick
Children in London, concurrently with William Lennox’s work
at the Boston Children’s Hospital.
These European centers of EEG activities had rather an
international flavor with strong North American influence.
This cannot be said about the evolving field of clinical EEG in
West Germany, which was dominated by its prestigious leader
Richard Jung in Freiburg. Jung’s greatness pertained to experi-
mental neurophysiology; he also had great interest in the clini-
cal fields and even in philosophy. This renaissance man
designed the outline for EEG training and the routine of the
EEG laboratory—unfortunately not without shortcomings,
which hamstrung the further development of clinical EEG in
West Germany.
The 1950s also saw EEG sprouting into related fields.
Depth electroencephalography with implanted intracerebral
electrodes was used in the humans for the first time by
Meyers and Hayne (1948) and Knott et al. (1950) at the
University of Iowa, Iowa City, and also by Hayne et al.
(1949a) in Chicago. These short recordings served the study
of EEG activity in the human basal ganglia and thalamus
with regard to basal ganglia dyskinesias and epilepsy. In the
following years, deep structures were also explored in
patients with psychiatric disorders until doubt was cast upon
the ethical basis of this invasive approach (see Chapter 33,
“Depth Electroencephalography”). In the 1960s, depth EEG
would find its true field in epileptic patients considered can-
didates for epilepsy surgery.
The origin of intraoperative electrocorticography dates back
to Foerster and Altenburger (1935). How was it possible that
Otfried Foerster, perhaps the greatest clinical neurologist ever
and an amazing self-taught master of neurosurgery, failed to
recognize the future potential of EEG? Did his mind work
mainly in the world of Sherringtonian concepts? Most of the
work in electrocorticography remains associated with the
Montreal Neurologic Institute and the names of Penfield and
Jasper (also see Chapter 34, “Electrocorticography”).
The related fields of EEG started to bloom in the 1950s. In a
study entitled “A Summation Technique for Detecting Small
Signals in a Large Irregular Background,” George D. Dawson
from the National Hospital, Queen Square, in London demon-
strated evoked potentials to electrical stimulation of the ulnar
nerve (Dawson, 1951). This required advanced analog technol-
ogy. Thus, Dawson became the father of evoked potential stud-
ies, which developed into a major outcropping of
electroencephalography, eventually constituting a field of its
own. The ingenious superimposition method of Dawson was
eventually superseded by the advent of computerized averaging
methods in the 1960s.
Computational techniques of wave analysis started early in
the history of EEG. First attempts were made by Hans Berger
(1932); he was assisted by the physicist Dietsch (1932), who
applied Fourier analysis to short EEG sections. Further work in
this field was produced by Grass and Gibbs (1938) and Knott
and Gibbs (1939). At the Massachusetts Institute of Technology
near Boston, Guillemin applied Fourier analysis to communica-
tion theory, and one of his students was Albert Grass who
“could not wait to get the Gibbs interested” (Grass, 1984). The
1950s saw the early generation of automatic frequency analyz-
ers approaching and eventually saw the end of these magnifi-
cent but mostly unused machines.
Eventually, the EEG branched out into the world of single
neurons, and the microelectrode technique was introduced in
the early 1950s. Microelectrodes can be made of metal such as
tungsten with tips of 1- to 3- m diameter; glass electrodes
filled with electrolytes such as KCI have tips of 0.5 m or even
smaller. Because of their characteristics, microelectrodes reach
very high impedance values (1–60 M ), which render conven-
tional EEG recording techniques unsuitable. The introduction
of the cathode follower by Toennies created the technical pre-
requisite for single-cell recordings.
Extracellular microelectrode recording was used on a larger
scale in the early 1950s (Jung et al., 1952; Li et al., 1952;
Moruzzi, 1952). About 10 years later, extracellular microelec-
trode studies were even done intraoperatively in humans. Far
more revolutionary was the introduction of the extremely labo-
rious intracellular microelectrode technology (Brock et al.,
1952, in the spinal cord; Phillips, 1961, in the cortex). This tech-
nique opened the gates to a new world of biochemical
processes. These insights taught lessons in humility to the elec-
trophysiologic neurophysiologist. There was no doubt that the
chemical changes were of primary significance, while the elec-
trical phenomena were more or less by-products.
We cannot leave the 1950s without mentioning epochal
developments in the field of sleep research. At the University of
Chicago, N. Kleitman stood out as one of the world’s leading
investigators of the organization of sleep. This institution pro-
duced the first study of rapid eye movement (REM) sleep
(Aserinsky and Kleitman, 1953), but it must be pointed out that
Blake and Gerard (1937) described a “null stage” in the EEG of
nocturnal sleep, thus indicating the desynchronization of EEG
in REM sleep but without observation of the accompanying
ocular, muscular, and other autonomic changes. William C.
Dement continued the work of Kleitman and, following his
move from Chicago to Stanford, became a world leader in the
study of nocturnal sleep.

Sleep research gradually became based on polygraphic
recording, and its share in the overall EEG research declined.
This development led to a constantly widening gap between
EEG and nocturnal sleep research (in the 1960s and the follow-
ing decades).
The Rest of the Story
The last 30 years of the history of EEG and related fields can be
gleaned directly from this book. The events of the 1960s, 1970s,
and 1980s are just too close for us to see with the eyes of the his-
torian. Nevertheless, modern trends are briefly discussed in this
final section.
The development of clinical and experimental EEG work
reached a high point around 1960 after 30 years of steady
progress. There is no doubt that the 1960s slowed down the
smooth progress. The interest of electroencephalographers in
academic institutions tended to shift from the tracing, with all
its waves and patterns, to automatic data analysis.
Computerization was the direction—tendencies reaching back
to Berger’s coworker Dietsch (1932), but flourishing in the
1960s and 1970s. This development had many positive aspects.
The names of Barlow, Brazier, Remond, Lopes da Silva,
Bickford, Saltzberg, Dumermuth, Matousek, D. O. Walter,
Cooper, Künkel, Lehmann, Gasser, Burch, Hjorth, Schenk,
Matejcek, and Low should be mentioned in this context. In par-
ticular, Cooley and Tukey (1965) have been credited with the
introduction of the fast Fourier transforms as the basis of
power spectral analysis.
This work led us into a “brave new world” of EEG comput-
erization and, as early as in 1967, we were told that custom-
ary EEG reading would soon be a thing of the past, replaced
by a fully automatic EEG interpretation. The fears of many
clinical electroencephalographers were unfounded; nobody
became jobless, because such an automatization of EEG read-
ing was fictional. It was found that EEG is far too complex for
such automation. Its interpretation requires the wonderful
computer that is located between the ears—the educated and
experienced brain. One simply must consider that the meth-
ods of those years—frequency or time domain—were limited
to an analysis of frequencies. Automatic spike detection had
barely reached its earliest stage.
The electroencephalographer needed to be aware that all
types of data computerization were nothing but the EEG in
disguise—“an analog of an analog.” Computerized frequency
analysis was here to stay and to prove to be of enormous value
not only in psychophysiologic research but also in the assess-
ment of neuropharmacologic effects.
In the 1970s, the evoked potential technique progressed
greatly. The introduction of the pattern changer in the visual
evoked potential technique made this method highly reliable;
the names of H. Speckrejse and R. Spehlmann ought to be men-
tioned in this context. In the field of auditory evoked potentials,
the location of primary cortical discharge was elusive for many
years, and the late vertex potential of limited clinical value. The
introduction of the far-field technique for the demonstration of
the brainstem auditory evoked potentials (Jewell), however,
proved to be extremely valuable. Analogous work in the field of
Chapter 1 ■ Historical Aspects of EEG 13
somatosensory evoked potentials is associated with the names
of Roger and Joan Cracco.
The 1960s and 1970s witnessed a regrettable alienation of
EEG and epileptology, which had existed before in an almost
perfect marriage. A sizable number of epileptologists lost inter-
est in EEG. Was it early enthusiasm about the introduction of
antiepileptic serum levels? Was the path to mastery of EEG
becoming too laborious? This situation changed in the 1980s
due to the rapidly increasing emphasis on EEG and related
techniques in the presurgical workup of patients considered
candidates for seizure surgery.
The 1970s and 1980s saw brilliant structural neuroimaging
techniques emerging: computed tomography and magnetic res-
onance imaging. This seemed to knock out electroencephalog-
raphy from the contributors to focal CNS diagnosis. Such a
knockout blow, however, was also more apparent than real. The
EEG, by its very nature, never was a structure-oriented test.
Whether the patient has a hemispheric brain tumor, a vascular
lesion, or a traumatic contusion, EEG can always demonstrate
the degree of dysfunctional changes around the lesion (or sec-
ondary diffuse cerebral dysfunction). The sad story is that the
neurologist of our day seems to lose interest in the realm of
function and dysfunction. This development must be halted
(and eventually will be).
Topical EEG diagnosis, however, has made a comeback of its
own in the form of computerized brain mapping. This fascinat-
ing recent development is associated chiefly with the name of
Frank Duffy. Again, it is the old EEG in new clothes: it can be
understood only by an expert of the conventional EEG.
Starting in the late 1960s, a completely new development
took place in the EEG exploration of the full-term and the pre-
mature newborn. The historical aspects of this development are
found in Chapters 11 and 49.
This historical overview is not complete without a few words
about the technicians (technologists) doing the EEG laboratory
routine work. They had to place electrodes with greatest accu-
racy and to obtain a readable tracing, even under the most
adverse conditions. They were considered the electroencephalo-
grapher’s attendants for a long time, even though their work
required considerable sophistication.
John R. Knott and Charles E. Henry invested incredible energy
into the founding of the American Society of EEG (later
Electroneurodiagnostic) Technologists, which came into being in
1962. Soon afterward, the first group of technicians underwent a
stiff examination that made them registered EEG technologists.
There have been similar developments in many other countries.
This evolution has been helpful in giving EEG technologists the
dignity they deserve, but this process is far from being completed.
With every record we read, we must be thankful for the work of
our technical staff and invest some of our energies in their con-
tinuous education and training, for the sake of better technical
EEG quality and thus for the sake of our patients.
This historical overview remains a fragmented account
because much remains untold in this story. However, it is hoped
that the historical perspective this chapter brings to the reader
will foster in our ranks insights that may help us avoid the mis-
takes of the past.
14 Part I ■ Basic Principles
EPILOGUE: THOUGHTS ABOUT
PRESENT AND FUTURE
Clinical Electroencephalogram
The role of clinical electroencephalography has been diminish-
ing throughout the past 30 years except for epileptology, which
has been using the tool of EEG video monitoring to its full
extent. This produces an enormous accumulation of data and
needs thorough analysis and interpretation—a highly time-
consuming task that is often “farmed out” or subcontracted to
outside readers of sometimes unacceptable EEG expertize.
Presence, number or absence of spikes, and spike-related dis-
charges are not the only criteria; full justice must be done to all
EEG abnormalities of nonparoxysmal character, and this task is
a lot more difficult. Such expensive long-term recordings should
be either artfully interpreted or not done at all.
Acutely requested EEG on an emergency basis nowadays
deals mainly with the question of “rule out status epilepticus”
(Varelas et al., 2003). With presently widespread confusions con-
cerning the limitations of nonconvulsive status epilepticus (see
Chapters 28 and 29), the answer needs profound neurologic–
epileptologic understanding. Cases of acute cerebral anoxia
(postcardiopulmonary arrest) may cause both convulsive as
well as nonconvulsive pictures with massive paroxysmal EEG
abnormalities but do not represent true status epilepticus
(see Chapter 23).
Outside the confines of epileptic conditions, the EEG is full
of information for the vast majority of neurologic diseases. It is
most regrettable that this rich source of information has been
badly underused in most neurologic teaching institutions and
hospitals in general (Niedermeyer, 2003b).
The Basis of Electroencephalogram
The cellular basis of EEG activity has been the topic of intensive
studies of extracellular current flow and voltage-dependent
intrinsic oscillations (see Buzsaki et al., 2003). Ongoing work
on cortical and subcortical generators is found in Chapter 3
(also see Ebersole, 2003). This shows that the genesis of EEG is
still a widely discussed issue.
Electroencephalogram and Neurocognition
This area has become the perhaps most fascinating aspect of
modern EEG interest (see Chapter 51). It is true that most of
this type of research has been done with other tools such as
functional magnetic resonance imaging, positron emission
tomography (PET) scanning, and single-photon emission com-
puted tomography (SPECT) (methods demonstrating regional
blood flow and metabolic needs). With the extension of EEG
into the ultrafast frequency ranges (Chapter 37), a powerful
upswing of EEG-oriented neurocognitive research in animals
and humans can be expected. Neurocognition has assumed the
position of the “Holy Grail” of all neuroscience, and EEG stands
a great chance to become a principal contributor. After years of
frantic efforts, it will be found that neuroscience can illuminate
the “brain-mind barrier” only to a certain degree. For the time
being, however, let us not disturb the glowing enthusiasm with
Cassandra calls.
REFERENCES
Major Historical Works
Brazier MAB. A History of the Electrical Activity of the Brain. The First
Half-Century. London: Pitman; 1961.
Critchley M. The Ventricle of Memory. New York: Raven Press; 1990.
Gloor P. Hans Berger on the Electroencephalogram of Man. Amsterdam:
Elsevier; 1969.
Grass AM. The electroencephalographic heritage. Am J EEG Technol.
1984;24:133–173.
Hassler R. Cécile und Oskar Vogt. In: Kolle K, ed. Grosse NervenŠrzte.
Vol. 2. Stuttgart: Thieme; 1959:45–64.
Jung R. Hans Berger und die Entdeckung des EEG nach seinen
Tagebüchern und Protokollen. In: Werner R, ed. Jenenser EEG-
Symposium: 30 Jahre Elektroenzephalographie. Berlin: VEB Verlag
Volk und Gesundheit; 1963:20–53.
Kolle K. Hans Berger. In: Kolle K, ed. Grosse NervenŠrzte. Vol. 1.
Stuttgart: Thieme; 1956:1–6.
Mazza S, Pavone A, Niedermeyer E. Mario Gozzano: the work of an
EEG pioneer. Clin Electroencephalogr. 2002;33:155–159.
Niedermeyer E. Benjamin Franklin and static electricity.
Considerations of past, present, and future. Am J End Technol.
2003a;43:26–29.
O’Leary JL, Goldring S. Science and Epilepsy. New York: Raven Press;
1976.
Schwab RS. Electroencephalography. Philadelphia: WB Saunders; 1951.
Upton M. Electronics for Everyone. New York: New American Library of
World Literature (Signet Key Books); 1960.
Werner R. Hans Berger zum Gedächtnis. In: Werner R, ed. Jenenser
EEG-Symposium: 30 Jahre Elektroenzephalographie. Berlin: VEB
Verlag Volk und Gesundheit; 1963:13–19.
Work of Specially Mentioned EEG Pioneers
Adrian ED. The Berger rhythm in the monkey’s brain. J Physiol.
1936;87:83P–84P.
Adrian ED, Bronk DW. The frequency of discharge in reflex and volun-
tary contractions. J Physiol. 1929;67:119–151.
Adrian ED, Matthews BHC. The interpretation of potential waves in
the cortex. J. Physiol. 1934;81:440–471.
Aserinsky W, Kleitman N. Regularly occurring episodes of eye motility
and concomitant phenomena during sleep. Science. 1953;118:
273–274.
Bartley SH. Analysis of cortical response to stimulation of the optic
nerve. Am J Physiol. 1932;101:4P.
Bartley SH, Newman EB. Recording cerebral action currents. Science.
1930;71:587.
Bartley SH, Newman EB. Studies on the dog’s cortex. Am J Physiol.
1931;99:1–8.
Berger H. Über das Elektrenkephalogramm des Menschen. 1st report.
Arch Psychiat Nervenkr. 1929;87:527–570.
Berger H. Über das Elektrenkephalogramm des Menschen. 4th report.
Arch Psychiat Nervenkr. 1932;97:6–26.
Berger H. Über das Elektrenkephalogramm des Menschen. 7th report.
Arch Psychiat Nervenkr. 1933;100:301–320.
Bishop GH, Bartley SH. Electrical study of the cerebral cortex as com-
pared to the action potential of excised nerve. Proc Soc Exp Biol
(New York). 1932;29:698–699.
Blake K, Gerard RW. Brain potentials during sleep. Am J Physiol.
1937;119:692–703.
Bremer F. Cerveau isolé et physiologie du sommeil. C R Soc Biol
(Paris). 1935;118:1235–1241.
Bremer F. Cerebral and cerebellar potentials. Physiol Rev. 1958;38:
357–388.

Brock LG, Coombs JS, Eccles JC. The recordings of potentials from
motor neurons with an intracellular electrode. J Physiol. 1952;117:
431–460.
Buzsaki G, Traub R, Pedley TA. The cellular basis of EEG activity. In:
Ebersole JS, Pedley TA, eds. Current Practice of Clinical
Electroencephalography. 3rd ed. Philadelphia: Lippincott Williams
and Wilkins; 2003:1–11.
Caton R. The electric currents of the brain. Br Med J. 1875;2:278.
Cooley JW, Tukey JW. An algorithm for the machine calculation of
complex Fourier series. Math Comp. 1965;19:267–301.
Danilevsky VD. Investigations into the Physiology of the Brain. Doctoral
thesis. University Charkov (cited in Brazier, 1961), 1877.
Davis H, Saul LV. Action currents in the auditory tracts of the midbrain
of the cat. Science. 1931;86:448–450.
Dawson GD. A summation technique for the detection of small signals in
a large irregular background. J Physiol (London). 1951;115:2P.
Dietsch G. Fourier-Analyse von Elektrenkephalogrammen des
Menschen. Pflugers Arch Ges Physiol. 1932;230:106–112.
Ebersole JS. Cortical generators and EEG voltage fields. In: Ebersole JS,
Pedley TA, eds. Current Practice of Clinical Electroencephalography.
3rd ed. Philadelphia: Lippincott Williams and Wilkins; 2003:12–31.
Fischer MH. Elektrobiologische Auswirkungen von Krampfgiften am
Zentralnervensystem. Med Klin. 1933;29:15–19.
Fischer MH, Löwenbach H. Aktionsströme des Zentralnervensystems
unter der Einwirkung von Krampfgiften. 1. Mitteilung Strychnin und
Pikrotoxin. Arch Exp Pathol Pharmakol. 1934a;174:357–382.
Fischer MH, Löwenbach H. Aktionsströme des Zentralnervensystems
unter der Einwirkung von Krampfgiften. 2. Mitteilung: Cardiazol,
Coffein und andere. Arch Exp Pathol Pharmakol. 1934b;174:
502–516.
Foerster O, Altenburger H. Elektrobiologische Vorgänge an der
mensehlichen Hirnrinde. Dtsch Z Nervenheilk. 1935;135:277–288.
Gastaut H. Effets des stimulations physiques sur l’E.E.G. de l’homme.
Electroencephalogr Clin Neurophysiol Suppl. 1949;2:69–82.
Gastaut H, Roger J, Corriol JH, et al. Les formes expérimentales
d l’épilepsie humaine. L’épilepsie induite par la stimulation
lumineuse intermittente ou épilepsie photogénique. Rev Neurol
(Paris). 1948;80:161–183.
Geddes LA. What did Caton see? Electroencephalogr Clin Neurophysiol.
1987;67:2–6.
Gerard RW, Marshall WH, Saul LJ. Cerebral action potentials. Proc Soc
Exp Biol (New York). 1933;30:1123–1125.
Gibbs EL, Gibbs FA. Diagnostic and localizing value of electroen-
cephalographic studies in sleep. Publ Assoc Res Nerv Ment Dis.
1947;26:366–376.
Gibbs EL, Fuster B, Gibbs FA. Peculiar low temporal localization of
sleep-induced seizure discharges of psychomotor epilepsy. Arch
Neurol Psychiatry (Chicago). 1948;60:95–97.
Gibbs FA, Davis H. Changes in the human electroencephalogram asso-
ciated with loss of consciousness. Am J Physiol. 1935;113:49–50.
Gibbs FA, Davis H, Lennox WG. The electroencephalogram in epilepsy
and in conditions of impaired consciousness. Arch Neurol
Psychiatry (Chicago). 1935;34:1133–1148.
Gibbs FA, Gibbs EL, Lennox WG. Epilepsy paroxysmal cerebral dys-
rhythmia. Brain. 1937;60:377–388.
Goltz FL. Ueber die Verrichtungen des Groszhirns. Pflueger’s Arch Ges
Physiol. 1888;42:419–467.
Grass AM, Gibbs FA. A Fourier transform of the electroencephalo-
gram. J Neurophysiol. 1938;1:521–526.
Hayne R, Belinson L, Gibbs FA. Electrical activity of subcortical
areas in epilepsy. Electroencephalogr Clin Neurophysiol.
1949a;1:437–445.
Chapter 1 ■ Historical Aspects of EEG 15
Hayne R, Meyers R, Knott JR. Characteristics of electrical activity of
human corpus striatum and neighboring structures. J Neurophysiol.
1949b;12:185–195.
Hoagland H, Rubin MA, Cameron DF. The electroencephalogram of
schizophrenics during insulin hypoglycemia and recovery. Am J
Physiol. 1937;120:559–570.
Jung R, Baumgarten RV, Baumgartner G. Mikroableitungen von
einzelnen Nervzellen im optischen Cortex der Katze. Die lich tak-
tiviertcn B-Neurone. Arch Psychiat Z Ges Neurol. 1952;189:
521–539.
Knott JR, Gibbs FA. A Fourier transform of the electroencephalogram
from one to eighteen years. Psychol Bull. 1939;36:512–513.
Knott JR, Gibbs FA, Henry CE. Fourier transforms of electroen-
cephalogram during sleep. J Exp Psychol. 1942;31:465–477.
Knott JR, Hayne RA, Meyers HR. Physiology of sleepwave characteris-
tics and temporal relations of human electroencephalograms
recorded from the thalamus, the corpus striatum and the surface of
the scalp. Arch Neurol Psychiatry (Chicago). 1950;63:526–527.
Kornmüller AE. Architektonische Lokalisation bioelektriseher
Erscheinungen auf der Grosshirnrinde. 1. Mitteilung:
Untersuchungen am Kaninchen bei Augenbelichtung. J Psychol
Neurol. 1932;44:447–459.
Kornmüller AE. Die Ableitung bioelektischer Effekte architektonischer
Rindenfelder vom uneröffneten Schadel. J Psychol Neurol.
1933;45:172–184.
Kornmüller AE. Der Mechanismus des epileptischen Anfalles auf
Grund bioelektrischer Untersuchungen am Zentralnervensystem.
Fortschr Neurol Psychiatry. 1935;7:391–400, 414–432.
Kornmüller AE. Die Bioelektrischer Erseheinungen der Hirnrindenfelder.
Leipzig: Thieme; 1937.
Kornmüller AE. Klinische Elektrenkephalographie. Munich: Lehmann;
1944.
Kornmüller AE. Die Elemente der Nervšsen TŠtigkeit. Stuttgart:
Thieme; 1947.
Lennox WG. The oxygen and carbon dioxide content of blood from
the internal jugular and other veins. Arch Intern Med.
1930;46:630–636.
Lennox WG. The cerebral circulation. Arch Neurol Psychiatry
(Chicago). 1931;26:719–724.
Lennox WG, Gibbs EL. The blood flow in the brain and the leg of man,
and the changes induced by alteration of blood gases. J Clin Invest.
1932;1:1155–1177.
Li CL, Jasper HH, McLennan H. Décharge d’unités cellulaires en rela-
tion avec les oscillations électriques de l’écorce cérébrale. Rev
Neurol (Paris). 1952;87:149–151.
Liberson WT. Recherches sur les électroencéphalogramme tran-
scraniens de l’homme. Travail Hum. 1937;5:431–463.
Lindsley DB, Bowden JW, Magoun HW. The effect of subcortical
lesions upon the electroencephalogram. Am Psychol. 1949;4:
233–234.
Loomis AL, Harvey EN, Hobart GA. Potential rhythms of the cerebral
cortex during sleep. Science. 1935;82:198–200.
Meyers HR, Hayne R. Electrical potentials of the corpus striatum and
cortex in Parkinsonism and hemiballism. Trans Am Neurol Assoc.
1948;73:10–14.
Morison RS, Dempsey EW. A study of thalamo-cortical relations. Am J
Physiol. 1942;135:281–292.
Moruzzi G. L’attività dei neuroni corticali durante il sonne e durante la
reazione elettroencefalografica di risveglio. Ricerca Sci. 1952;22:
1165–1173.
Moruzzi G, Magoun HW. Brain stem reticular format and activation of
the EEG. Electroencephalogr Clin Neurophysiol. 1949;1:455–473.
16 Part I ■ Basic Principles
Niedermeyer E. The clinical relevance of EEG interpretation. Clin
Electroencephalogr. 2003b;34:93–98.
Penfield W, Jasper HH. Epilepsy and the Functional Anatomy of the
Human Brain. Boston: Little, Brown; 1954.
Phillips CG. Some properties of pyramidal neurones of the motor cor-
tex. In: Wolstenholme GEW, O’Conner M, eds. The Nature of Sleep.
Boston: Little, Brown; 1961:4–24.
Rothmann H. Zusammenfassender Bericht über den Rothmann’schen
groszhirnlosen Hund nach klinischer und anatomischer
Untersuchung. Z Ges Neurol Psychiat. 1923;87:247–313.
Sherrington CS. Integrative Action of the Nervous System. New Haven,
CT: Yale University Press; 1906.
Travis LE, Dorsey JM. Action current studies of simultaneously active
disparate fields of the central nervous system of the rat. Arch Neurol
Psychiat. 1932;28:331–338.
Travis LE, Herren RY. The relation of electrical changes in the brain to
reflex activity. J Comp Psychol. 1931;12:23–29.
Travis LE, Knott JR. Brain potential studies. I. Perseveration time to
light. J Psychol. 1936;3:97–100.
Varelas PN, Spanaki MV, Hacein-Bey L, et al. Emergent EEG.
Neurology. 2003;61:702–704.
Vogt C, Vogt O. Sitz und Wesen der Krankheiten im Lichte der topistis-
chen Hirnforschung und des Variierens der Tiere. Leipzig: Barth;
1937.
Walter WG. The location of brain tumors by electroencephalogram.
Proc R Soc Med. 1936;30:579–598.
Walter WG. Slow potential waves in the human brain associated with
expectancy, attention and decision. Arch Psychiat Nervenkr.
1964;206:309–322.
Walter WG, Dovey VJ, Shipton H. Analysis of electrical responses
of the human cortex to photic stimulation. Nature. 1946;
158:540–541.
Walter WG, Cooper R, Aldridge VJ, et al. Contingent negative varia-
tion. An electric sign of sensorimotor association and expectancy in
the human brain. Nature. 1964;203:380–384.
Yakovlev PI. Vladimir Mikhailovich Bekhterev (1857–1927). In:
Haymaker W, ed. The Founders of Neurology. Springfield, IL:
Charles C Thomas; 1953:244–247.
 CHAPTER
Neurophysiologic Basis of
EEG and DC Potentials
ERWIN-JOSEF SPECKMANN, CHRISTIAN E. ELGER, AND ALI GORJI
2
T
he clinical electroencephalographer correlates central numbers. Thus, nerve cells are usually covered with several
nervous system (CNS) functions as well as dysfunctions thousand synapses (2). The glia cells are imbedded between
and diseases with certain patterns of the electroen- nerve cell somata, dendrites, and axons. They usually have sev-
cephalograhy (EEG) on an empirical basis. Obviously, this eral processes that make contact with somata and processes of
method has been found valuable in clinical practice. Therefore, nerve cells; they may also make contact with vessels. This histo-
why should the clinical electroencephalographer study the basic logic arrangement results in a cerebral extracellular space con-
elementary processes underlying the EEG? There is little doubt sisting of very narrow intercellular clefts (3).
that the range of EEG interpretations can be much widened and
misinterpretations avoided when the underlying elementary NEURONAL MEMBRANE POTENTIALS:
processes are also considered. This is true especially for convul-
INTRACELLULAR RECORDINGS
sive disorders and cerebral metabolic disturbances. For exam-
ple, an isoelectric EEG can be caused by selective pCO2 increase Essential potentials that can be demonstrated with intracellular
while the brain is sufficiently supplied with O2. On the other recordings are characterized briefly. When the membrane of the
hand, in the presence of practically normal pCO2 levels, cere- nerve cell body is penetrated by a microelectrode, a potential of
bral hypoxia may be the cause. It will be pointed out below that about 60 to 70 mV with negative polarity in the intracellular
the prognosis may be quite different in these two cases. space can be recorded. This membrane potential is subject to
various fluctuations that are elicited chiefly by synaptic activi-
POTENTIAL GENERATION AT NEURONAL ties. Their mechanisms are shown in greater detail in Figure 2.2.
AND GLIAL ELEMENTS: MEMBRANE As can be derived from this schematic illustration, the neuron
POTENTIALS AND FIELD POTENTIALS
The basic mechanisms that give rise to potentials recorded out-
side the CNS elements will be described. Such extracellular
potentials are generally known as field potentials (1).
In the course of this presentation, the morphology of gen-
erator structures is discussed briefly. Then, the electrical activ-
ity demonstrable with intracellular recordings from neurons
and glia cells is described. On the basis of this information,
the principles of the generation of extracellular field poten-
tials are outlined and the various types of field potentials are
characterized.

GENERATOR STRUCTURES
The CNS essentially consists of nerve cells and glia cells. The
arrangement of neurons usually shows a specific type of lami-
nar character. Glia cells are located between neurons. As shown
in Figure 2.1, several processes emerge from the nucleus-con-
taining cellular soma (body) of the nerve cell. These processes
can be divided into two types according to their function. Most
of the processes are dendrites that branch off into numerous
small ramifications. Every cell also has an axon that may split
up into multiple collaterals. Such an axon provides contact with
other nerve cells or with other target organs. In the case of
interneuronal connections, the contact consists of synapses that Figure 2.1 Schematic drawing of morphology and histology of neu-
cover the dendrites, the soma, and the axon hillock in large ronal and glial elements.

17
18 Part I ■ Basic Principles
Figure 2.2 Membrane potential (MP) changes and current flows dur-
ing synaptic activation. A: The MP of the postsynaptic neuron and the
MP of the presynaptic fibers are recorded by means of intracellular
microelectrodes. Action potentials in the excitatory and inhibitory
presynaptic fiber lead to excitatory postsynaptic potential (EPSP) and
inhibitory postsynaptic potential (IPSP), respectively, in the postsynap-
tic neuron. Two EPSPs sum up to a superthreshold potential, triggering
an action potential in the postsynaptic neuron. B: During EPSP and
IPSP, ionic current flows occur through as well as along the neuronal
membrane, as shown by arrows. The density of and signs indicate
the polarization of the subsynaptic (dark area) as well as that of the
postsynaptic membrane during synaptic activation.
from which the soma membrane potential is recorded has
synaptic connections. The corresponding presynaptic struc-
tures are also explored with microelectrodes. If an action poten-
tial travels along the fiber, which ends in an excitatory synapse,
an excitatory postsynaptic potential (EPSP) occurs in the
following neuron (Fig. 2.2A). If two action potentials travel
along the same fiber with a short interval, there will be a sum-
mation of EPSP triggering an action potential on the postsy-
naptic neuron after reaching the membrane threshold. If an
action potential travels along a fiber ending in an inhibitory
synapse, then hyperpolarization will occur, representing an
inhibitory postsynaptic potential (IPSP) (4–6).
Because of the time course of the various membrane poten-
tial fluctuations, the postsynaptic potentials are thought to
contribute primarily to the generation of the extracellular field
potentials in question (1,5,7,8). For this reason, the ionic
mechanisms of these potentials are discussed in greater detail.
The individual events of this process are presented with a mag-
nified time base (Fig. 2.3). With the elicitation of an EPSP, a net
inflow of cations occurs across the subsynaptic membrane. As
shown in Figure 2.2B, a potential gradient develops along the
neuronal membrane in the intra- and extracellular spaces.
Because of this potential gradient, cations move along the
nerve cell membrane through the extracellular space in the
direction of the subsynaptic region. An inversely directed flow
takes place in the intracellular space. With the generation of an
IPSP, there is an outflow of cations from the nerve cell and/or
an inflow of anions into the nerve cell. These changes first
increase the membrane potential at the subsynaptic membrane
in comparison with the surrounding segments of the mem-
brane. For this reason, a potential gradient develops along the
nerve cell membrane, as in the case of the EPSP genesis. This
potential gradient causes, in the extracellular space, a flow of
cations from the subsynaptic region to the surrounding por-
tions of the membrane. An inverse process develops in the
intracellular space (5).
The ion fluxes in the extracellular space are of paramount
significance in the generation of field potentials. Therefore,
these processes are further discussed in the following chapters.
GLIAL MEMBRANE POTENTIALS:
INTRACELLULAR RECORDINGS
In addition to the neurons, glial cells may also play a role in the
generation of extracellular field potentials (9,10). Therefore, the
bioelectric properties of glial cells are summarized.
If a glia cell is penetrated with a microelectrode, a membrane
potential can be recorded with a polarity similar to that of the
nerve cells. The size of this membrane potential approximates
the potassium equilibrium potential and hence somewhat
exceeds the membrane potential of nerve cells. In contrast to
neurons, glial cells fail to show any action potentials, and there
are also no postsynaptic potentials. Thus, in contrast to neu-
rons, glial cells do not show characteristic potentials that distin-
guish them unmistakably from other cells. The glial membrane
potential, however, is also not constant. An augmentation of the
extracellular potassium concentration (potassium activity)
causes depolarization of glial cells (Fig. 2.4A). Concentration
changes of other ions cause only negligible alterations of the
glial cell membrane potential. The glial cell is hence compara-
ble to a potassium electrode (9,11).
The dependency of the glial membrane potential on the
extracellular potassium concentration is the reason for a
functional linkage with adjacent neuronal structures.
Neuronal activity is associated with outflow of potassium
ions. As shown schematically in Figure 2.4B, repetitive firing
of neurons gives rise to increased extracellular potassium
concentration and hence to glial cell depolarization (12,13).
If the potassium concentration does not affect the entire glial
cell membrane and remains increased only locally, then
potential gradients build up along the glial cell, giving rise to
 Chapter 2 ■ Neurophysiologic Basis of EEG and DC Potentials 19

Figure 2.3 Basic mechanisms underlying

generation of potentials (electroencephalogra-
phy [EEG]) and of magnetic fields (magnetoen-
cephalography [MEG]) in the extracellular space
of central nervous system. The description is
based on the assumption that an extended neu-
ronal process, for example, a dendrite, is locally
depolarized by activation of an excitatory synapse.

intra- and extracellular current flows similar to the ones
described in reference to neuronal synaptic transmissions
(Fig. 2.2). Glial cells frequently have widespread processes
and furthermore may have close connections with each other.
For this reason, potential fields of considerable spatial
extension may develop on the basis of the aforementioned
mechanisms (1,10,14,15). In view of the above-described
functional interconnections, it is quite likely that in the gen-
esis of extracellular field potentials an amplifying effect can
be attributed to the glial cells.

Figure 2.4 Membrane potential (MP) changes of glia cells induced by an increase in the extracellular K concentration
(arrows in the schematic drawings). A: Potassium is applied extracellularly to the glia cell. B: The potassium concentra-
tion is increased due to an activation of a neighboring neuron. (From original tracings from Kuffler SW, Nicholls JG, Orkand
RK. Physiologic properties of glial cells in the central nervous system of amphibia. J Neurophysiol. 1966;29:768–787.)
20 Part I ■ Basic Principles

FIELD POTENTIALS
It has been shown in the preceding section that primary trans-
membranous currents generate secondary ionic currents along
the cell membranes in the intra- and extracellular spaces. The
portion of these currents that flows through the extracellular
space is directly responsible for the generation of field poten-
tials (Fig. 2.3). Particular significance must be ascribed to the
synaptic processes as causing events for the field potentials in
question, especially for their time course. In accordance with
these statements, the generation of extracellular field potentials
will be discussed as exemplified by extracellular fields accompa-
nying synaptic activity (5,8,15,16). The discussion of these
events will again make use of a very protracted time axis
(Fig. 2.3). The explanation of the events is given in reference to
the schematic view in Figure 2.5. This figure shows a widely
stretched neuronal element, with one end segment lying close
to the surface of a central nervous structure. At both ends of
this neuronal unit, the microelectrodes ME1 and ME2 are
inserted. At the same time, the extracellular electrodes E1 and E2
are located at the surface and at the deeper end of the neuronal
element. The potentials picked up from the intra- and extracel-
lular electrodes are shown in the vicinity of each electrode. The
potential recorded from the surface of the nervous structure is
accentuated by thicker lines. Figure 2.5 shows active excitatory
and inhibitory synapses, either close to the surface or located in
the depth. As described elsewhere, the activation of an excita-
tory synapse leads to a net inward flow of cations. If this state-
ment is applied to Figure 2.5A1, then it becomes evident that Figure 2.5 Membrane potential (MP) changes and field potentials
the upper end of the neuronal element will be depolarized in (FPs) elicited by the activation of excitatory and inhibitory synapses
comparison with other segments of the same cell. Accordingly, in the central nervous system. The elementary processes are explained
the synaptic current flow causes an EPSP at the microelectrode by means of a neuronal element (hatched area), the one end of which
ME1. This local depolarization then gives rise to further intra- contracts the surface of a structure in the central nervous system. The
and extracellular ionic currents along the nerve cell membrane. MP of the neuron element is recorded at both ends by the microelec-
Because of the intracellular movements of positive charges, trodes ME1 and ME2. The extracellular field is picked up at the surface
depolarization in the area of microelectrode ME2 also takes of the neuronal structure by the electrode E1, as well as in the vicin-
place. This depolarization, however, is less steep and of smaller ity of ME2 by the electrode E2. Active excitatory and inhibitory
amplitude. At the superficially located extracellular electrode synapses are marked by open triangles and black triangles (S), respec-
E2, the inflow of positive charges into the neuronal element tively. A1: The inward current at S generates an EPSP that appears in
causes a negative field potential. The extracellular electrode E2 the region of ME1, as well as in that of ME2. Because S is located
is, metaphorically speaking, approached by positive charges so superficially, the FP generated, due to the direction of the extracellu-
that a positive field potential will develop in this area. The point lar current flow (arrows), is of negative polarity at the surface (E1)
of reversal of the field potentials is localized between electrodes and of positive polarity in the deeper recording (E2). A2: The activa-
E1 and E2. The exact position of the point of reversal depends tion of a deep excitatory synapse elicits a current flow with inverse
on the distribution of extracellular impedances. direction as compared with A1. Therefore, the extracellular FP con-
Current flows of reversed direction (in reference to the sists of a positive deflection at the surface and a negative one at the
recording electrodes) will occur if the active excitatory synapse depth. B1: The outward current at S generates an IPSP in the region
is located at the deeper end of the neuronal element (Fig. of ME2, as well as in that of ME1. Due to the direction of the extracel-
2.5A2). In this case, positive charges approach the superficially lular current flow, the FP generated consists of a positive fluctuation
located electrode (E1) (again speaking metaphorically) and in the depth (E2) and a negative one in the surface recording (E1).
remove themselves from the deeply located electrode (E2). This B2: The current flow during the activation of a superficial inhibitory
arrangement of the active synaptic structures causes a positive synapse is inverse as compared with B1. Therefore, the FP recorded
field potential at the surface and a negative one at the deep elec- from the surface consists of a positive fluctuation. Differences in the
trode. The current flows accompanying the activation of time course of the various potentials are caused by the electrical prop-
inhibitory synapses located in deeper and in more superficial erties of the tissue.
areas, respectively, are shown in Figure 2.5B. As can be derived
from this illustration, the activation of a deep inhibitory
 Chapter 2 ■ Neurophysiologic Basis of EEG and DC Potentials 21

synapse (Fig. 2.5B1) produces a current flow that is largely sim-

ilar to the one generated by the activation of a superficial exci-
tatory synapse (Fig. 2.5A1). In the same manner, there are also
similar current flows in the extracellular space when a superfi-
cial inhibitory synapse (Fig. 2.5B2) or a deeply located excita-
tory synapse (Fig. 2.5A2) is activated. Accordingly, a negative
field potential will develop at the surface of a central nervous
structure (in the schematic view of Fig. 2.5) whenever a super-
ficial excitatory or a more deeply located inhibitory synapse is
activated. The corresponding principle applies to generation of
the superficial field potentials of positive polarity.
Due to their higher densities especially the currents flowing
through the intracellular spaces give rise to magnetic fields sur-
rounding the elements. These magnetic fields can be picked up
from the scalp as magnetoencephalography (MEG; Fig. 2.3).

Types of Field Potentials

The field potentials, whose generation has been described, can
be subdivided into different types. If field potentials are
recorded against an inactive reference point with an upper fre-
quency limit of about 100 Hz, then two types of field potentials
can be distinguished, depending on the time constant of the
amplifying recording device. In the case of a time constant of
1 second or less, the extracellular field potentials correspond
with that which is commonly known as the EEG. If the record-
ing is carried out with an infinite time constant, that is, with
direct current (DC) amplifier, then slower potentials can also be
picked up. Potentials recorded with this technique are generally
known as DC potentials (1,8,15,17). Thus, DC potentials com-
prise slow as well as fast field potentials. The fast components
correspond with the potential fluctuations of the EEG. Due to
different time constants, however, the faster potential compo-
nents may differ from each other as far as their time course is
concerned when recordings are done either with conventional
EEG amplifiers or with DC amplifiers.
Thus far, technical problems have made it difficult to carry
out DC recordings from the scalp. Except for special areas of
application, DC recordings are usually performed in animal
experiments. DC potentials directly reflect the state of activity
of central nervous cells and therefore contribute to the explana- Figure 2.6 Recording of neuronal membrane potentials using voltage-
tion of the mechanisms of genesis of cerebral field potentials sensitive dyes, and principles and schematic example of application.
(14,18). For this reason, DC potentials will be discussed jointly A: (1) A dye is incorporated into the double lipid membrane of nerve
with EEG waves. cells and illuminated by light with dye-specific wave length; simultane-
For the sake of comparison, Figure 2.6 shows the EEG and ously the membrane potential is recorded with an intracellular micro-
the DC potentials during convulsive activity, hypercapnia, and electrode against a reference electrode in the extracellular space.
asphyxia. As shown in this illustration, a tonic–clonic convul- (2) Changes of the membrane potential (MP) starting from the resting
sion is associated with a negative DC shift (8,13–15,19–21). level passing a decrease (depolarization) with action potentials superim-
Furthermore, it can be seen that the hypercapnia-induced dis- posed and a subsequent increase (hyperpolarization) and eventually
appearance of the EEG is associated with a monophasic positive returning to resting level. (3) In correspondence to the different MP levels
DC shift. In the case of EEG extinction due to primary fluorescence and absorption of the dye change. With fluorescent dyes a
asphyxia, however, there are characteristic patterns of DC fluc- depolarization is associated with a decrease and a hyperpolarization with
tuation. Hence, similar findings in the conventional EEG may an increase of fluorescence (symbols). B: (1) Two neurons in a popula-
be associated with different DC shifts. tion; one stays in the resting state and the other changes its MP as in
A2. (2) By the aid of a microscope and a connected array of diodes
Original Remarks Made by Ernst Niedermeyer (squares) the different MP changes of both neurons can be detected via
What does the term DC shift mean? What is DC? Speaking from the different optical behaviors. (Speckmann E-J. Das Gehirn meiner Kunst.
experience, many electroencephalographers have no clear Kreativität und das selbstbewußteGehirn. Münster: Daedalus; 2008.)
22 Part I ■ Basic Principles

concept regarding DC potentials or DC shifts. One cannot

blame them because, for strange reasons, “DC” has two mean-
ings in this context:
1. DC means direct current (and this is, of course, common-
place): a current without oscillations; a current derived from
a battery source; a current maintained in one direction
through a circuit. A more imperfect DC is produced by a rec-
tifier, used to change alternating current (AC) into DC. For
multilingual readers, DC is courant continu in French,
Gleichstrom in German, and corrente continuo in Italian.
Electroneurophysiologically, DC shifts are ultraslow poten-
tials, about as slow as 0.1 to 0.2/sec. This, however, is not true
DC. Such slow activity is just a bit more “DC-like” since it
does not show the faster “AC-like” activity. One simply has to
live with this kind of misnomer.
2. DC also means direct coupling (and this is much less
known). What coupling? The coupling between the stages of
EEG amplification. Conventional EEG machines have stages
coupled by capacitors. Now one has to remember that capac-
itors (a) reject DC and (b) determine the time constant.
Even a very long time constant (several seconds duration)
may not suffice for the recording of DC potentials. Direct
coupling is a capacitor-free coupling between the stages of Figure 2.7 Simultaneous detection of membrane potentials (MP) of all
amplification and provides the optimal condition for DC neurons in a population (voltage-sensitive dye) and conventional
recording. This is technically quite difficult in clinical con- recording of the local field potential (FP) at the same time. Living brain
ventional EEG recording but easier under experimental neu- tissue (0.5 mm thick) from the temporal lobe of a patient who under-
rophysiologic conditions in animals.
went epilepsy surgery. A: Recording of the local FP (“local EEG”).
Hence, be aware of the dual significance of the term DC. (1) Resting state; (2 to 4) epileptic discharges of different intensities.
Epileptic discharges appeared spontaneously, that is, they were not
Validation of Field Potentials induced experimentally. B: MP changes indicated by the intensity of flu-
As to Their Functional Significance orescence of the dye (black, decrease of the MP, depolarization).
Correlating field potentials with other signals accompanying Resting state (1); similar epileptic potentials in the FP (2 and 3) are
neuronal activity may contribute to a validation of them. There associated with different extents of neuronal depolarizations and simi-
is a variety of methods other than electroencephalography for lar extents of neuronal depolarizations with different epileptic poten-
detecting brain activity. Thus, single photon emission tomogra- tials in the FP (3 and 4). (Speckmann E-J. Das Gehirn meiner Kunst.
phy (SPECT), positron emission tomography (PET), functional Kreativität und das selbstbewußteGehirn. Münster: Daedalus; 2008.)
magnetic resonance imaging (fMRI), and intrinsic optical imag-
ing (IOI) are based on metabolic changes associated with
increases of local neuronal activity. The latter methods are “very Figure 2.7. The living brain slices are stained with fluorescence
indirect” (22). EEG including evoked potentials (EP) and MEG (or absorption) dyes (Fig. 2.7A1). With depolarization and
represent “more direct ones” since they measure the field effects hyperpolarization the fluorescence is decreased and increased,
of the proper neuronal activity and therewith of the information respectively (Fig. 2.7A2 and A3). The changes in fluorescence are
processing brain activity. For the analysis of neuronal network measured via a microscope by an array of detectors and there-
functions, the immediate and simultaneous recording of mem- with the actual membrane potentials of the neurons are
brane potentials of all neurons in a population by application of observed (Fig. 2.7B1 and B2).
voltage-sensitive dyes is the “only direct” method available yet The method of simultaneous measurement of neuronal
(23–28). All these methods have advantages and disadvantages. membrane potentials of all neurons in a population is success-
Thus, the functional imaging using voltage-sensitive dyes cannot fully applied in living human brain slices (0.5 mm thick) in
be applied in patients for several reasons, for example, prerequi- vitro obtained from neurosurgical interventions (tumor and
site of direct access to the brain structure to be investigated, pho- epilepsy surgery) (25–27,29).
totoxicity, and pharmacologic side effects of the dyes. But, this A comparison of the field potentials, that is, the local EEG,
method is helpful to analyze the principles of functional mean- and of the neuronal membrane potentials detected by the aid
ing of field potentials in living human brain slices in vitro, espe- of voltage-sensitive dyes is given in Figure 2.8. The tissue is a slice
cially with spontaneously occurring epileptic discharges. preparation from the temporal neocortex resected from a patient
A principle and schematic example of recording neuronal suffering from pharmacoresistant complex partial seizures. Most
membrane potentials using voltage-sensitive dyes is displayed in of these living human brain slices show spontaneous epileptic
 Chapter 2
Figure 2.8 EEG (time constant: 1 second; upper frequency limit:
100 Hz) and DC/ EEG recordings (DC recording: upper frequency limit,
100 Hz) during a generalized seizure induced by pentylenetetrazol
(A), during hypercapnia (B), and during asphyxia (C). Original record-
ings were obtained from cats and rats. Note the different time scales.
EEG potentials, that is, epileptic discharges not induced by exper-
imental procedures. One can derive the following from the
recordings:
1. During epileptic discharges only a certain portion of the neu-
rons in the population is active simultaneously, that is, a com-
plete synchronization is missing (Fig. 2.7, numbers 2 to 4).
2. Similar epileptic potentials in the EEG (Fig. 2.7, numbers
2 and 3) are associated with different extents of neuronal
depolarizations and similar extents of neuronal depolar-
izations with different epileptic potentials in the EEG
(numbers 3 and 4).
WAVE GENERATION
In the preceding sections, the generation of single field poten-
tials was described. In this section, the principles of the genera-
tion of wavelike potential fluctuations are outlined. This is
followed by the discussion of the laminar distribution of such
potentials in the cerebral cortex.
To present the generation of wavelike potential fluctuations
on the surface of a central nervous structure, a simple model as
shown in Figure 2.9 is used. This model consists of two extended
pyramidal neurons of vertical orientation. Terminals of afferent
fibers make contact with the superficial dendrites of both neu-
rons via excitatory synapses. The bioelectrical activity of these
structures is recorded with intracellular microelectrodes. The
■ Neurophysiologic Basis of EEG and DC Potentials 23
Figure 2.9 Principles of wave generation. The excitatory synapses of
two afferent fibers contact the superficial dendritic arborization of two
longitudinal neuronal elements. The afferent fiber activity is recorded by
means of the intracellular electrodes E1 and E2, and the membrane
potentials (MPs) of the dendritic elements are recorded by the electrodes
E3 and E4. The field potential at the surface of the neuronal structure
(cortex) is led by the electrode E5. Synchronized groups of action poten-
tials in the afferent fibers (E1, E2) generate wavelike EPSPs in the den-
dritic areas (E3, E4) and corresponding field potentials in the EEG and
DC/ EEG recording (E5). Tonic activity in the afferent fibers results in a
long-lasting EPSP with small fluctuations. During this period the EEG
(5b) shows only a reduction in amplitude, whereas the DC/ EEG record-
ing (5a) reflects the depolarization of the neuronal elements as well.
microelectrodes E1 and E2 are located in the ascending fibers and
the microelectrodes E3 and E4 are in the superficial dendrites of
the postsynaptic neurons. To pick up the extracellular field
potentials, the electrode E5 lies on the surface of the central
nervous structure.
As shown in tracings 1 and 2, action potentials occur synchro-
nously in the afferent fibers. There are grouped discharges that
are temporarily supplanted by tonic activity. The ascending
action potentials elicit individual EPSP in the upper dendrites of
the neurons; these EPSPs are subsequently summated into major
depolarizations in accordance with the discharge frequency. As
shown in tracings 3 and 4, amplitude and duration of the depo-
larizations depend on the discharge pattern of the afferent fibers.
The synaptic activity at the superficial structures gives rise to
extracellular current flows resulting in superficial field potentials.
With the use of DC recording techniques, the superficial field
potentials reflect the potential fluctuations of the dendritic mem-
brane. If, however, the superficial field potentials are recorded
with a time constant of 1 second or less, then only the fast fluctu-
ations of the superficial field potentials are demonstrable.
Thus far, the principles of genesis of EEG and DC waves
have been shown in the schematic view of Figure 2.9.
Accordingly, the generation of physiologic EEG waves may be
explained as follows. If a grouped and synchronous influx
24 Part I ■ Basic Principles

takes place in afferent fiber systems toward the superficial

generator structures, then EEG waves evolve that are of high
amplitude and distinctly separated from each other. In case
of a periodic sequence of the afferent bursts, the recording of
the field potentials shows sinusoidal potential fluctuations.
This mechanism has been presumed by several groups of
investigators as the principle of the generation of the alpha
rhythm and slower periodic EEG waves. According to these
workers, thalamocortical feedback loops are believed to play
a significant role in the generation of the alpha rhythm
(see Chapters 3 and 4 for detailed descriptions of the genera-
tion of rhythmic activities) (1,30).
If the afferent influx of impulses occurs at a high frequency
for a longer period and/or synchronously, then negative field
potentials with small fluctuations will result from the extracel-
lular current flows. Accordingly, the EEG recording will pick
up only waves of smaller amplitude and mostly higher fre-
quency. In the DC recording, however, the prolonged depolar-
ization of the superficial structures caused by the afferent
high-frequency influx will express itself by a negative DC
potential shift (31,32). There is a close correlation between the Figure 2.10 Surface (1) and laminar recordings (2 to 6) of EEG waves
amplitude of the negative DC shift and average discharge fre- of the cortex. The schematic drawing symbolizes conical neuronal ele-
quency in the afferent fiber systems. This mechanism may ments densely packed with synapses. (Drawings from original tracings
apply principally to the generation of beta activity and other obtained in experiments in the rat’s motor cortex during pentobarbital
EEG waves of higher frequencies (see also cortical sources of anesthesia.)
beta and gamma rhythms in Chapter 3 and Fig. 3.16). A
decrease of the amplitudes of the EEG waves can also occur
when the afferent activity is diminished. In this case, however, CORTICAL FIELD POTENTIALS
the depression of EEG waves is accompanied by a positive DC DURING EPILEPTIFORM ACTIVITY
shift (see Fig. 2.17) (33).
The principles of generation of individual and wavelike In the following subsections, the generation of cortical field
field potentials at the surface of central nervous structures potentials during convulsive activity is discussed. The first sub-
such as the cerebral cortex have been described. If the wave- section deals with focal activity, and the second discusses gener-
like potential fluctuations are recorded not only from the alized tonic–clonic convulsive activity. For methodical reasons,
cortical surface, but also from different cortical layers, then it we refer to data derived from experimental work in animals.
can be shown that potential fluctuations in the latter record-
ings may differ considerably from those at the surface. Focal Activity
These differences imply polarity, frequency, and amplitude If a convulsive substance such as penicillin is applied to the sur-
(1,34,35). Such a recording from the cortex of the rat is shown face of the cerebral cortex, steep negative potentials of high
in Figure 2.10. According to this illustration, field potentials amplitude can be picked up from the area of application after a
reverse their polarity between electrode 1 (on the surface) and short latency period. These discharges repeat themselves in
electrode 2 (located 300 m beneath the cortical surface). Two stereotyped form and periodicity (Fig. 2.11A) (13,36,37). If the
and sometimes more of such phase reversals may be observed membrane potential of a cortical neuron is simultaneously
in deeper recording sites depending on the experimental con- recorded with a microelectrode while a second microelectrode
ditions (phase reversals of alpha oscillations are shown in picks up the corresponding field potentials, then potential fluc-
Chapter 3, Figs. 3.9 and 3.10). The vertical distribution type of tuations occur as shown in Figure 2.11B. It can be derived from
field potential will be discussed in greater detail in connection this illustration that the monotonously recurrent negative field
with the generation of cortical field potentials during convul- potentials are associated with equally stereotyped membrane
sive activity. potential fluctuations. These oscillations of the membrane
In the course of the discussion of cerebral field potentials, it commence with a steep depolarization that, having exceeded
was pointed out that particular significance must be attributed to the membrane threshold, triggers a series of action potentials.
synaptic activity. A view of the laminar distribution of neurons in This is followed by a plateau that, after 80 to 100 msec, changes
the cortex and the dense coverage of these unitary structures with into a steep repolarization, and frequently also into a hyperpo-
synapses makes it clear that different patterns of potentials must larization. These membrane potential fluctuations have proved
necessarily occur in different layers when populations of synapses to be characteristic in the epileptiform activity of individual
are activated in a different manner. This should be clarified by the neurons. They are generally known as paroxysmal depolariza-
schematic drawing in Figure 2.10. tion shifts (PDSs) (13,38).
 Chapter 2
Figure 2.11 EEG (A) and membrane potential (MP) A changes of a
pyramidal tract neuron and extracellular field potential (FP) recorded in
the vicinity of the impaled neuron (B) during focal interictal activity
elicited by application of penicillin to the cortical surface (hatched area
in A). Drawings of original tracings from experiments in the rat. The
sweep speed in B is five times that in A. The recording sites are shown
in the schematic drawings.
Investigation of potential distribution within the cerebral
cortex after the local application of penicillin yields a variety
of findings. An appropriate model is shown in Figure 2.12. In
this experiment, recordings of interictal field potentials were
carried out from the cortical surface, from inside the cortex,
and from the spinal cord. The spinal field potentials permit
the observation of electrical activity descending from the cor-
tex to the spinal cord. In Figure 2.12A, negative field poten-
tials are recorded from the cortical surface and from the two
upper intracortical contacts after the application of penicillin
together with penicillin-metabolizing enzyme penicillinase.
■ Neurophysiologic Basis of EEG and DC Potentials 25
There are, however, field potentials with predominantly posi-
tive components in the deeper contacts 4 to 6. If penicillin is
applied to the surface without penicillinase, then negative
field potential will also develop in deeper cortical layers. If it
is assumed that the negative field potentials mirror the direct
epileptiform activity of neuronal structures (Fig. 2.11), then it
must also be assumed that deeper cortical elements are
involved in convulsive activity shown in of Figure 2.12B in
contrast with Figure 2.12A. This is further supported by the
observation that neuronal activity descending to the spinal
cord and producing characteristic spinal field potentials
occurs only under the experimental conditions shown in
Figure 2.12B. If one compares the recordings in Figure 2.12A
and B, it becomes clear that, with a monotonous epileptiform
potential at the cortical surface, the intracortical potential dis-
tribution and the occurrence of descending activity may differ
considerably (34,39–44).
If penicillin is applied to deeper cortical laminae
(Fig. 2.12C), then negative field potentials will be confined to
that region. These potentials are consistently accompanied by
descending activity to the spinal cord. Under these conditions,
there is frequently nothing but a positive potential fluctuation
of minor amplitude at the cortical surface (34,40).
In summary, it can be derived from the described experi-
mental models that, in focal convulsive activity limited to the
cortex, the surface potential does not necessarily reflect the bio-
electrical events in deeper cortical layers.
Generalized Tonic–Clonic Activity
Here, possible mechanisms involved in the generation of cortical
field potentials during tonic–clonic convulsive activity are
described. Again, data are based on experimental observations in
animals. Tonic–clonic convulsive activity was triggered by
repeated injections of pentylenetetrazol (also see Refs. 13 and 37).
Figure 2.12 Cortical field potentials recorded at the
surface (1) and from within the cortex (2 to 6) and
spinal field potentials (7) during interictal activity.
The interictal activity was elicited by penicillin.
A,B: Potential distribution after surface application of
the drug. In A, the spread of penicillin is limited by the
use of penicillinase. C: Potential distribution after intra-
cortical application of penicillin at recording point 4.
The areas directly involved in the epileptiform activity
as indicated by negative field potentials are marked by
hatching in the schematic drawings. Spinal field poten-
tials are linked to the occurrence of negative field
potentials in lamina V (B and C, 4). Distance between
the intracortical electrodes, 300 m. (From original trac-
ings from Elger CE, Speckmann EJ, Caspers H, et al.
Focal interictal epileptiform discharges in the cortex of
the rat: laminar restriction and its consequences for
activity descending to the spinal cord. In: Klee MR, Lux
HD, Speckmann EJ, eds. Physiology and Pharmacology
of Epileptogenic Phenomena. New York, NY: Raven
Press; 1981.)
26 Part I ■ Basic Principles
Figure 2.13 Simultaneous recordings of EEG and DC/ EEG (A) and of
DC/ EEG and membrane potential (MP) of a pyramidal tract neuron (B)
during generalized tonic–clonic seizures elicited by pentylenetetrazol.
(Drawings after original tracings from experiments in the cat’s motor
cortex. The sweep speed in B is 10 times that in A.)
Figure 2.13A shows a tonic–clonic convulsion recorded with
a conventional EEG amplifier, as well as with a DC amplifier.
There is a negative DC shift from the baseline during a convul-
sive seizure. This negative DC shift gradually recedes during the
termination of the convulsions and frequently changes into a
transient positive after shift (13–15,18,19,21,41).
When the membrane potential of a pyramidal tract neuron
of lamina V is recorded during a convulsive seizure, it can be
shown that under these conditions typical PDSs become mani-
fest (Fig. 2.13B). If these PDSs are correlated with the potential
fluctuations in the DC recording, it can be noticed that the
PDSs in pyramidal tract neurons are coupled at the beginning
of the convulsive seizure with superficial negative potential
fluctuations and at the end of the convulsive seizure with sur-
face-positive potential fluctuations (Fig. 2.13B) (1,18,45).
In addition to the field potentials of the cortical surface and
the membrane potentials of the pyramidal tract cells, field
potentials were also recorded in the fifth lamina. Under these
conditions, it can be shown that every PDS is associated with a
negative monophasic field potential in the depth (Fig. 2.14A).
These stereotyped potential fluctuations in deep cortical layers
correspond with field potentials at the cortical surface with
either monophasic negative or positive (Fig. 2.14A1 and A3) or
with polyphasic (Fig. 2.14A2) configurations. This statement
does not merely apply to individual ictal potentials but is also
true for prolonged trains of potentials during the convulsion.
As Figure 2.14B shows, paroxysmal depolarizations of pyrami-
dal tract cells may be accompanied by a sequence of either neg-
ative or positive potentials on the cortical surface. If one
correlates these various field potentials on the cortical surface
with the slow DC shifts occurring during the convulsion (also
see Fig. 2.14A), then it can be demonstrated that the surface-
negative field potentials are associated primarily with a slight
DC shift and that surface-positive field potentials will appear
Figure 2.14 Single potential fluctuations at the cortical surface
(DC/ EEG) and concomitant membrane potential (MP) of a pyramidal
tract cell (PTC) and field potentials (FP) in the PTC layer during gener-
alized tonic–clonic seizures. The seizure activity was induced by
pentylenetetrazol. A: The negative potential (1), the positive–negative
fluctuation (2), and the positive potential (3) in the DC/ EEG recording
coincide with monophasic negative FP and stereotyped paroxysmal
depolarization shift in the neuron. The negative DC shift occurring dur-
ing the seizure is indicated by a dashed line in the upper row.
Monophasic negative potentials in the DC/ EEG recording occur with
small and monophasic positive fluctuations along with a marked DC
displacement. B: The relations between DC/ EEG potentials and MP of
PTC as described for A1 and A3 also hold true for trains of potentials
(1 and 2). (From original tracings from Speckmann EJ, Caspers H,
Jansen RWC. Laminar distribution of cortical field potentials in relation
to neuronal field activities during seizure discharges. In: Brazier MAB,
Petsche H. eds. Architectonics of the Cerebral Cortex, IBRO Monograph
Series. Vol 3. New York, NY: Raven Press; 1978:191–209.)
when the negative DC shift at the cortical surface reaches and
exceeds a critical value (1,18,45,46).
These data are interpreted with flow charts in Figure 2.15.
The amplitude of the negative DC shift at the cortical surface
depends greatly on the amount of the afferent influx of
impulses to the generator structures in the superficial cortical
laminae. This predominantly asynchronous afferent influx is
symbolized by the width of hatched arrows in Figure 2.15.
Accordingly, the afferent influx in Figure 2.15A is smaller than
that in Figure 2.15B. Therefore, there is a smaller DC shift in
Figure 2.15A and a prominent one in Figure 2.15B. In the case
of Figure 2.15A, a synchronized inflow of impulses from sub-
cortical structures is assumed to reach the cortex (widened
afferent fiber in schematic view). As a consequence, pyramidal
tract cells will be stimulated to generate a PDS, and structures
close to the surface will be depolarized through the mediation
of interneurons. Accordingly, in such a constellation of
 Chapter 2 ■ Neurophysiologic Basis of EEG and DC Potentials 27

Figure 2.15 Flow charts of neuronal processes possibly responsible for the generation of DC/ EEG waves of opposite polar-
ity during a generalized tonic–clonic seizure. (Hatched arrows) Symbols for continuous asynchronous input to the cortex;
(heavy lines) symbols for phasic volleys giving rise to single convulsive discharges; (PTC) pyramidal tract cell; (IN)
interneuron; (MP) membrane potential; (UA) extracellularly recorded unit activity. A: During a moderate asynchronous
input to the cortex (small hatched arrow), a burst of UA triggers a paroxysmal depolarization shift in a PTC.
Simultaneously, it leads to a depolarization of superficial neuronal structures and therewith to a negative fluctuation in the
DC/ EEG recording at the cortical surface. B: With an increased asynchronous input to the cortex (wide hatched arrow),
the DC potential shifts to a more negative level than in A (1). When in these conditions a phasic volley reaches the cor-
tex, paroxysmal depolarization shifts are also triggered in PTC, whereas the enhanced asynchronous UA is interrupted
mainly due to inactivation. The latter process results in a disfacilitation of the upper neuronal structures and therewith to
a positive fluctuation of the superficial DC/ EEG potential (2). (From original tracings from Speckmann EJ, Caspers H,
Jansen RWC. Laminar distribution of cortical fluid potentials in relation to neuronal field activities during seizure dis-
charges. In: Brazier MAB, Petsche H, eds. Architectonics of the Cerebral Cortex. IBRO Monograph Series. Vol 3. New York,
NY: Raven Press; 1978:191–209.)

excitatory processes, the paroxysmal depolarization in the
depth will be coupled with a surface-negative field potential.
With augmentation of the already existing afferent inflow of
impulses, the interneurons involved will necessarily exhibit a
heightened level of excitation (Fig. 2.15B). If an additional
highly synchronized afferent influx of impulses takes place
under these conditions, then further PDSs will be triggered in
the pyramidal tract cells, but, in the interneurons, the previ-
ously existing high-frequency activity will be temporarily inter-
rupted, chiefly due to inactivation. This causes a decline of the
excitatory inflow of impulses to the superficial cortical struc-
tures. This disfacilitation gives rise to a positive field potential
at the cortical surface. In this manner, a massive afferent inflow
of impulses provides the basis for a correlation of positive
epicortical field potentials with stereotyped paroxysmal depo-
larizations and monophasic negative field potentials in the
depth (18,45).
SPREADING DEPRESSION
In this section, the generation and propagation of cortical field
potentials during spreading depression (SD) is discussed. SD is
a strong and rapid depolarization of neuronal tissues and prop-
agates slowly (3 to 5 mm/min) as a wave in brain tissue (47).
The SD phenomenon is exclusive to the CNS and appears to
influence both the neuronal and the glial cells. The most
characteristic physiologic feature of SD is a negative DC poten-
tial shift with maximal amplitude of 5 to 30 mV and duration
of 30 to 90 seconds (Fig. 2.16A) (47). SD recordings by extracel-
lular microelectrodes inserted into the gray matter of the neo-
cortex demonstrate that SD can be biphasic or triphasic with
the main negative component. SD begins with a first small pos-
itive component that is sometimes not observed. The main fea-
ture of the SD is a negative wave of high amplitude, which is
then followed by a positive-going wave of smaller amplitude,
but longer duration. The white matter beneath the neocortical
gray substance becomes positive, while the gray matter itself
undergoes the negative wave. Apical dendrites were preferen-
tially involved in the generation of the negative DC deflection.
When recordings were made in hippocampal CA1 region,
SD invariably started earlier and ended later in the layer of the
dendritic trees than among the cell body. SD flows in the oppo-
site direction compared with the current underlying
tonic–clonic seizure discharges, which is inward in the neuron
soma layers (48,49).
A brief period of excitation heralds SD, which is immediately
followed by prolonged neuronal activity depression. This is
replaced by a sustained increase in the neuronal activity; SD
also produces transient suppression of synaptic transmission
that is replaced by a sustained increase in the efficacy of synaptic
transmission. SD significantly enhanced the amplitude of field
EPSP following a transient suppressive period and increased the
28 Part I ■ Basic Principles
Figure 2.16 Propagation of negative DC potential shift in the neocor-
tex and its effect on neuronal excitability. A: Recording of slow spread
of negative DC potential wave after injection of KCl in human neocorti-
cal tissue. Voltage variations were recorded simultaneously by four
electrodes (DC1 to DC4), which are set apart by 1 mm. KCl was injected
in layer VI. Injecting and recording electrodes arranged as shown.
B: Suppression of evoked potentials (EP) during spreading of negative
DC potential fluctuation. The DC potential was recorded from layer III
of a human neocortical slice. EP was elicited by a single electrical
stimu lation (ST) applied to afferent fibers. Note the transient suppres-
sion of EP during propagation of spreading depression. (Modified from
Gorji A. Spreading depression: a review of the clinical relevance. Brain
Res Brain Res Rev. 2001;38:33.)
induction of long-term potentiation in the neocortex (50). The
prodromal excitation as well as late hyperexcitability is mani-
fested in appearance of a burst of multiple action potentials and
intense synaptic noise representing synchronized firing of neu-
rons. The widely spread synchronization seems best explained
by electrical continuity that could be provided by gap junctions.
Patent gap junctions may provide a pathway not only for electric
currents and ions, but also for intracellular “second” messengers
and other active ingredients in cytosol (48).
SD is associated with transient, but major, localized redistrib-
utions of ions between the extra- and intracellular spaces. The
[K ] o rapidly rises and can reach levels as high as 40 to
80 mmol/L. This causes neuronal excitation followed by depolar-
ization and a period of electrical silence during which the DC
potential at the brain surface falls. Increases in [K ]o could result
from a depolarization of both presynaptic and postsynaptic
processes. In addition, presynaptic depolarization evokes
Ca2 -dependent synaptic transmission and glutamate release.
The latter in turn enhances depolarization of postsynaptic ele-
ments and causes further increases in [K ]o. In addition to potas-
sium and proton release from the cells, sodium, calcium, and
chloride enter together with water causing cells to swell and the
volume of the extracellular compartment to be decreased. SD is
accompanied by an increase of glucose utilization and O2 con-
sumption. Recovery of SD depends on energy metabolism (51,52).
SD can be initiated by different stimuli. The most common
model of SD initiation is KCl application to the neuronal
tissues. Other methods of SD induction include mechanical,
electrical, and thermal stimulations, application of the excita-
tory amino acids such as glutamate and aspartate, the Na –K
ATPase inhibitor ouabain, and metabolic inhibitors such as
NaCN (53). The ionic processes responsible for the generation
of SD are complex. Several ionic currents are involved; both
high [K ] o and low[Ca2 ] o can support either SD or epilepti-
form discharges depending on the state of the various mem-
brane ion channels. Insight into the interplay of these different
factors can best be obtained by the study of computer models
solidly based on biophysical data. Thus, the model of Kager
et al. (2008) revealed how the ignition of SD can result from the
shift in the balance between slowly inactivating (persistent)
inward and outward currents in apical dendrites of pyramidal
cells in favor of the former. The effects of hypoxia and anoxia,
through changes in Na/K pump and glial buffering, can also be
put in evidence with these models (54,55).
Processes similar to SD in animal and human neocortex
observed in experimental studies are thought to take place in a
number of pathologic conditions in humans. SD belongs to the
domain of the pathophysiology of the brain, and there are rea-
sons to believe that it is involved in different clinical disorders,
including migraine, cerebrovascular diseases, head injury,
spinal cord disorders, epilepsy, and transient global amnesia
(51). SD can alter the neurochemistry of cortical and subcorti-
cal structures, modulating oxygen distribution, cell survival in
these structures, and behavior. Direct alterations of electrical
activity of cortical neurons by the locally spreading wave can
lead to clinical symptoms (e.g., the aura phase of migraine).
The most common symptoms in the aura phase are visual, aris-
ing from dysfunction of occipital lobe neurons. The positive
(excitatory) neurologic symptoms, for example, flashing lights,
are usually followed by negative (suppressive) ones, for exam-
ple, scotoma or hemianopia, in this phase. Subdural recordings
in patients demonstrated that SD is critically involved in vari-
ous disorders associated with acute neuronal injury including
traumatic and spontaneous intracerebral hemorrhage (56).
CORTICAL FIELD POTENTIALS DURING
GAS TENSION CHANGES IN TISSUE
This section deals with the alterations of epicortical field poten-
tials and concomitant changes of the membrane potentials
caused by deviations of the gas tension in brain tissue. Such
changes of the gas tension may occur when, for instance, the
 Chapter 2
pulmonary and circulatory function is disturbed or when the
local cerebral blood flow is inadequate.
First, the alterations of epicortical field potentials during
selective hypercapnia are discussed; then, those associated with
primary asphyxia are considered. It is shown that EEG changes
may be similar under both conditions. The cortical DC poten-
tial, however, shows typical shifts that permit inferences con-
cerning the cause of the accompanying EEG changes. The
discussion of the effects of gas tension alterations on the bio-
electrical activity of the CNS is based, again, on data derived
from experimental work in animals.
Hypercapnia
If the CO2 tension in the brain tissue is increased in a selective
manner, typical reactions of the cortical field potentials as well
as of the membrane potential and the postsynaptic potentials of
individual neurons are found. These findings are shown in a
summarized schematic view in Figure 2.17.
The animal experiments on which Figure 2.17 is based were
carried out with the use of the so-called apnea technique. With
this technique, interference of the effects of hypercapnia with
simultaneous effects of hypoxia could be avoided. According to
this technique, the experimental animal is ventilated for at least
a half hour with pure oxygen. Thereafter, artificial ventilation is
discontinued while the trachea of the animal remains con-
nected with the O2 reservoir. Under these conditions, the CO2
tension progressively rises in the tissue for about 15 minutes
without a concomitant fall of the oxygen tension below the
baseline level.
Figure 2.17 Effects of an isolated hypercapnia on epicortical field
potentials (EEG, DC/ EEG) and on membrane potential (MP). With
increasing pCO2, the EEG disappears even if the pO2 is above normal
levels. The disappearance of the EEG is associated with a positive DC
shift and a hyperpolarization of most of the neurons. Simultaneously,
the amplitudes of stimulus (St) evoked EPSP are markedly reduced.
(From original tracings from Speckmann EJ, Caspers, H. The effect of
O2 and CO2 tensions in the nervous tissue on neuronal activity and DC
potentials. In: Remond A, ed-in-chief. Handbook of Electroence-
phalography and Clinical Neurophysiology. Vol 2. Part C. Amsterdam:
Elsevier; 1974:71–89.)
■ Neurophysiologic Basis of EEG and DC Potentials 29
With isolated increment of the CO 2 tension in the cerebral
tissue by means of the apnea technique, the amplitude of the
conventional EEG decreases progressively. This amplitude
reduction affects first the waves of higher frequency and then
those of lower frequency. Prior to the extinction of normal
EEG activity, there is once again a phase characterized by
high-frequency EEG activity in the range of 50 to 70 Hz
(1,57). The extinction of the EEG is associated with a shift of
the DC potential in a positive direction. If the CO 2 tension is
then lowered again by reventilation, the EEG waves return in
the original spectral composition after a short latency. At the
same time, the positive DC shift resolves (Fig. 2.17).
Experiments in animals have shown that, with reduction of
the pCO2, the EEG returns to normal activity even though the
hypercapnia-induced suppression lasted for 1 hour or more.
In these cases, a positive DC deflection of monophasic charac-
ter was found to occur during the whole period of apnea
(33,57,58).
Under the aforementioned conditions, the recording of the
membrane potential of a cortical nerve cell shows a hyperpolar-
ization while the CO2 tension is increased. Extensive experi-
mental studies in animals have demonstrated that such a
hyperpolarization is caused primarily by a reduction of the
EPSP (Fig. 2.17; also see Ref. 58). Consideration of field poten-
tials, membrane potentials, and EPSP shows that epicortical DC
potentials reflect neuronal hyperpolarization. The disappear-
ance of the EEG waves is presumed to be caused mainly by the
reduction of postsynaptic activity.
Asphyxia
Primary asphyxia exemplified by respiratory arrest after air ven-
tilation is associated with combined CNS effects of hypercapnia
and hypoxia. The effects of gas tension changes on the field
potentials and on the membrane potential of individual neu-
rons are schematically shown in Figure 2.18. In the correspon-
ding animal experiments, the artificial ventilation with air was
either temporarily (Fig. 2.18A) or persistently (Fig. 2.18B)
interrupted.
With such an interruption of artificial ventilation with
air, the conventional EEG waves disappear within less than
1 minute. This process is accompanied by a negative DC poten-
tial shift from the baseline, which has been characterized as ini-
tial negativity (1 in Fig. 2.18). While the EEG shows an
isoelectric line in the further course of asphyxia, additional
potential shifts are detectable with DC recording technique.
The initial negativity is followed by a positive DC shift termed
intermediate positivity (2 in Fig. 2.18). If reventilation is per-
formed in this phase of asphyxia, an additional positive DC
shift is observed, appropriately termed reactive positivity (3 in
Fig. 2.18A). According to the analysis of the experimental work,
the intermediate and the reactive types of positivity are due to
an increase of CO2 tension in the brain tissue. With the resolu-
tion of the reactive positivity, a restitution of the fast field
potentials occurs that is also demonstrable with the conven-
tional EEG. A comparison of the DC shifts and the alterations
of the membrane potentials shows a parallelism of both events
(1,14,15,33,57,58).
30 Part I ■ Basic Principles
Figure 2.18 Alterations of EEG, DC/ EEG, and neuronal membrane
potential (MP) during primary asphyxia. A: The abolition and the reap-
pearance of EEG during a transient asphyxia goes in parallel with typi-
cal DC shifts: (1) initial negativity, (2) intermediate positivity, and (3)
reactive positivity. These DC fluctuations are accompanied by corre-
sponding reactions of the MP. B: With continuing asphyxia, the EEG
remains abolished and the intermediate positivity (2) turns over into a
terminal negativity (4). The latter DC negativity corresponds to a break-
down of neuronal membrane potential. (From original tracings from
Speckmann EJ, Caspers H. The effect of O2 and CO2 tensions in the
nervous tissue on neuronal activity and DC potentials. In: Remond A,
ed-in-chief. Ha ndbook of Electroen cepha logra phy a nd Clinica l
Neurophysiology. Vol 2. Part C. Amsterdam: Elsevier; 1974:71–89.)
If the interruption of the artificial ventilation is continued
for a longer period of time, then the intermediate positivity
converts into the so-called terminal negativity (4 in Fig. 2.18B).
This negative DC shift correlates with the breakdown of the
neuronal membrane potential. The terminal effects are due to a
critical lack of oxygen. The terminal negativity may be
reversible for a substantial period of time under certain experi-
mental conditions if the artificial ventilation is resumed and the
reduction of the cerebral circulation is counteracted with circu-
lation support measures (58).
In summary, a comparison of EEG and DC potentials in
selective hypercapnia and primary asphyxia shows that the
recording of cortical field potentials with DC amplifiers pro-
vides a more accurate picture of the actual functional state of
nerve cells.
REFERENCES
1. Speckmann E-J, Caspers H, eds. Origin of Cerebral Field Potentials.
Stuttgart: Thieme; 1979.
2. Palay SL, Chan-Palay V. General morphology of neurons and neu-
roglia. In: Kandel ER, ed. Handbook of Physiology/The Nervous
System. Vol 1. Part 1. Bethesda, MD: American Physiological
Society; 1977:5–37.
3. De Robertis EDP, Carrea R, eds. Biology of Neuroglia/Progress in
Brain Research. Vol 15. New York, NY: Elsevier; 1964.
4. Eccles JC. The Physiology of Synapses. Berlin: Springer; 1964.
5. Hubbard JI, Llinas R, Quastel DMJ. Electrophysiological Analysis of
Synaptic Transmission/Monographs of the Physiological Society.
London: Edward Arnold; 1969.
6. Shepherd GM. The Synaptic Organization of the Brain. London:
Oxford University Press; 1974.
7. Creutzfeldt O, Houchin J. Neuronal basis of EEG waves. In:
Remond A, ed-in-chief. Handbook of Electroencephalography and
Clinical Neurophysiology. Vol 2. Part C. Amsterdam: Elsevier;
1974:5–55.
8. Speckmann EJ, Caspers H, Elger CE. Neuronal mechanisms
underlying the generation of field potentials. In: Elbert T,
Rockstroh B, Lutzenberger W, et al., eds. Self-Regulation of the
Brain and Behavior. New York, NY: Springer; 1984:9–25.
9. Kuffler SW, Nicholls JG. The physiology of neuroglial cells. Erg
Physiol. 1966;57:1–90.
10. Somjen GG, Trachtenberg M. Neuroglia as generator of extracellu-
lar current. In: Speckmann EJ, Caspers H, eds. Origin of Cerebral
Field Potentials. Stuttgart: Thieme; 1979:21–32.
11. Kuffler SW, Nicholls JG, Orkand RK. Physiological properties of
glial cells in the central nervous system of amphibia. J Neuro-
physiol. 1966;29:768–787.
12. Orkand RK, Nicholls JG, Kuffler SW. Effect of nerve impulses on
the membrane potential of glial cells in the central nervous system
of amphibia. J Neurophysiol. 1966;29:788–806.
13. Speckmann E-J. Experimentelle Epilepsieforschung. Darmstadt:
Wissenschaftliche Buchgesellschaft; 1986.
14. Caspers H, Speckmann E-J, Lehmenkühler A. Electrogenesis of
cortical DC potentials. In: Kornhuber HH, Deecke L, eds.
Motivation Factor and Sensory Processes of the Brain: Electrical
Potentials, Behaviour and Clinical Use. Progress in Brain Research.
Vol 54. New York, NY: Elsevier; 1980:3–15.
15. Caspers H, Speckmann E-J, Lehmenkühler A. Electrogenesis of
slow potentials of the brain. In: Elbert T, Rockstroh B,
Lutzenberger W, et al., eds. Self-Regulation of the Brain and
Behavior. New York, NY: Springer; 1984:26–41.
16. Rall W. Core conductor theory and cable properties of neurons. In:
Kandel ER, ed. Handbook of Physiology/The Nervous System. Vol 1.
Part 1. Bethesda, MD: American Physiological Society;
1977:39–97.
17. Caspers H, ed. DC potentials recorded directly from the cortex. In:
Remond A, ed-in-chief. Handbook of Electroencephalography
and Clinical Neurophysiology. Vol 10. Part A. Amsterdam: Elsevier;
1974.
18. Speckmann E-J, Caspers H. Cortical field potentials in relation to
neuronal activities in seizure conditions. In: Speckmann E-J,
Caspers H, eds. Origin of Cerebral Field Potentials. Stuttgart:
Thieme; 1979:205–213.
19. Caspers H, Speckmann E-J. DC potential shifts in paroxysmal
states. In: Jasper HH, Ward AA Jr, Pope A, eds. Basic Mechanisms of
the Epilepsies. Boston, MA: Little, Brown & Company; 1969:
375–388.
20. Gumnit RJ, Matsumoto H, Vasconetto C. DC activity in the depth
of an experimental epileptic focus. Electroencephalogr Clin
Neurophysiol. 1970;28:333–339.
21. Speckmann E-J, Elger CE. The neurophysiological basis of epilep-
tic activity: a condensed overview. In: Degen R, Niedermeyer E,
eds. Epilepsy, Sleep, and Sleep Deprivation. Amsterdam: Elsevier;
1984:23–34.
22. Speckmann E-J. Das Gehirn meiner Kunst. Kreativität und das
selbstbewußteGehirn. Münster: Daedalus; 2008.
23. Cohen LB, Salzberg BM. Optical measurement of membrane
potential. Rev Physiol Biochem Pharmacol. 1978;83:35–83.
 Chapter 2
24. Ebner TJ, Chen G. Use of voltage-sensitive dyes and optical record-
ings in the central nervous system. Prog Neurobiol. 1985;46: 463–506.
25. Köhling R, Höhling J-M, Straub H, et al. Optical monitoring of neu-
ronal activity during spontaneous sharp waves in chronically epilep-
tic human neocortical tissue. J Neurophysiol. 2000;84: 2161–2165.
26. Köhling R, Reinel J, Vahrenhold J, et al. Spatio-temporal patterns
of neuronal activity: analysis of optical imaging data using geo-
metric shape matching. J Neurosci Methods. 2002;114:17–23.
27. Straub H, Kuhnt U, Höhling J-M, et al. Stimulus induced patterns
of bioelectric activity in human neocortical tissue recorded by a
voltage sensitive dye. Neuroscience. 2003;121:587–604.
28. Grinvald A, Hildesheim R. VSDI: a new era in functional imaging
of cortical dynamics. Nat Rev Neurosci. 2004;5:874–885.
29. Gorji A, Straub H, Speckmann E-J. Epilepsy surgery: perioperative
investigations of intractable epilepsy. Anat Embryol (Berlin).
2005;210:525–537.
30. Andersen P, Andersson SA. Physiological Basis of the Alpha Rhythm.
New York, NY: Meredith; 1968.
31. Caspers H. Relations of steady potential shifts in the cortex to the
wakefulness–sleep spectrum. In: Brazier MAB, ed. Brain Function.
Berkeley, CA: University of California Press; 1963:177–213.
32. Goldring S. DC shifts released by direct and afferent stimulation.
In: Remond A, ed-in-chief. Handbook of Electroencephalography
and Clinical Neurophysiology. Vol 10. Part A. Amsterdam: Elsevier;
1974:12–24.
33. Caspers H, Speckmann E-J. Cortical DC shifts associated
with changes of gas tensions in blood and tissue. In: Remond A,
ed-in-chief. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 10. Part A. Amsterdam: Elsevier; 1974:41–65.
34. Elger CE, Speckmann E-J. Penicillin-induced epileptic foci in
the motor cortex: vertical inhibition. Electroencephalogr Clin
Neurophysiol. 1983;56:604–622.
35. Petsche H, Muller-Paschinger IB, Pockberger H, et al. Depth profiles
of electrocortical activities and cortical architectonics. In: Brazier
MAB, Petsche H, eds. Architectonics of the Cerebral Cortex. IBRO
Monograph Series. Vol 3. New York, NY: Raven Press; 1978:257–280.
36. Klee MR, Lux HD, Speckmann E-J, eds. Physiology and
Pharmacology of Epileptogenic Phenomena. New York, NY: Raven
Press; 1982.
37. Purpura DP, Penry JK, Tower DB, et al., eds. Experimental Models
of Epilepsy. New York, NY: Raven Press; 1972.
38. Jasper HH, Ward AA, Pope A, eds. Basic Mechanisms of the
Epilepsies. Boston, MA: Little, Brown & Company; 1969.
39. Elger CE, Speckmann E-J. Focal interictal epileptiform discharges
(FIED) in the epicortical EEG and their relations to spinal field
potentials in the rat. Electroencephalogr Clin Neurophysiol.
1980;48:447–460.
40. Elger CE, Speckmann E-J, Caspers H, et al. Focal interictal epilep-
tiform discharges in the cortex of the rat: laminar restriction and
its consequences for activity descending to the spinal cord. In:
Klee MR, Lux HD, Speckmann E-J, eds. Physiology and
Pharmacology of Epileptogenic Phenomena. New York, NY: Raven
Press; 1981.
41. Gumnit R. DC shifts accompanying seizure activity. In: Remond A,
ed-in-chief. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 10. Part A. Amsterdam: Elsevier; 1974:66–77.
■ Neurophysiologic Basis of EEG and DC Potentials 31
42. Petsche H, Pockberger H, Rappelsberger P. Current source density
studies of epileptic phenomena and the morphology of the rabbit’s
striate cortex. In: Klee MR, Lux HD, Speckmann E-J, eds.
Physiology and Pharmacology of Epileptogenic Phenomena. New
York, NY: Raven Press; 1981.
43. Speckmann E-J, Elger CE, eds. Epilepsy and Motor System.
Baltimore: Urban & Schwarzenberg; 1983.
44. Wieser HG. Electroclinical Features of the Psychomotor Seizure. A
Stereoencephalographic Study of Ictal Symptoms and
Chronotopographical Seizure Patterns Including Clinical Effects of
Intracerebral Stimulation. Stuttgart: Gustav Fischer; 1983.
45. Speckmann E-J, Caspers H, Janzen RWC. Laminar distribution of
cortical field potentials in relation to neuronal activities during
seizure discharges. In: Brazier MAB, Petsche H, eds. Architectonics
of the Cerebral Cortex. IBRO Monograph Series. Vol 3. New York,
NY: Raven Press; 1978:191–209.
46. Speckmann E-J, Caspers H, Janzen RWC. Relations between corti-
cal DC shifts and membrane potential changes of cortical neurons
associated with seizure activity. In: Petsche H, Brazier MAB, eds.
Synchronization of EEG Activity in Epilepsies. New York, NY:
Springer; 1972:93–111.
47. Leao AAP. Spreading depression of activity in the cerebral cortex.
J Neurophysiol. 1944;7:359–390.
48. Somjen GG. Mechanisms of spreading depression and hypoxic
spreading depression-like depolarization. Physiol Rev.
2001;81:1065–1096.
49. Gorji A, Scheller D, Straub H, et al. Spreading depression in
human neocortical slices. Brain Res. 2001;906:74–83.
50. Berger M, Speckmann EJ, Pape HC, et al. Spreading depression
enhances human neocortical excitability in vitro. Cephalalgia.
2008;28:558–562.
51. Gorji A. Spreading depression: a review of the clinical relevance.
Brain Res Brain Res Rev. 2001;38:33–60.
52. Smith JM, Bradley DP, James MF, et al. Physiological studies of
cortical spreading depression. Biol Rev Camb Philos Soc.
2006;81:457–481.
53. Bures J, Buresova O, Krivanek J. The Mechanisms and Applications
of Leao’s Spreading Depression of Electroencephalographic Activity.
New York, NY: Academic Press; 1974.
54. Kager H, Wadman WJ, Somjen GG. Conditions for the triggering
of spreading depression studied with computer simulations.
J Neurophysiol. 2002;88:2700–2712.
55. Somjen GG, Kager H, Wadman WJ. Computer simulations of neu-
ron–glia interactions mediated by ion flux. J Comput Neurosci.
2008;25(2):349–365.
56. Strong AJ, Fabricius M, Boutelle MG, et al. Spreading and synchro-
nous depressions of cortical activity in acutely injured human
brain. Stroke. 2002;33:2738–2743.
57. Caspers H, Speckmann E-J, Lehmenkühler A. Effects of CO2 on
cortical field potentials in relation to neuronal activity. In:
Speckmann E-J, Caspers H, eds. Origin of Cerebral Field Potentials.
Stuttgart: Thieme; 1979:151–163.
58. Speckmann E-J, Caspers H. The effect of O2 and CO2 tensions in the
nervous tissue on neuronal activity and DC potentials. In: Remond
A, ed-in-chief. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 2. Part C. Amsterdam: Elsevier; 1974:71–89.
 CHAPTER

Cellular Substrates of Brain Rhythms*

FLORIN AMZICA AND FERNANDO H. LOPES DA SILVA
Q
uasiperiodically recurring waves, henceforth termed
oscillations, are, besides evoked potentials, one of the
essential components of the electroencephalogram
(EEG). The curiosity of understanding the cellular counterparts
of EEG waves started almost simultaneously with the discovery
and description of different EEG patterns during the early 1930s
and 1940s as an effort to relate the global graphic aspect of the
EEG with its more intimate cellular counterparts. In this chap-
ter, we will discuss how various patterns of oscillatory activity
develop in cellular brain networks and are further reflected at
the EEG level. The EEG is essentially an electric phenomenon;
therefore, we will focus on the ionic currents active during vari-
ous oscillating patterns but we will also identify the (minimal)
anatomical structures necessary and sufficient for the genesis of
the aforementioned activities. This approach has unavoidably
oversimplifying virtues, while the EEG signal is a complex one.
This reality will be recreated through the coalescence of various
oscillations that coexist during a given behavioral context.
From its very beginning, the EEG was characterized through
its oscillatory nature. Thus the dominant frequency of an oscil-
lation became its main descriptor. This trend was further
enforced with the advent of spectral analysis (e.g., Fourier
transforms) and is still largely used nowadays. Although this
might enclose certain advantages, it also contains numerous
pitfalls that prevent the reaching of correct conclusions. This
situation is generated by the fact that a given spectral compo-
nent might ambiguously result either from the presence in the
EEG of oscillations or from the contribution of particular
waveforms, not necessarily rhythmic, to the EEG spectrum.
In the following we will organize this chapter according to
the frequency bands traditionally encountered in the EEG
praxis, however, emphasizing the possible sources of haziness
related to the interpretation of EEG signals. Furthermore, as a
section devoted to the understanding of the cellular mecha-
nisms underlying the genesis of the oscillatory behavior of the
EEG, the intracellular data constitutes a very strong, although
not absolute, source of knowledge. The reader should, however,
keep in mind that, when extrapolating from cellular data to
EEG, one should not overlook essential details, such as
1. Cellular recordings are almost exclusively performed in ani-
mals, whose philogenetical development, behavioral, and
structural peculiarities are often neglected.
3
2. The study of the intrinsic properties of cells is mostly carried
out in vitro or in cultures, preparations that are as far as one can
imagine from the complex reality of the whole brain, both in
terms of network linkages and physiologic state. At best, these
preparations would correspond to a deeply comatose brain.
3. The study of the network interactions, however multisite
they might be, is still based on recordings from spatially dis-
crete and limited locations. The assumption of continuity
should therefore bear reserve.
4. Anesthesia is often an unavoidable companion of animal
studies, and the elimination of the pharmacologic effect is
time and again more complicated than a mere subtraction.
SLOW DELTA RHYTHMS
Grey Walter (1) was the first to assign the term “delta waves” to
particular types of slow waves recorded in the EEG of humans.
Although Walter introduced the term delta waves in correspon-
dence to pathologic potentials due to cerebral tumors, with
time delta activities became more related to sleep and anesthe-
sia. The IFSECN (2) defines delta waves as waves with a dura-
tion of more than 0.25 seconds (which implies a frequency
band between 0 and 4 Hz). Thus the delta term is associated
with frequency bands rather than with phenomena generating
specific electrographic patterns. There have been various stud-
ies aiming at disclosing the relationship between cellular activ-
ities and EEG (3) and the sources of delta activities (4,5). In the
following it will be emphasized that the 0- to 4-Hz frequency
range reflects more than one phenomenon and that definitions
based exclusively on frequency bands may conceal the underly-
ing mechanism. Studies in the last two decades have unveiled
the electrophysiologic substrates of several distinct activities in
the frequency range below 4 Hz during sleep and anesthesia.
Their interaction within corticothalamic networks yields a
complex pattern whose reflection at the EEG level takes the
shape of polymorphic waves.
It should also be stressed from the beginning that delta activ-
ities cover two EEG phenomena: waves and oscillations, with no
clear separations between the two, as far as spectral analysis is
concerned. The term oscillation designates a repeated variation
of a parameter (e.g., current, voltage) between two extreme val-
ues, with an optional requirement of regularity of the variation.

*This chapter was initiated and written in the previous editions by the late Dr. Mircea Steriade (1924–2006). The present authors wish to pay tribute to his outstand-
ing work that has much advanced our understanding of neuronal mechanisms underlying EEG rhythms. They hope to continue and preserve the spirit of his
contributions.

33
34 Part I ■ Basic Principles
However, most of the biologic phenomena resulting from large
population interactions, such as the EEG, are rarely associated
with clocklike rhythmicity. At the other limit of the definition is
the wave, as a single variation of a parameter between two
extreme values. From the above definitions it appears that oscil-
lations are made of waves.
As for now, there are at least two cellular sources of delta
activities: one originating in the thalamus and the other one in
the cortex.
Thalamic delta oscillations have been found in a series of in
vitro studies. They revealed that a clocklike oscillation within the
delta frequency range is generated by the interplay of two intrin-
sic currents of thalamocortical cells. Whereas most of brain
oscillations are generated by interactions within networks of
neurons, but also glial cells, the thalamic delta oscillation is an
intrinsic oscillation depending on two inward currents of thala-
mocortical cells (7–9). It was shown that thalamocortical neu-
rons recorded in vitro display rhythmic bursts of high-frequency
spikes with an interburst frequency of 1 to 2 Hz. This oscillation
results from the interplay between two membrane currents: the
transient calcium current (It) underlying the low-threshold
spike (LTS) and a hyperpolarization-activated cation current
(Ih). Ih is a noninactivating inward (anomalous) rectifier carried
by sodium and potassium. The model proposed for the genesis
of this clocklike delta oscillation is depicted in Figure 3.1A: when
the membrane potential hyperpolarizes (more negative than
–70 mV), the Ih becomes activated and produces a depolarizing
sag, which in turn will activate the It (deinactivated by the hyper-
polarized membrane). The ensuing LTS deactivates the Ih but
also generates a hyperpolarizing overshoot that will reactivate
the Ih, thus restarting a new oscillatory cycle.
This oscillation was also found in vivo, in the thalamocorti-
cal neurons of the cat, after decortications (10–12).
Thalamocortical neurons from a variety of sensory, motor, asso-
ciational, and intralaminar thalamic nuclei were able to display
a clocklike delta rhythm either induced by imposed hyperpolar-
izing current pulses (Fig. 3.1B) or spontaneously (Fig. 3.1C).
Interestingly, the deafferentation achieved through decortica-
tion, by removing the powerful depolarizing impingement from
corticothalamic neurons and setting thalamic cells at a more
hyperpolarized membrane potential (around –70 mV), was
instrumental in permitting the delta oscillation to emerge.
However, in intact preparations, with functional corticothalamic
loops, the regular thalamic delta oscillation was absent or largely
prevented by the ongoing cortical activity (13). At the cellular
level, the occurrence of the Ih was blocked during the depolariz-
ing phases of the slow oscillation ( 1 Hz; see below), due to the
interference of synaptic currents and the ensuing decreased
membrane resistance with the intrinsically generated Ih. These
more recent findings raised concerns as to the emergence of this
intrinsically generated thalamic rhythm at the level of the EEG.
The existence of cortical delta oscillations was suggested on
the basis that delta waves, at 1 to 2 Hz, survive in the EEG of
athalamic cats (14). Whether such procedures generate a phys-
iologic or a pathologic pattern remains an open question;
nonetheless, they create a deafferentation of the cortex.
Extracellular recordings of cortical activity during pathologic
delta waves (as obtained by lesions of the subcortical white mat-
ter, the thalamus, or the mesencephalic reticular [RE] forma-
tion) have shown a relationship between the firing probability
and the surface-positive (depth-negative) delta waves, whereas
the depth-positive waves were associated with a diminution in
discharge rates (15). These field–unit relationships led to the
assumption that the depth-positive component of delta waves
reflects the inhibition of pyramidal-shaped neurons by local-
circuit cells, implying maximal firing of inhibitory interneurons
during the depth-positive delta waves. However, this has not
been found. It was then suggested that, far from resulting exclu-
sively from inhibitory postsynaptic potentials (IPSPs), EEG
delta waves are rather generated by summation of long-lasting
afterhyperpolarizations produced by a variety of potassium
currents in deeply lying pyramidal neurons (16,17).
The discovery of a novel slow-wave sleep- or anesthesia-related
oscillation (around, but generally less than 1 Hz) (18), together
with its thorough investigation at the cellular level (see below), set
the basis for revisiting the cellular basis of delta rhythms. The fre-
quency of the slow oscillation depends on the depth or type of
anesthesia, or the sleep stage: it is mainly between 0.3 and 0.6 Hz
under urethane anesthesia, between 0.6 and 0.9 Hz under keta-
mine-xylazine anesthesia, and between 0.7 and ~1 Hz during nat-
ural sleep. This oscillation, termed slow oscillation by its authors,
was first described in intracellular recordings from neocortical
neurons in anesthetized animals and in EEG recordings from
humans (18). It was subsequently found during natural slow-
wave sleep of animals (19) and humans (20–24).
The slow oscillation is generated within the cortex because it
survives after thalamectomy (25), is absent from the thalamus
of decorticated animals (26), and is present in cortical slices
(27). At the intracellular level, cortical neurons throughout lay-
ers II to VI (with physiologically identified thalamic and/or cal-
losal inputs, and with projections to the thalamus and/or
homotopic points of the contralateral hemisphere) displayed a
spontaneous oscillation recurring with periods of 1–1.5 to
5 seconds, depending on the anesthetic, and consisting of pro-
longed depolarizing and hyperpolarizing components (Fig.
3.2A). The long-lasting depolarizations of the slow oscillation
consisted of excitatory postsynaptic potentials (EPSPs), fast
prepotentials (FPPs), and fast IPSPs, reflecting the action of
synaptically coupled GABAergic local-circuit cortical cells (18).
Data also demonstrated the contribution of both NMDA-
mediated synaptic excitatory events and a voltage-dependent
persistent sodium current, INa(p), to the depolarizing component
of the slow oscillation. On the other hand, the long-lasting
hyperpolarization, interrupting the depolarizing events, is asso-
ciated with network disfacilitation in the cortex. This is sup-
ported by the fact that the membrane input resistance is highest
during the long-lasting hyperpolarizing component of the slow
oscillation (28). The disfacilitation is probably achieved by a
progressive depletion of the extracellular calcium ions during
the depolarizing phase of the slow oscillation (29) (Fig. 3.2B–C).
All major cellular classes in the cerebral cortex, as identified
by electrophysiologic characteristics and intracellular staining,
display the slow oscillation: regular spiking and intrinsically
bursting cells, as well as local-circuit inhibitory basket cells
Figure 3.1 Delta oscillations in thalamocortical cells result from the interplay between two intrinsic currents, Ih and It. A: Proposed model for interac-
tion between these intrinsic currents. Activation of the low-threshold calcium current, It, depolarizes the membrane toward threshold for a burst of
sodium-dependent fast action potentials. The depolarization inactivates the portion of Ih that was active immediately before the calcium spike.
Repolarization of the membrane due to It inactivation is followed by a hyperpolarizing overshoot, due to the reduced depolarizing effect of Ih. The hyper-
polarization in turn deinactivates It and activates Ih, which depolarizes the membrane toward threshold for another calcium spike. (Modified from
McCormick DA, Pape HC. Properties of a hyperpolarization-activated cation current and its role in rhythmic oscillation in thalamic relay neurones.
J Physiol. 1990;431:291–318.) B: Delta oscillation of cat ventrolateral thalamocortical cell triggered by hyperpolarizing current pulses (0.7 nA in 1, 1 nA
in 2, 1.1 nA in 3, and 1.2 nA in 4). Note increasing number of cycles at a frequency of 1.6 Hz. C: Latero-posterior (LP) thalamocortical cell after decor-
tication of areas projecting to LP nucleus. The cell oscillated spontaneously at 1.7 Hz. A 0.5-nA depolarizing current (between arrows) prevented the
oscillation, and its removal set the cell back in the oscillatory mode. Three cycles after removal of depolarizing current in 1 are expanded in 2 to show
high-frequency bursts crowning the low-threshold calcium spike. In this and following figures, polarity of EEG and field potentials is as for intracellular
recordings (positivity up). (Modified from Steriade M, Curró Dossi R, Nuñez A. Network modulation of a slow intrinsic oscillation of cat thalamocorti-
cal neurons implicated in sleep delta waves: cortically induced synchronization and brainstem cholinergic suppression. J Neurosci. 1991;11:3200–3217.)

35
36 Part I ■ Basic Principles

Figure 3.2 A: Pyramidal cell from the somatosensory cortex (area 3b) during the slow ( 1 Hz) oscillation. The left panel
shows intracellular recordings and simultaneously recorded EEG in the vicinity (~ 1 mm) of the cell. The EEG was recorded
by means of coaxial electrodes located on the surface and at a depth of approximately 0.6 mm. The cell oscillated at 0.9 Hz
with depolarizing phases corresponding to depth-EEG negative (surface-positive) potentials. The right panel shows the cor-
responding cell stained with Neurobiotin (calibration bar in mm). (Modified from Contreras D, Steriade M. Cellular basis
of EEG slow rhythms: a study of dynamic corticothalamic relationships. J Neurosci. 1995;15:604–622.) B: Fluctuations dur-
ing the slow oscillation. Relationships between intracellular membrane potential, extracellular calcium [Ca] out, and field
potential. Periodic neuronal depolarizations triggering action potentials were interrupted by periods (300–500 msec) of
hyperpolarization and silenced synaptic activity. [Ca]out dropped by about 0.25 mM during the depolarizing phase reach-
ing a minimum just before the onset of the hyperpolarization. Then, [Ca] out rose back until the beginning of the next cycle.
C: Thirty cycles were averaged (spikes from the neuronal signal were clipped) after being extracted around the onset of
the neuronal depolarization. The vertical dotted lines tentatively indicate the boundaries of the two phases of the slow
oscillation (B–C modified from Massimini M, Amzica F. Extracellular calcium fluctuations and intracellular potentials in
the cortex during the slow sleep oscillation. J Neurophysiol. 2001;85:1346–1350.).
 Chapter 3 ■ Cellular Substrates of Brain Rhythms 37

Figure 3.3 Effect of lidocaine inactivation on synchrony between intracellular activities and FPs. A: Double intracellular
recording in the anterior and posterior parts of the suprasylvian gyrus (see scheme). Synchrony was present between all
three recording sites. Intracortical injection of lidocaine (40 L, 20% ; bottom) inactivated the electrical activity close to
the cannula in the middle part of the gyms, disrupted the synchrony between pools of neurons, and between individual
neurons. B: Sequential correlation analyses of intracellular and field potential data. Two-minute periods were analyzed
with the sequential correlation method. (Modified from Amzica F, Steriade M. Disconnection of intracortical synaptic link-
ages disrupts synchronization of a slow oscillation. J Neurosci. 1995;15:4658–4677.)

(18,30,31). All these neuronal types exhibit similar relation-
ships with the EEG components of the slow oscillation: during
the depth-positive EEG wave cortical neurons are hyperpolar-
ized, whereas during the sharp depth-negative EEG deflection
cortical neurons are depolarized (Fig. 3.2). The spectacular
coherence between all types of cortical neurons and EEG wave-
forms raised the question of the underlying synchronizing
mechanisms. Dual intracellular recordings in vivo revealed that
intracellular potentials were highly synchronized among neu-
rons (31,32). This synchronization, displaying time lags as a
function of the distance between cortical areas (33), mainly
relies on the integrity of intracortical synaptic linkages.
Although the coherence was impaired after local blockage of
axonal transmission through local injections of lidocaine (Fig.
3.3), the slow oscillation survived at both locations and con-
served a rudiment of synchronization suggesting that, besides
direct intracortical linkages, other projections, possibly cortico-
thalamo-cortical, as well as networks of gap junctions (see
below) might contribute to the synchronization of the slow
oscillation.
Such cortical coherence is expected to equally involve the
thalamus. Indeed, RE thalamic cells exhibit similar and time-
locked variations of their membrane potential, with prolonged
depolarizations interrupted by prolonged hyperpolarizations
(34) in reflection of the excitatory cortico-RE projections. The
depolarizing component in RE neurons is transmitted to thala-
mocortical neurons, via GABAergic projections, where it trig-
gers rhythmic IPSPs, leading to rebound spike bursts (Fig. 3.4),
which in turn will generate phasic excitations in the cortex,
shaping the steady depolarization of cortical neurons.
The better comprehension of the mechanisms determining
the pacing of the slow oscillations and the switch between
depolarizing and hyperpolarizing phases came from experi-
ments considering the possible dialogue between neurons and
glial cells. Dual simultaneous intracellular recordings from neu-
rons and adjacent glial cells were performed during sleeplike
patterns produced by ketamine-xylazine anesthesia, exploring
the possibility that glia may not only passively reflect, but also
influence the state of neuronal networks (35–37). The behavior
of simultaneously recorded neurons and glia is illustrated in
38 Part I ■ Basic Principles

Figure 3.4 The slow oscillation (~ 0.9 Hz) in dual simultaneous intracellular recordings from a regular-spiking cell
in cortical area 4 and a thalamocortical cell in the ventrolateral nucleus. Cat under ketamine-xylazine anesthesia.
Arrow in A points to a low-threshold spike burst. An expanded cycle is shown in B. Notes: (a) Depth-positive
(upward) EEG waves are associated with hyperpolarization of cortical and thalamic cells, whereas the sharp depth-
negatives are associated with depolarization and action potentials in cortical cell, while the thalamic neuron displays
a rebound spike burst with a delay of 150–200 msec; (b) brief sequence of EEG spindles after depth-negative (surface-
positive) sharp deflection; and (c) fast depolarizing waves (40–50 Hz) in cortical neuron during the sustained depo-
larization. (Unpublished data by M Steriade and D Contreras.)

Figure 3.5. During spontaneously occurring slow oscillations,
the onset of the glial depolarization did, in the vast majority of
the cases, follow the onset of the neuronal depolarization with
an average time lag of around 90 msec (Fig. 3.5). This time lag
is, however, longer than the one obtained from pairs of neurons
(around 10 ms in Ref. 33). The glial depolarization reflects with
virtually no delay the potassium uptake (38). As potassium is
also continuously recaptured by neurons through Na /K -ATP
pumps, the actual onset of the glial depolarization may thus
mark the moment where the increase of extracellular potassium
concentration exceeds the capacity of the neuronal pumps. This
behavior suggests that glial cells are essential for maintaining
extracellular potassium within controlled values.
Toward the end of the depolarizing phase, the glial membrane
begins to repolarize before the neuronal membranes (Fig. 3.5)
(37). Interestingly, this observation may expose the mechanism
ruling the onset of the hyperpolarizing phase of the slow oscilla-
tion. It was initially proposed that this hyperpolarization relies
either on active GABA inhibition or on the activation of slow
calcium-dependent potassium currents. In the first case, the
release of GABA by neurons would either depolarize glial cells
through direct action on GABAA receptors (39,40) or indirectly
through the potassium output resulting from the neuronal acti-
vation of neuronal GABAB receptors. The alternative implication
of calcium-dependent potassium currents should be discarded
because of the fact that the extracellular potassium starts to drop
before the onset of the hyperpolarizing phase. In addition, the
glial membrane potential returns to control value at the end of
each of the recurring oscillatory cycles, leading to the assumption
glial cells might control the pace of the oscillation through
changes in the concentration of the extracellular potassium (38).
The overall synchronization of the slow oscillation in the
cortex is also assisted by glial cells, which are imbedded in a gap
junction-based network, through the phenomenon of spatial
buffering (41–43). Spatial buffering evens local increases of
extracellular potassium by transferring it from the extracellular
milieu to neighboring glial cells, then through gap junctions,
and along the potassium concentration gradient, at more dis-
tant sites with lower concentrations of potassium, where it is
again expelled in the extracellular space. This pathway is pre-
ferred to the direct one through the extracellular space because
of the tortuosity of the latter (reviewed in Ref. 44). The rela-
tionship between intracellular glial potentials and extracellular
potassium (38) shows little accumulation of potassium in the
extracellular space ( 1.1 mM/cycle), and this causes a minimal
imbalance between the extra- and intracellular potassium con-
centrations. Given that normally glial cells deal with low
amounts of potassium but have a high propensity to uptake it
(43,45,48), it is likely that spatial buffering would occur at a
reduced spatial scale, thus contributing to an uniform distribu-
tion of the potassium around the membrane of neurons (38).
The local spatial buffering during slow oscillations might
play two roles: (i) it contributes to the steady depolarization of
neurons during the depolarizing phase of the slow oscillation
in a Nernst-related manner and (ii) it modulates the neuronal
excitability. The latter mechanism could favor the synaptic
interaction within cortical networks at the onset of the depo-
larizing phase of the slow oscillation, but it could equally

induce a gradual disfacilitation as the extracellular potassium
reaches higher values toward the top of the glial depolariza-
tion. The amplitude of the potassium variations during the
slow sleep oscillation ranges between 1 and 2 mM, which,
when added to the physiologic values of resting concentration
(~3 mM), may assist cortical neurons in oscillating between
hyper- and hypoexcitability (47–49). This hypothesis was con-
firmed after testing the excitability of the neurons during the
various phases of the slow oscillation cycle with cortical elec-
tric stimuli (37). The results demonstrate that the neuronal
excitability is maximal toward the beginning of the depolariz-
ing cycle and almost zero toward the end of the depolarizing
phase of the slow oscillation.
THE SLOW OSCILLATION
AND THE K-COMPLEX
One of the functional correlates of the slow cortical oscillation is
the fact that, at the EEG level, each sequence of
depolarizing–hyperpolarizing episodes within an oscillatory
cycle corresponds to a waveform called K-complex. Although
the phenomenon was first described in 1938 as a landmark of
slow-wave sleep (50), its study remained at the EEG level for
Chapter 3 ■ Cellular Substrates of Brain Rhythms 39
Figure 3.5 Simultaneous recording
of a neuron–glia pair in the cortex of
the cat during slow oscillatory pat-
terns. A: Relationship between neu-
ronal and glial depolarizations on
the one hand, and between neu-
ronal hyperpolarization and glial
repolarizations on the other hand
(note enlargement within the box).
The glial impalement occurred dur-
ing this period of the recording and
is depicted as a brisk drop of the
potential from the extracellular
(0 mV) to the intracellular (–90 mV)
level. B: Histogram of membrane
potential variations in the neuron
(in gray) and glial cell (in black). C:
Typical time relationship between
the onset and offset moments of the
two phases of a slow oscillatory
cycle. Note that the onset of the
depolarizing phase starts in the neu-
ron, but that the end of this period is
anticipated in the glia. (Modified
from Amzica F, Massimini M. Glial
and neuronal interactions during
slow wave and paroxysmal activities
in the neocortex. Cereb Cortex.
2002;12:1101–1113.)
decades (see the excellent review in Ref. 51) during which
researchers struggled with the inherent shape variability intro-
duced by various bipolar electrode configurations, reference sys-
tems, and filter settings. These technical difficulties, together
with the relative vagueness of the definition, often led to discor-
dant conclusions and to difficult interpretations. The cellular
mechanisms of the K-complex were only recently described (52)
and disclosed that the K-complex is a recurrent event with the
rhythmicity of the slow oscillation (less than but close to 1 Hz)
and that, through its shape, it is one of the main contributors to
the delta spectrum (Fig. 3.6) (21,35).
The rhythmicity of K-complexes during the onset of sleep is
less obvious due to the lesser organization of the slow oscilla-
tion. This corroborates with the fact that bipolar recordings are
often the only ones allowing the detection of K-complexes in
the initial part of slow-wave sleep. However, with the deepening
of sleep, the slow oscillation becomes more regular and faster
(the frequency approaches 1 Hz). During late stages of sleep
most of the K-complexes may be confounded with delta waves.
The frequency shift (usually from 0.6 to 0.9 Hz) of the slow
oscillation, as well as the shape modification of the K-complex
(Fig. 3.7), accompanying the deepening of sleep can be
explained by taking into account the intrinsic properties of the
40 Part I ■ Basic Principles

Figure 3.6 K-complex in human sleep. Scalp monopolar recordings with respect to the contralateral ear (see figurine).
A: Short episode from a stage 3 period of sleep. The two arrows point to a K-complex (KC) followed by a spindle sequence
( ) and to a KC in isolation. The two K-complexes are embedded into a slow oscillation at about 0.6 Hz. Note the syn-
chrony of K-complexes in all recorded sites and the diminution of their amplitudes in the occipital area. B: Average of
50 K-complexes aligned on the positive peak (in this figure positivity is downwards) of the upper channel
(vertical dotted line). C: Power spectrum of the C3-lead for a period of 80 seconds of stable stage 3 activity
containing the period depicted in A. The three frequency bands (S, , and ; further illustrated in D) are rep-
resented in the power spectrum. Moreover, the slow (S) activity displays a high peak, distinct from the onset of delta
( ) activity (see inset). D: Frequency decomposition of the C3-lead (upper trace) into three frequency bands:
slow (S, 0–1 Hz), delta ( , 1–4 Hz), and sigma ( , 12–15 Hz). It is shown that the K-complex results from a
combination of slow and delta waves. (Modified from Amzica F, Steriade M. The K-complex: its slow ( 1-Hz) rhythmicity
and relation to delta waves. Neurology. 1997;49:952–959.)

neuronal membrane. At sleep onset, neurons in the midbrain
reticular formation and mesopontine cholinergic nuclei dimi-
nish their firing rate (53,54), thus removing a steady excitatory
drive from thalamocortical neurons and allowing the
membrane potential of cortical neurons to reach more hyper-
polarized levels. The progressive hyperpolarization of thalamo-
cortical neurons is a constant correlate of sleep deepening (55)
and induces a gradual reduction of the duration of the depolar-
izing phase of the slow cortical oscillation together with the
acceleration of the rhythm. The former action contributes
mostly to the shape of the field potentials, while the second
determines the oscillatory frequency.
Due to the controversial criteria for shape recognition, the
study of K-complexes has relied mostly on evoked K-complexes
because stimulation generally produces stereotyped responses.
Although of certain experimental value, the evoked K-
complexes have set a bias in the understanding of the underly-
ing mechanisms because it was forgotten that the K-complex is
mainly a spontaneous phenomenon (50). The comparison at
the cellular level between spontaneous and evoked K-
complexes shows, in agreement with the original observation of
Davis et al. (56), that the latter appears as a secondary discharge.
Both spontaneous and evoked K-complexes display similar
depth profiles and current source density (CSD) distributions
 Chapter 3 ■ Cellular Substrates of Brain Rhythms 41

Figure 3.7 Rhythmic KCs in human EEG during natural slow-wave sleep. A: Four leads recorded from the
two hemispheres during stage 2 sleep show quasirhythmic (around 0.5 Hz) KCs. The expanded insets display
a simple K-complex (KC, left) and a K-complex followed by a spindle (right). Stage 3–4 EEG (below) is char-
acterized by a more regular oscillation of the K-complexes (0.7 Hz). Asterisks mark the most obvious
K-complexes in order to suggest their rhythmicity. B: Averaged K-complexes (n 200) from stages 2 (left)
and 3–4 (right). The gray surface around the averaged K-complexes covers the standard deviation. The low-
est trace (dotted line) results from the filtering of the average K-complex in the delta band (1–4 Hz). C: Power
spectra of 2-minute epochs containing the ones displayed above. Note a principal peak at 0.5 Hz for the stage
2 episode (left), surrounded by other lower peaks as evidence for distributed rhythmicity. At variance, deep
sleep shows a dominant peak at 0.7 Hz. The two FFT graphs are scaled with the same ordinate. (Modified
from Amzica F, Steriade M. The K-complex: its slow ( 1-Hz) rhythmicity and relation to delta waves.
Neurology. 1997;49:952–959.)
42 Part I ■ Basic Principles
(52), further supporting their common origin and generating
mechanisms.
The complex electrographic pattern of slow-wave sleep
results from the coalescence of the slow oscillation with sleep
spindles (see below) and, possibly, thalamic delta oscillations
(57). The weight of each of these major components is dynam-
ically modulated during sleep by synaptic coupling, local-
circuit configurations, and the general behavioral state of the
network. The discussion of the functional significance of the
K-complex is determined by the role of the slow sleep oscilla-
tion of which the K-complex is the electrographic reflection.
Through its rhythmic occurrence at a frequency that is lower
than the one of other sleep rhythms (spindles and delta), and
due to its wide synchronization at the cortical level, the sharp
onset of the K-complex provides a synchronous input to thala-
mic neurons, thus triggering and/or grouping sequences of
spindles or delta oscillations.
THETA RHYTHMS
Theta waves are usually defined as occurring within the 4- to
7-Hz frequency range. The normal theta activity should not be
confused with pathologic theta waves, described as a slowing
down of alpha activity, expressed during cerebral blood flow
reduction (58) or metabolic encephalopathies (59). Rhythmic
activities in the theta frequency range are conspicuous in limbic
areas in various animal species as extensively reviewed by
Buzsáki (60).
Limbic Theta Rhythms in Humans
The presence of theta rhythms was first denied in humans
(61,62), but they were later recorded in the hippocampus of
epileptic patients with indwelling electrodes (63–65) or by way
of mesio-temporal corticography with foramen ovale elec-
trodes (66), and also using magnetoencephalography (MEG) in
normal subjects (67,68). In all these studies human hippocam-
pal theta rhythm was studied under specific behavioral condi-
tions. In the study of Arnolds et al. (69) the hippocampal
activity was recorded while the subject was freely moving and
performing a cognitive task; the frequency and rhythmicity of
the hippocampal theta component while the subject was writ-
ing was consistently higher than while sitting or walking. In a
word association task the amplitude, frequency, and rhythmic-
ity showed a significant rise during the period of silent mental
activity immediately following a verbal cue to which the subject
was requested to give a verbal answer. Ekstrom et al. (65) found
evidence for movement-related theta oscillations in human
hippocampus and suggested that both cortical and hippocam-
pal oscillations play a role in attention and sensorimotor inte-
gration. Kahana et al. (64) investigated hippocampal EEG
activity using subdural recordings from epileptic patients dur-
ing spatial navigation in computer-generated mazes and found
that theta rhythmic activity occurred more frequently when the
subject was confronted with solving complex mazes. In addi-
tion, theta oscillations were more evident during recall trials
than during learning trials. Bódizs et al. (66) recorded in the
human hippocampus a rhythmic activity with frequency at the
lowest end of the theta frequency range that presented larger
power spectrum during rapid eye movement (REM) sleep than
during other states and suggested that this oscillation may be
considered the counterpart of the hippocampal theta of mam-
malian REM sleep. Tesche and Karhu (67) recorded MEG sig-
nals in human subjects and reconstructed from these signals the
corresponding hippocampal sources of theta rhythmic activity.
They reported that during the presentation of a memory task
the duration of theta bursts increased with memory load and
suggested a relation between “stimulus-locked hippocampal
theta” and the processing of working memory. Using a whole-
head 275- channel MEG system in normal subjects, Cornwell et
al. (68) reported hippocampal and parahippocampal theta
oscillations during a spatial navigation task (virtual reality
Morris water maze). They carried out source analyses by means
of which they were able to show larger theta activity in the left
anterior hippocampus and parahippocampal cortices during
goal-directed navigation relative to aimless movements. Thus,
similar to other species (see below), the human theta may rep-
resent a dynamic state emerging from hippocampal networks
engaged in spatial navigation and in memory processes.
Limbic Theta Rhythms in Animals
Stewart and Fox (69), working on urethane-anesthetized pri-
mates, recorded hippocampal theta activity similar to that of rats
but at a slightly higher frequency and appearing as relatively
short bursts. Arnolds et al. (63) made a comparative study of
hippocampal theta rhythms in cat, dog, and human and showed
that all three displayed the same type of qualitative changes in
spectral parameters (peak frequency, peak amplitude, and
rhythmicity) in relation to behavioral tasks involving motor
activities. Particularly in cat, remarkable relationships between
theta spectral properties and vestibular stimulation (e.g., body
acceleration) and/or eye movements were put in evidence. Theta
rhythms appear conspicuously in rodents such as rabbits (first
description of the rhythm was made by Green and Arduini (70))
and rats in relation with sensory processing and different types
of movements (71) and have been recorded in several structures
of the cortical limbic system (72,73). Depending on species and
conditions, these rhythmic may extend from 3–4 to 10 Hz,
which is somehow larger than the conventional range of the
EEG theta rhythm (4 to 7 Hz). This is the reason why these lim-
bic theta oscillations are also named “rhythmic slow activity
(RSA)” in animal neurophysiology to avoid confusion with the
classic human EEG theta frequency band. The two designations
are commonly used in the literature.
In the rat, theta oscillations are most regular in frequency
and present the largest amplitude in the stratum lacunosum-
moleculare of the hippocampal CA1 region but are also found
in the dentate gyrus and the CA3 region. In addition to the hip-
pocampal formation, theta oscillations have been observed in
the subiculum, entorhinal cortex, perirhinal cortex, amygdala,
and cingulate cortex (74). Although these areas are capable of
displaying theta oscillations that can be recorded extracellularly,
they are not able to generate theta activity on their own, that is,
isolated from the rest of the brain, unless manipulated by intra-
cellular current injection or by the addition of pharmacologic

substances. In this respect it is interesting to mention that neu-
rons of several limbic structures have intrinsic membrane prop-
erties that might facilitate their entrainment in theta
oscillations as shown in vitro. This was demonstrated for some
types of CA1 interneurons depolarized near spike threshold by
current injection that generated rhythmic firing at about 7 Hz
(75) and also for neurons of the perirhinal cortex that can dis-
play oscillations around 8 to 9 Hz when depolarized at thresh-
old levels (76). These intrinsic properties could facilitate the
entrainment of these neurons in theta frequency oscillations.
Theta Rhythm Generation
Since the early years of this research line several studies were
carried out with the aim of finding a system that may act as
theta “pacemaker.” The search for a theta pacemaker has led to
controversial results: first it was claimed that the medial
septo/diagonal band-hippocampal cholinergic system, driven
by the brainstem reticular core, is the “pacemaker” of theta
(77,78). The medial septum provides cholinergic and
GABAergic input to the hippocampus. While cholinergic neu-
rons innervate both principal cells and interneurons,
GABAergic projection neurons from the medial septum make
synapses on hippocampal interneurons, including parvalbu-
min-positive basket cells (79).
More recent investigations revealed that hippocampal
interneurons with long-range projections also innervate the
cells of origin in the medial septum. Further, the supramammil-
lary nucleus of the hypothalamus strongly connected to the
medial septum, and to extended networks of the brainstem and
diencephalon, has also been proposed to play a role in pace-
making and modulating hippocampal theta (80).
The fact that theta rhythms survive after the blockade of
cholinergic muscarinic inputs from the septal area using
atropine brought to the fore the existence of an additional non-
cholinergic source of theta activity in the entorhinal cortex (81).
Indeed, a dipole of theta activity was found also in the entorhi-
nal cortex with two amplitude maxima, one superficial in layers
I–II and the other in layer III (82–85). We may conclude that
the classic view of a septal/diagonal band pacemaker impinging
rhythmic activity on the pyramidal cells of the hippocampus,
within the theta frequency band, is too simplistic.
Many investigations analyzed the characteristics of the two
types of hippocampal theta: one atropine-sensitive and the other
atropine-resistant (86). Removal of the entorhinal cortex makes
hippocampal theta atropine-sensitive, so that it may be con-
cluded that the atropine-resistant theta originates from the
entorhinal inputs to the hippocampus (81,87). These inputs are
glutamatergic and it was shown that NMDA synapses on the dis-
tal apical dendrites of CA1 pyramidal neurons are important
sources of theta-generating inputs (73). Which pathways are
responsible for the atropine-sensitive theta component is still
unclear despite many investigations using sophisticated methods
such as high-density 96-site silicon probes enabling recordings
simultaneously from the dentate gyrus, CA3 and CA1 areas (88).
Cholinergic inputs can cause excitation of interneurons that
may be responsible for theta rhythmic discharges of hippocampal
interneurons (79,89). These interneurons, discharging at the theta
Chapter 3 ■ Cellular Substrates of Brain Rhythms 43
frequency, can thus cause rhythmic IPSPs on their target pyrami-
dal cells. Consequently somatic outward currents at the level of
the CA1 pyramidal layer can contribute to the theta extracellular
field (81,90). Further cholinergic agonists applied to hippocampal
slices cause rhythmic depolarizations of lacunosum-moleculare
interneurons at the theta frequency, which are blocked by atropine
(91). This indicates that muscarinic induction of theta-frequency
membrane potential oscillations in lacunosum moleculare
interneurons may contribute to the generation of rhythmic inhi-
bition that paces intrinsically generated theta activity in CA1
pyramidal cells. In addition, the CA3 can also contribute to the
generation of theta field potentials, since the recurrent network of
CA3 pyramidal cells, and possibly hilar mossy cells, may form an
intrahippocampal oscillator that may contribute to the atropine-
sensitive component (73).
A classic view (92) is that hippocampal CA1 extracellular
field during theta can be accounted for by a two-dipole model.
According to this view the CA1 pyramidal cells receive rhyth-
mic inputs in the theta frequency range from other sources that
consist of atropine-sensitive and atropine-resistant inputs. The
former would drive the somata and the latter the distal den-
drites. The atropine-sensitive theta rhythm would be mainly
caused by Cl--mediated IPSPs on pyramidal cells (93,94).
Further intrinsic voltage-dependent membrane potential oscil-
lations may modulate the response to a theta frequency driving,
although hippocampal pyramidal neurons do not oscillate in
isolation (60). They show, however, in vitro, resonant properties
in the theta frequency range that are determined by the inter-
play between two main ionic currents with different kinetics,
the Ih and Im currents. The former is a hyperpolarization-
activated cation current and the latter is a muscarinic K cur-
rent. In contrast, neurons of the entorhinal cortex layer II show
subthreshold oscillations at the theta frequency that appear to
result from the interplay of the Ih and the depolarization-
activated persistent Na current (60).
More recent studies (88), however, show that hippocampal
theta is associated with multiple current sinks and sources with
a wide range of phase relationships between the firing of
pyramidal cells and of interneurons to the theta recorded
extracellularly from different hippocampal layers.
Intrahippocampal Inhibitory System
and Theta Generation
The phase relationship between spike firing and the polarity of
the extracellular field depends on the relative strengths of the
dendritic depolarization and somatic inhibition. The insightful
review of Somogyi and Klausberger (95) summarizes the com-
plex interrelationship between the firing of different popula-
tions of inhibitory interneurons and pyramidal cells during
theta oscillations (Fig. 3.8). During theta oscillations several
types of interneurons exhibited considerable variability in the
relation between firing probability and the mean theta phase
recorded extracellularly. Figure 3.8 shows the relationships of
the firing of different kinds of interneurons not only with theta
waves, but also with high-frequency ripples and oscillations in
the gamma frequency range. Here we limit ourselves to a
description of the former. Axo-axonic cells displayed the highest
44 Part I ■ Basic Principles

Figure 3.8 Spatiotemporal interaction between pyramidal cells and several classes of interneurons during network oscillations,
shown as a schematic summary of the main synaptic connections of pyramidal cells (P), parvalbumine (PV)-positive-expressing
basket, axo-axonic, bistratified, oriens-lacunosum moleculare (O-LM), and three classes of cholecystokinin (CCK)-expressing
interneurons. The firing probability histograms show that interneurons innervating different domains of pyramidal cells fire with
distinct temporal patterns during theta and ripple oscillations, and their spike timing is coupled to field gamma oscillations to dif-
fering degrees. The same somatic and dendritic domains receive differentially timed input from several types of GABAergic
interneuron. ACh, acetylcholine. (Modified from Klausberger T, Somogyi P. Neuronal diversity and temporal dynamics: the unity
of hippocampal circuit operations. Science. 2008;321:53–57.)

firing rate around the positive peak of extracellular theta field
recorded at the level of the pyramidal layer, that is, when the
average pyramidal cells are most hyperpolarized. Parvalbumin
immunopositive basket cells fired during the descending phase
of the theta oscillations were recorded at the same level.
Bistratified cells innervating strati radiatum and oriens were
fired during the trough of the extracellular theta in the pyram-
idal cell layer. The oriens-lacunosum moleculare cells terminat-
ing on the most distal dendrites in conjunction with the
entorhinal input were also fired during the trough of the theta
field potential.
In addition it should be noted that some inhibitory interneu-
rons make synapses with other inhibitory interneurons such as
described by Banks et al. (96) who demonstrated an interaction
between two groups of CA1 interneurons defined according to
the kinetics of the corresponding IPSCs: GABAA generating
IPSPs with slow kinetics and GABAA with fast kinetics that may
contribute to theta and gamma rhythms, respectively. Most
interesting is the fact that incoming stimuli in stratum lacuno-
sum moleculare do not only inhibit pyramidal cells, by activat-
ing “GABA-slow” IPSCs in these cells, but simultaneously inhibit
“GABA-fast” interneurons. The authors hypothesize that this
inhibitory–inhibitory interaction contributes to nested
theta/gamma rhythms in the hippocampus. In Chapter 4, we
describe that an alteration of this inhibitory interaction may
constitute an important mechanism for the occurrence of
epileptogenic activity in the hippocampus (97). Fuentealba et al.
(98) also described more recently a subpopulation of
GABAergic interneurons expressing enkephalin that innervate
parvalbumin-positive interneurons and thus may contribute to
the organization of rhythmic activities in CA1 area.
The peak of the intracellular theta recorded in the dendrites
of pyramidal cells slightly lags the peak of the extracellular theta
in the pyramidal cell layer (99).
An important new finding (100) that further emphasizes the
role of interneurons in the generation of hippocampal theta
was obtained using transgenic mice where the fast synaptic
GABAergic inhibition was removed from parvalbumin-positive
interneurons. In CA1, these interneurons comprise bistratified
and possibly oriens-lacunosum moleculare cells that target
dendrites, and basket cells that target somata and axo-axonic
cells (101). Each of the perisomatic inhibitory parvalbumin-
positive cells innervates more than 1000 pyramidal cells (102)
and causes IPSPs in the pyramidal cell layer. Through synaptic
GABAA receptors, parvalbumin-positive basket cells recipro-
cally inhibit each other and receive inhibition from the medial
septum and other local interneurons (103). Wulff et al. (100)
found that this interneuron-specific ablation of the inhibitory
septohippocampal projection can severely impair theta
rhythm in the hippocampus. This study implies that the

intrahippocampal systems of inhibition of parvalbumin-
positive cells contribute to theta generation.
From all these data it emerges that to understand theta gen-
eration in the hippocampus and associated areas it is necessary
to integrate single neuronal data with knowledge of local net-
works, their local synaptic wiring, and the corresponding
dynamical properties. In other words, it is necessary to make
explicit computational models (104) including networks of dif-
ferent types of interneurons as discussed in Chapter 4, as well as
pyramidal neurons and granule cells; the latter should be simu-
lated using compartmental models to account for the spatial
distribution of specific inputs to dendrites and somata.
ALPHA RHYTHMS
This section summarizes the few elements presently known
about the mechanism(s) and origin(s) of alpha rhythm occur-
ring in the frequency range of 8 to 13 Hz, mainly around 10 Hz.
Its initial description (105) also marked the beginning of EEG.
A more recent and exhaustive review of different EEG patterns
of alpha rhythms in the waking adult can be found in
Niedermeyer (106). Besides the classic alpha rhythm of the
visual cortex, there are rhythmic activities in the same fre-
quency range that can be recorded from the somatosensory cor-
tex (called the mu rhythm) and the temporal cortex (called the
tau rhythm). Occipital alpha waves usually occur during
reduced visual attention, while the mu rhythms of the somato-
motor cortex occur as the subject is in a state of muscular relax-
ation. In this way, the presence of dominant activity within the
alpha frequency range has been interpreted, since Adrian and
Matthews (107), as indicating an “idling state” of the brain
although paradoxical findings with alpha enhancement by
attention tasks (108,109) were reported. These observations are
now well understood since a number of observations, both of
visual alpha and somatomotor mu rhythms, have shown that
enhancement and attenuation of these rhythmic activities can
be recorded simultaneously at different sites and successively
depending on a given task. This can be explained taking into
consideration the interaction between different thalamocortical
modules as explained in Chapter 45 (event-related alpha desyn-
chronization [ERD]/event-related alpha synchronization [ERS]
phenomena). In this respect, the experimental findings of Rihs
et al. (110,111) showing enhancement of occipital alpha associ-
ated with active suppression of unattended positions during a
visual spatial orienting task provides further evidence for the
facilitative role of alpha-power decreases (ERD) versus the
inhibitory role of alpha-power enhancement (ERS) of atten-
tional processes. With few exceptions, the study of the cellular
behavior underlying alpha oscillations was prevented mainly by
the fact that alpha rhythms appear only during wakefulness and
no valid anesthesia model was established. The fact that the
alpha band overlaps with another EEG phenomenon, the sleep
spindles (or sigma rhythms, see next section), led to a legitimate
although failed attempt to study it under barbiturate anesthesia
(112). Thus, most of our present knowledge is based on scalp
recordings in humans and laminar profiles of cortical field
potentials in animals.
Chapter 3 ■ Cellular Substrates of Brain Rhythms 45
Alpha Waves: Cortical Dipolar Sources
and Thalamic Influences
In a series of experimental studies performed on dogs, Lopes da
Silva and colleagues established that alpha rhythms are the
undertaking of the cerebral cortex, mostly in the visual areas,
although they can also be recorded in the visual thalamus (lat-
eral geniculate and pulvinar nuclei) (113). In the visual cortex,
alpha waves are generated by an equivalent dipole layer cen-
tered at the level of the somata and basal dendrites of pyrami-
dal neurons in layers IV and V (Fig. 3.9A) (114). Furthermore,
the coherence of alpha waves within the visual cortex was only
partially dependent on thalamic sources measured in the same
animal (Fig. 3.9B) (114,115) leading to the conclusion that hor-
izontal intracortical linkages are essential to the spread of alpha
activities, with only moderate implication of the thalamus.
The cortical generator of alpha rhythms was confirmed and
extended by more recent studies performed in awake monkeys
using fine microelectrode arrays implanted across the visual
cortical areas V2, V4, and the inferior temporal cortex by
Bollimunta et al. (116) (Fig. 3.10). CSD analysis was combined
with CSD–multiple unit activity coherence to identify intracor-
tical alpha current generators. In V2 and V4, alpha current gen-
erators were found in all layers, with the infragranular neurons
(layer V) acting as primary local generator. In contrast, in the
inferior temporal cortex, alpha current generators were found
only in supragranular and infragranular layers, with the supra-
granular generator acting as primary local generator. Granger
causality analysis revealed that the influence of the infragranu-
lar on the supragranular cells has both a direct component and
a component mediated by the granular layer cells. Most inter-
esting, the analysis of the coherence between CSD sinks and
local multiunit activity in infragranular and granular layers
showed that these cells are depolarized and tend to generate
action potentials during the local current sinks; however, at the
supragranular layer the coherence between the CSD sinks and
the multiunit activity was poor, suggesting that the cells of this
layer are inhibited due to inhibitory inputs from layer V
interneurons that innervate superficial pyramidal cells.
Alpha Waves: Synaptic Processes Studied In Vitro
Furthermore the nature of the synaptic processes underlying
alpha frequency oscillations in the cortex was studied in in vitro
preparations, although it is always a matter of discussion
whether in vitro preparations submitted to pharmacologic
manipulations and displaying this kind of oscillations provide
reliable models of the situation in the intact brain, where the
neurons are in different physiologic conditions. Nonetheless,
Flint and Connors (117) were able to obtain oscillations in the
alpha frequency range when slices of rat somatosensory cortex
were activated by low extracellular magnesium in vitro. These
oscillations were NMDA-dependent since they were reversibly
abolished by the NMDA receptor antagonist D-2-amino-
5-phosphonovaleric acid (AP-5) but were relatively unaffected
by the non-NMDA receptor antagonist 6,7-dinitroquinoxaline-
2,3-dione (DNQX). The duration of oscillatory events was
increased by blocking GABAA receptors with bicuculline or by
46 Part I ■ Basic Principles

Figure 3.9 Alpha rhythm in dog. A: (Left) Photomicrograph of a section of the marginal gyrus (visual cortex). The elec-
trode bundle consisted of seven wires with a bare tip of 0.15 mm. (Middle) An epoch of alpha activity recorded simulta-
neously from six sites corresponding to those at left, against a common frontal cranial reference (negativity upward). Note
that there is a polarity change between the most superficial sites (E25, 24) and the deepest sites (E22, 21, 19). (Right) Phase
shift between the most superficial site (E25) and deeper lying sites computed by using spectral analysis based on the aver-
age of a number of epochs within the frequency band 11.2–13.1 Hz. The mean values of the phase shift and the correspon-
ding standard deviations are indicated. (Modified from Lopes da Silva FH, Storm Van Leeuwen W. The cortical source of
the alpha rhythm. Neurosci Lett. 1977;6:237–241.) B: Schematic view of the relationship between thalamic and cortical
alpha activity. The lines indicate the amount of influence that a thalamic signal has on the coherence between a pair of
cortical signals (shaded area) recorded during alpha activity. This was measured by partializing the intercortical coherence
on the thalamic signals from the lateral geniculate nucleus (LGN) and pulvinar (PUL) or other cortical signals. Despite the
fact that PUL has an influence on corticocortical domains of alpha activity, the intercortical factors play a significant role
in the establishment of cortical domains of alpha activity. (Modified from Lopes da Silva FH, Vos JE. Mooibroek J, et al.
Relative contributions of intracortical and thalamocortical processes in the generation of alpha rhythms, revealed by par-
tial coherence analysis. Electroencephalogr Clin Neurophysiol. 1980;50:449–456.)
 Chapter 3 ■ Cellular Substrates of Brain Rhythms 47

Figure 3.10 A: Multielectrode with 14 equally spaced (200 m) contacts. B: A short segment (200 msec) of alpha
rhythm. C: CSD displayed as a color-coded plot, sinks in red and sources in blue. D: Spectra profiles shown for three sites.
E: Coherence between CSD and multiunit activity (MUA) showing high coherence for the granular and infragranular lay-
ers, and in contrast with the superficial layer. (Modified from Bollimunta A, Chen Y, Schroeder CE, et al. Neuronal mech-
anisms of cortical alpha oscillations in awake-behaving macaques. J Neurosci. 2008;28:9976–9988.)(See color insert.)

activating metabotropic glutamate receptors with trans-
1-aminocyclopentane-1,3-dicarboxylic acid (trans-ACPD).
A similar kind of study in slices of the cat lateral geniculate
nucleus maintained in vitro (118) showed that activation of the
metabotropic glutamate receptor, mGluR1a, which is postsy-
naptic to corticothalamic fibers induces synchronized oscilla-
tions at alpha and theta frequencies in a subset of
thalamocortical neurons that can be synchronized by gap junc-
tions. The frequency of the oscillations, whether alpha (8 to 13
Hz) or theta (2 to 7 Hz), depended on the concentration of the
agonist trans-ACPD. These oscillations are driven by a subset
(~25%) of lateral geniculate thalamocortical neurons display-
ing high threshold burst activity occurring at membrane poten-
tials more depolarized than –55 mV, which depends on
Ni2 -sensitive Ca2 channels. In these in vitro conditions the
alpha/theta oscillations do not appear to require chemical
synapses because they can be sustained by coupling through
gap junctions between thalamocortical neurons. These in
vitro results may indicate that activation of mGluRs of corti-
cothalamic synapses is an important mechanism for the gener-
ation of EEG alpha and theta activity but, as stated above, one
should keep in mind that the mechanisms underlying oscilla-
tions in the alpha frequency range elicited in this kind of in
vitro conditions may differ from those that are responsible for
alpha rhythm generation in the intact brain. A comprehensive
review of the basic findings with respect to the generation and
organization of alpha rhythms, both in vivo and in vitro, can be
found in Ref. 119.
Remarkably, interesting additional findings in the last few
years have enriched our understanding of how alpha rhythms
are generated and have posed some new challenges for further
research. Lörincz et al. (120) confirmed that a sparse network of
high threshold bursting thalamocortical cells coupled through
gap junctions is capable of generating alpha rhythmic activity
in cat slices of the lateral geniculate nucleus and ventrobasal
complex (somatosensory thalamus). They also found that this
activity can be induced in this preparation by administering a
cholinergic agonist and can be abolished by muscarinic M3
receptor antagonists. However, this was evident in only about
26% of the thalamocortical cells.
An important issue is that of the remaining majority of thal-
amocortical cells. These authors suggest that the remaining
cells are probably under phasic inhibition from local interneu-
rons activated by the bursting thalamocortical cells. It still
remains unclear how these different neuronal populations
behave in the intact brain where thalamocortical cells receive
48 Part I ■ Basic Principles
several important inputs from the cortex, RE nucleus of the
thalamus, and several brainstem sources that use various neu-
rotransmitters, particularly serotonin, noradrenalin, and hista-
mine that are mainly released during the waking state but not
during REM sleep (121). An important input to these thalamic
nuclei is the cholinergic input that arises from brainstem and
basal forebrain nuclear groups. It has been shown that cholin-
ergic inputs are excitatory to thalamocortical cells causing a
direct depolarization associated with a decrease of a K cur-
rent; at the same time there is a decrease of the inhibitory input
arising from the thalamic RE neurons, since the latter are
hyperpolarized by cholinergic activation of a K conductance,
as was shown in vitro in guinea pig (122) and in vivo in cat
(123,124). In this way these cholinergic inputs favor the block-
age of rhythmic spindles, a phenomenon that is generated in
the thalamus and shares the same frequency range with alpha
oscillations (see below). It is possible that the major effect of
cholinergic inputs on thalamic oscillations in vivo may be to
induce their blockage through muscarinic inhibition of RE
thalamic neurons, although more subtle effects appear to exist
in some thalamic subpopulations, but the relevance of these
effects for alpha oscillatory activity is not yet clear. These recent
findings provide new insights but it is evident that more
research in in vivo conditions is necessary in order to achieve a
deeper understanding of the mechanisms responsible for the
generation of alpha oscillations of different kinds and localization.
SPINDLE (SIGMA) RHYTHMS
Classically, spindles have been regarded as one of the first signs
of EEG synchronization during the early stage of quiescent
sleep. This type of oscillation is defined by the presence of two
distinct rhythms: waxing and waning spindle waves at 7 to
14 Hz within sequences lasting for 1 to 2 seconds, and the peri-
odic recurrence of spindle sequences with a slow rhythm, gen-
erally 0.2 to 0.5 Hz (Fig. 3.11A). Spindles are generated within
the thalamus because they survive in the thalamus after decor-
tications and high brainstem transection (125). More radical
procedures, including complete removal of the striatum and
rhinencephalon, do not affect thalamic spindles either (14).
Sleep spindles have been mostly studied in cats (in vivo) and
ferrets (in vitro). Humans and cats have similar sleep cycles,
EEG patterns, and ultrastructural organization of their thala-
mus. Both intrinsic properties of various thalamic neurons
(126) and network linkages (for review, see Ref. 127) contribute
to the patterning of spindles.
As indicated above, the thalamic circuits comprise two main
structures: (i) the dorsal thalamus, made of several nuclei, each of
them containing both relay (thalamocortical) and local-circuit
(inter-) neurons and (ii) the RE nucleus of the thalamus (see
scheme in Fig. 3.11A). The latter is a thin sheet of GABAergic
neurons that covers the rostral, lateral, and ventral surfaces of the
thalamus (128). It mainly receives inputs from the cerebral cor-
tex and dorsal thalamus, but also from the rostral sector of the
brainstem and basal forebrain. The RE–thalamic–cortical–RE
loop constitutes a resonating circuit that reinforces the spindle
oscillation. Moreover, spindles generated within the RE nucleus
travel along the dorsal thalamic relay pathway en route to the cor-
tex, which probably is responsible for the genesis of the dipole
that allows them to surface in the EEG. Interestingly, no spindles
are recorded in the EEG of kittens during the first 6 to 7 days of
life, although they are already present in thalamic recordings 3 to
4 hours after birth (129), suggesting that no viable dipole is gen-
erated in the thalamus. The brainstem and basal forebrain pro-
jections exert activating effects on the above-mentioned circuit
with the result of inhibiting spindles.
The RE nucleus has been pointed out as the pacemaker for
spindle oscillations because spindles were abolished in the dor-
sal thalamus after disconnection from the RE (130) but were
preserved in the rostral part of the RE nucleus severed from the
dorsal thalamus (131). Further support is provided by the fact
that in cat’s anterior nucleus, which does not receive inputs from
the RE nucleus (132,133), spindles are absent (134). Similarly,
spindles are absent in the cingular cortex (74) and habenular
nuclei, other structures that are devoid of RE inputs. Modeling
studies of isolated RE neurons, with minimal or more realistic
ionic models of RE cells (135–137), further strengthened the
idea that the deafferented RE nucleus is a spindle pacemaker.
Spindle oscillations, in order to produce a coherent pattern
at the level of the target structures, need to be synchronized at
the very site of their genesis. Besides the axonal projections
allowing obvious connectivity between RE neurons, two other
mechanisms have been proposed. The first relies on the dendro-
dendritic GABAergic contacts between RE neurons (138–140)
and the second on the functional interconnection of RE neu-
rons through gap junctions (141).
A fundamental question concerns the triggering factor of a
spindle. Steriade and colleagues (131) have assumed that any
excitatory drive stimulating RE cells would start the process.
The decreased activity of brainstem cholinergic neurons at the
onset of sleep (53) contributes to the overall hyperpolarization
of thalamocortical cells, thus bringing the membrane potential
in the range where bursting discharges can occur. Such clusters
of high-frequency action potentials would excite the dendrites
of RE neurons and would trigger the dendrodendritic ava-
lanche leading to the final synchronization of the whole RE
nucleus. It should be emphasized that thalamic-RE loops may
assist the onset of a spindle; however, its synchronization can-
not rely on the dorsal thalamic circuitry because there is scarce,
if any, communication between thalamocortical neurons and
dorsal thalamic nuclei (142,143).
Once a spindle has been started through one of the mecha-
nisms mentioned above, the pacing of the actual oscillations
within the spindle sequence depends on the RE–thalamic dia-
logue. This was investigated in ferret thalamic visual slices con-
taining geniculate and perigeniculate (RE) nuclei (144–146) and
with multisite recordings within the thalamus (147,148). Results
show that the bursting of RE neurons imposes powerful
GABAergic IPSP in thalamocortical neurons. The end of this inhi-
bition triggers a rebound LTS, crowned by a high-frequency burst
of action potentials, which in turn will again excite the target RE
cells. This is also the moment where intrinsic properties of thala-
mic, both thalamocortical and RE, neurons play an important
role. The amount of hyperpolarization of the thalamocortical cells

determines the degree of deinactivation of the LTS (149). Another
intrinsic property of thalamic cells is the voltage-dependent, non-
inactivating or very slowly inactivating persistent sodium current,
INa(p) (149), which modulates postinhibitory rebound depolariza-
tions. The blockage of the INa(p) by intracellular injection of qua-
ternary derivatives of local anesthetics transformed spindles of
thalamocortical cells into a single, long-lasting period of hyperpo-
larization (because unopposed by INa(p)) without any rhythmic
rebounds within the frequency range of spindles (150). Finally,
the calcium-dependent potassium current, IK(Ca) , or other
voltage-dependent potassium currents (151,152) may prolong the
long-lasting IPSPs generated in thalamocortical neurons by RE
neurons and, thus, assist in the production of the LTS and in
inducing rebound bursts in thalamocortical targets.
Chapter 3 ■ Cellular Substrates of Brain Rhythms 49
Figure 3.11 Spindle oscillations in reticular
thalamic (RE), thalamocortical (Th-Cx, ventro-
lateral nucleus), and cortical (Cx, motor area)
neurons. A: Circuit of three neuronal types and
two rhythms (7–14 Hz and 0.1–0.2 Hz) of spin-
dle oscillations in cortical EEG. B: Intracellular
recordings in cats under barbiturate anesthesia.
(Modified from Steriade M, Deschênes M.
Intrathalamic and brainstem-thalamic networks
involved in resting and alert states. In:
Bentivoglio M, Spreafico R, eds. Cellula r
Thalamic Mechanisms. Amsterdam: Elsevier;
1988:37–62.)
In an intact sleeping brain, the RE–thalamic network under-
goes periodic, synchronized excitatory inputs from the cortex
such as K-complexes (see above), which then shape and modu-
late the duration of spindles (153). Corticothalamic stimuli
produce shorter and waning spindles that lack the initial wax-
ing component, probably because the volleys entrained from
the very beginning large populations of thalamic cells.
Moreover, the synchronization of spindles in the thalamus was
widespread, with virtual zero time lag, in intact preparations,
but dropped drastically after decortications (147,148) (Fig. 3.12).
This result stands in some contrast to a propagating pattern of
spindles in thalamic slices (154) likely due to the absence of
corticothalamic inputs of this preparation and to the overall
diminished synaptic activity of the slice.
Figure 3.12 Control of spatiotemporal coherence of thalamic spindles by the cerebral cortex in cat. Top panels: Disruption of the spatiotemporal coher-
ence of thalamic oscillation after removal of the neocortex. Spatiotemporal maps of electrical activity across the thalamus were constructed by plotting
time (time runs from top to bottom in each column; arrow indicates 1 second), space [from left to right, the width of each column represents 8 mm in
the anteroposterior axis of the thalamus), and local field potential (LFP) voltage (from blue to yellow, color represents the amplitude of the negative
deflection of thalamic LFPs; the color scale ranged in ten steps from the baseline [blue] to 100 V [yellow]). Time was divided into frames, each rep-
resenting a snapshot of 4 msec of thalamic activity. A total of 40 seconds is represented (9880 frames). Each frame consisted of eight color spots, each
corresponding to the LFP of one electrode from anterior to posterior (left to right in each column). Middle panels (experiments): Decay of correlation
with distance. Cross-correlations were computed for all possible pairs of thalamic sites, and the value at time zero from each correlation was repre-
sented as a function of the intersite distance for six different consecutive epochs of 20 seconds. Spatial correlation was calculated for thalamic record-
ings in the intact brain (left) and after removal of cortex (right). Bottom panel (model): Decay of correlation with distance (in units of sites). Similar
computation of cross-correlations as in the above panel from experiments, in the presence of cortex (left) and after decortication (right). (Modified from
Contreras D, Destexhe A, Sejnowski TJ, et al. Control of spatiotemporal coherence of a thalamic oscillation by corticothalamic feedback. Science.
1996;274:771–774. and unpublished data by A Destexhe, D Contreras, TJ Sejnowski and M Steriade.) (See color insert.)
50

The main functional correlate of sleep spindles is the block-
age of incoming stimuli on their way to the cortex. This could
result from the fact that thalamocortical neurons alternate
between IPSPs and LTS bursts. Both events preclude the relay-
ing of afferent synaptic signals either due to the shunting nature
of the former or due to the all-or-none feature of the latter. The
functional blockage of sensory transmission at the thalamic
level promotes the cortical deafferentation, further contribut-
ing to the process of falling asleep.
A final observation concerns the overlapping frequency
range of spindles and alpha waves. It should be emphasized that
both behavioral context and mechanisms are dissimilar. Alpha
waves generally occur during relaxed wakefulness, and in some
studies they are considered to regulate afferent and efferent sig-
nals (155). At difference, spindle oscillations are occurring dur-
ing unconsciousness, thus making the idea of an
alpha-to-spindle continuum obsolete. Nonetheless the same
basic neural circuitries seem to be involved in both kinds of
oscillatory activities—sleep spindles and alpha rhythms—but
working according to different functional modes.
FASTER (BETA AND GAMMA) RHYTHMS
At the opposite side of the EEG spectrum of slow sleep rhythms
( 15 Hz) are fast oscillations (generally 20 to 50 Hz) that are
proper to vigilance states associated with wakefulness, but also
paradoxical (REM) sleep. The brain substrate of EEG activation
upon arousal has begun to be understood since the pioneering
work of Moruzzi and Magoun (156). They reported an “activa-
tion” response to the stimulation of brainstem structures con-
sisting of the suppression of spindles and slower EEG rhythms.
The cellular mechanisms of this suppressing mechanism have
recently been analyzed: the blockage of spindles occurs at the
very site of their genesis (the RE nucleus) through the action of
acetylcholine serotonin and norepinephrine. Although acetyl-
choline hyperpolarizes RE neurons (124,157), while the latter
two depolarize them (158), the combined action of these neu-
rotransmitters is far from being understood. The intrinsically
generated thalamic delta oscillation is blocked through the
depolarizing action of both acetylcholine (10) and
monoamines (159) on thalamocortical cells. Finally, slow corti-
cal delta activities are prevented by cholinergic actions of the
nucleus basalis neurons (30,160).
Beta–Gamma Rhythms: Functional Implications
The disappearance of slow sleep waveforms during the activat-
ing response is also accompanied by the appearance of peculiar
fast rhythms characterizing wakefulness (161). This study
reported, in addition to the ocular syndrome of arousal, a clear-
cut enhancement in amplitude of the spontaneous EEG rhythms
and their regular acceleration to 40 to 45 Hz instead of the till
then observed flattening of the cortical EEG. Since then, several
papers have mentioned the presence of 20 to 40 Hz waves in var-
ious cortical areas, during different conditions of increased
alertness. For instance, fast rhythms were observed in a canine
subject in the occipital cortex while the dog paid intense atten-
tion to a visual stimulus (162), during accurate performance of
Chapter 3 ■ Cellular Substrates of Brain Rhythms 51
a conditioned response to a visual stimulus in monkey (163),
during tasks requiring fine finger movements and focused atten-
tion in monkey motor cortical cells (164), and during behavioral
immobility associated with an enhanced level of vigilance while
a cat was watching a visible but unseizable mouse (165). Other
studies have described stimulus-dependent oscillations at 25 to
45 Hz of the focal EEG and/or neuronal firing in the olfactory
system (166) and visual cortex (167–171).
The underlying mechanism responsible for cortical gamma
oscillations was further unraveled by the finding that synchro-
nized gamma frequency oscillations are enhanced with arousal
and attention (172) or by electrical stimulation of the mesen-
cephalic RE formation (173), and are mediated by acetylcholine
muscarinic receptors. In addition, the spectral power of local field
potentials in the gamma frequency range in the visual cortex of
the cat is closely correlated with hemodynamic signals (174).
The main functional implication of these rhythms in the cor-
tex was proposed to rely on the degree of spatial and temporal
synchronization, thus allowing, at a given moment, the aggrega-
tion of various cortical areas with the purpose of creating global
and coherent properties of patterns, a prerequisite for scene seg-
mentation and figure-ground distinction (175). The issue of this
“feature binding” has, however, been toned down by the fact that
such fast rhythms are coherent also within thalamocortical and
cortical networks during states such as sleep and deep anesthe-
sia reputed for the absence of any conscious behavior
(19,176,177). In these studies it was established that thalamocor-
tical cells do spontaneously oscillate within the beta–gamma fre-
quency range and this occurs coherently with field potentials
recorded from the related cortical areas (Fig. 3.13).
Most likely there exist different kinds of beta–gamma
rhythms with different properties and behavioral/cognitive
associations. Since the report by Gray and Singer (169) that the
firing probability of neurons of the visual cortex of the cat, in
response to the presentation of appropriate visual stimuli, oscil-
lates with a frequency in the gamma frequency range, along
with the observation that no evidence for similar oscillations
were found in the thalamus, these gamma oscillations are con-
sidered to be generated intracortically. This observation led to
the assumption that these gamma oscillations may reflect a gen-
eral mechanism that is capable of binding together activities of
spatially separate cortical areas (179). In general we may state
that synchronization of neuronal networks by way of common
oscillations in the gamma frequency range may enable fast
routing and processing of information in the cortex (180).
Genesis of Beta–Gamma Rhythms
The genesis of fast (beta–gamma) activities lies at the crossroads
of intrinsic and network properties. Cortical neurons generate
upon depolarization intrinsic oscillations in a range around 40
Hz (181,182). This is equally valid for a subclass of rostral
intralaminar thalamic neurons (183) (Fig. 3.14), a property that
is in line with their implication in fast oscillations during acti-
vated states associated with neuronal depolarizations, waking,
and REM sleep. Intralaminar nuclei are particularly fit for
wide range synchronization due to their widespread projections
to the cerebral cortex (125) including the visual areas (184).
52 Part I ■ Basic Principles
Despite these intrinsic properties that make neurons prone
to generate fast oscillations, their imbedding in complex cir-
cuits, especially in the cortex, is required for the short- and
long-range synchronization, in order to make the rhythm
emerge at the EEG level. As examples of different kinds of
circuits that may engage fast rhythmic activities we may men-
tion the following two findings: Freeman (180) postulated that
the generation of 40- to 80-Hz activity in the olfactory bulb
depends on feedback inhibitory circuits involving local-circuit
GABAergic neurons acting on output elements, the mitral cells,
while Chagnac-Amitai and Connors (185) showed that rat
Figure 3.13 Episodes of tonic activation are
associated with coherent fast rhythms (40 Hz)
in cortical EEG and intracellularly recorded
thalamocortical neuron. Cat under ketamine-
xylazine anesthesia. A: Four traces represent
simultaneous recordings of surface and depth
EEG from motor cortical area 4, extracellular
discharges of neuron from the rostrolateral part
of the thalamic reticular (RE) nucleus, and
intracellular activity of thalamocortical neuron
from ventrolateral (VL) nucleus. EEG, RE, and
VL cells displayed a slow oscillation (0.7–
0.8 Hz) during which the sharp depth-negative
(excitatory) EEG waves led to IPSPs in VL cell,
presumably generated by spike bursts in a cor-
tically driven GABAergic RE neuron. Part
marked by horizontal bar, taken from a short-
lasting period of spontaneous EEG activation, is
expanded in B (arrow), with EEG waves and
field potentials from the RE nucleus filtered
between 30 and 50 Hz; part marked by horizon-
tal bar in this panel is further expanded in C to
illustrate relations between action potentials of
VL cell and depth-negative waves in cortical
EEG at a frequency of 40 Hz. D: Cross-
correlations (CROSS) between action potentials
and depth-EEG shows clear-cut relation, with
opposition of phase, between intracellularly
recorded VL neuron and EEG waves. (Modified
from Steriade M, Contreras D, Amzica F, et al.
Synchronization of fast (30–40 Hz) sponta-
neous oscillations in intrathalamic and thalam-
ocortical networks. J Neurosci. 1996;16:
2788–2808.)
somatosensory slices with slightly reduced inhibition display
activities around 37 Hz generated by networks of intrinsically
pyramidal-shaped (excitatory) cells.
The complexity of the cortical generators of fast
(beta–gamma) rhythmic activities is illustrated further by the
observation that, under anesthesia, they do not show field rever-
sal at any depth of the cortex (19). Volume conduction was ruled
out because the negative field potentials of the fast oscillations
were systematically associated, at all depths, with neuronal firing
and were not observable in the underlying white matter. This
fact can be interpreted as being due to the existence of several
 Chapter 3 ■ Cellular Substrates of Brain Rhythms 53

Figure 3.14 Fast oscillatory patterns of neuron recorded from the dorsolateral part of the centrolat-
eral (CL) intralaminar thalamic nucleus, characterized by exceedingly high-frequency (800 to
1000 Hz) spike bursts recurring rhythmically at 50 to 60 Hz. Intracellular recording in cat under bar-
biturate anesthesia. A: Activities triggered by depolarizing current pulses ( 1.2 nA, 50 msec) at dif-
ferent levels of membrane potential (as indicated at left). Two examples are illustrated for each
membrane potential level. At bottom, presence of low-threshold spike (LTS) leading to high-
frequency (800 to 1000 Hz) spike burst at –78 mV and its absence at –84 mV. Note fast-recurring (50
to 60 Hz) spike doublets or spike triplets at relatively depolarized levels (–64 and –58 mV), a feature
that is not seen in other types of thalamocortical neurons. B: Oscillatory patterns similar to those
elicited by current injection occurred spontaneously at similar membrane potentials (from top to bot-
tom, –58, –64, –68, and –78 mV). (Modified from Steriade M, Curró Dossi R, Contreras D.
Electrophysiologic properties of intralaminar thalamocortical cells discharging rhythmic (approxi-
mately 40 Hz) spike-bursts at approximately 1000 Hz during waking and rapid eye movement sleep.
Neuroscience. 1993;56:1–9.)

intracortical distributed microsinks and microsources as shown
by current-source-density analyses (176). Thus, the absence of
potential reversal was due to the lower intensity of vertical cur-
rents, compared with that of transmembrane currents.
Despite their presence during sleep and under anesthesia,
fast (beta–gamma) oscillations are mainly associated with
increased levels of alertness, thus closely following the onset of
activity in cholinergic aggregates of the brainstem and basal
forebrain. The cholinergic activation induces a twofold increase
of cortical EEG waves around 40 Hz, a potentiating effect that
is mediated by muscarinic receptors as it is blocked by scopo-
lamine (30,183). Beyond the depolarizing effect exerted by
54 Part I ■ Basic Principles
acetylcholine on thalamocortical and cortical neurons, this
neurotransmitter hyperpolarizes a subpopulation of cortical
glial cells maintained in culture (186). When tested in vivo,
acetylcholine hyperpolarized most of the glial cells (Fig. 3.15),
in parallel with an overall decrease in the extracellular potas-
sium concentration (187) (Fig. 3.15; see also next section). This
effect was also accompanied by an increased cerebral blood
flow, leading to the interpretation that the glial hyperpolariza-
tion is due to a boosted transport of potassium across capillary
membranes. This is further supported by the fact that EEG acti-
vation, when recorded with DC amplifiers, generally displays a
persistent positive DC shift (187). In addition, brain activation
was accompanied in glial cells by a reduced membrane capaci-
tance suggesting that the interglial syncytium shuts down dur-
ing wakefulness, preventing nonspecific synchronization
through spatial buffering mechanisms (as mentioned previ-
ously). This feature may thus leave the synchronization of
beta–gamma activities strictly confined to synaptic networks,
achieving a more specific aggregation of the neuronal popula-
tions involved.
Recent studies in cortical slices revealed particularly interest-
ing properties with respect to the generation of gamma and
beta rhythmic activities. Roopun et al. (188) showed that
gamma (30 to 70 Hz) rhythms could be generated in the super-
ficial somatosensory cortical layers II/III, while beta rhythms
(20 to 30 Hz) were generated in deep layer V mainly associated
with the activity of fast-spiking interneurons when kainate was
applied to the slices. It is interesting to note that in layer IV both
activities may coexist (Fig. 3.16). A remarkable difference in the
physiologic mechanisms underlying these two types of cortical
oscillations is that the beta rhythms of the deep layers are
reduced by carbenoxolone, which blocks gap junction conduc-
tances, but are not affected by the blockade of synaptic trans-
mission. In contrast, gamma rhythms in superficial layers are
reduced by the blockade of chemical synaptic transmission.
The fact that gamma and beta oscillations rely on different
mechanisms is further emphasized in experiments with trans-
versal cuts within the somatosensory cortical slices that com-
pletely separate layers I to II from layers V to VI. The gamma
rhythmic activity of the superficial layers and the beta rhythmic
activity of the deep layers survived the cut, showing that they
can be maintained separately. The two kinds of rhythmic activ-
ities also respond differently to pharmacologic agents. While
the blockage of GABAA receptors in these slices abolished
gamma rhythms, it enhanced beta rhythms. This means that
gamma oscillations depend on the discharge of pyramidal cells
coupled to GABA-mediated local inhibitory circuits. We should
emphasize, however, that demonstrations of this kind of prop-
erties in vitro are not a guarantee that precisely the same mech-
anisms are also essential in vivo. Nonetheless in vitro studies as
this one yield a proof-of-principle that such mechanisms can be
operational in the cortex.
It was reported in the rat in vivo that benzodiazepines, which
enhance GABAA receptor-mediated synaptic transmission, also
increased beta oscillations in the EEG. In this sense, Mandema
and Danhof (189) reviewed data showing that changes in the
EEG spectral amplitude within the beta frequency band provide
a remarkable quantitative indicator of the pharmacologic effect
of benzodiazepines. In addition to the classic beta and gamma
oscillations, episodes of oscillations at high frequency (100 to
600 Hz) and short duration (~50 to 100 ms) have been
described in the brain of rodents and human, both under nor-
mal and epileptic conditions. These oscillatory bursts are called
ripples (see Fig. 3.8) or fast ripples depending on frequency and
are discussed in Chapter 37.
Fast (Beta–Gamma) Rhythms in Human EEG
While the mechanisms of generation of gamma and beta
rhythms in cortical slices, in vitro, present clear different prop-
erties, the situation in vivo in humans is even more complex.
On the one hand, beta EEG or MEG activities recorded over the
Rolandic cortical area (about 15 and 30 Hz) display a close tem-
poral relationship with peripheral EMG activity during isomet-
ric contractions, as shown using MEG recordings (190–192).
The MEG signals in the beta frequency range lead the EMG sig-
nals in time, and the time lag increases with increasing
brain–muscle distance. This implies that this beta rhythmic
activity is associated with rhythmic firing of neurons of the
motor cortex that generate the commands to drive spinal
motoneurons (193) (see also Chapter 42). On the other hand,
however, there are other conditions where beta rhythms are not
associated with a state of active neuronal firing in the sense
described above. This is the case of an increase of beta activity
after a finger movement (postmovement beta rebound at 16 to
21 Hz) or foot movement (19 to 26 Hz) (194) when the mus-
cles relax. The combination of EEG recording with the applica-
tion of transcranial magnetic stimulation showed that during
this beta rebound the corticospinal excitability was diminished
(195). The dynamics of changes of beta rhythms under differ-
ent behavioral conditions is further detailed in Chapter 45.
Related to the beta rebound after a movement it was shown that
a surge of gamma activity (above 30 Hz) may immediately pre-
cede a finger movement (196), which is probably directly asso-
ciated with the activation of motor cortical neurons.
DC EEG POTENTIALS
Although the chapter is devoted to oscillations in the brain, this
section will briefly raise the issue of less conventional, very slow
EEG signals, which are overwhelmingly left out of the clinical
focus (see, however, Refs. 197 and 198). Interestingly, the first
EEG machines used by Berger for his first EEG recordings were
true DC amplifiers. This changed with the introduction of
increasingly performing filters offering the possibility to limit
the intrusion of artifacts in clinical recordings. Filters also pre-
cluded the expression of neurophysiologic relevant informa-
tion, which, in time, limited the knowledge of brain
mechanisms. The consecutive conventional theories in elec-
troencephalography considered that most of the electrical
activity recorded on the scalp, including very slow waves, is gen-
erated by neuronal networks with particular emphasis on corti-
cal dipoles. Several studies have also demonstrated the
involvement of glial cells in generating slow local field poten-
tials during spreading depression (199,200), seizures
 Chapter 3 ■ Cellular Substrates of Brain Rhythms 55

Figure 3.15 Comparison between electrically elicited and spontaneously occurring activation in a pair of simultane-
ously impaled neuron and glia in a cat under ketamine-xylazine anesthesia. A: Sequence of slow ( 1 Hz) oscillation,
brainstem cholinergic activation, and return to slow oscillation. Depolarizing phases of the slow oscillation are indi-
cated with a black arrow head (up state), while hyperpolarizing phases are marked with an empty arrow head (down
state). After activation, the sustained neuronal depolarization is accompanied by glial hyperpolarization and the DC
field potential assumes rather positive values. B: Histograms of neuronal (panel 1) and glial (panel 2) Vm during the
slow oscillation (in gray) and during activation (black trace). Histograms were calculated over the respective under-
lined epochs in A. They illustrate the incidence of different values of Vm , sampled at 20 kHz, over periods indicated
by bar below traces. Note the bimodal histograms during sleep (up states correspond to the black arrow head, down
states to the empty arrow head) and the unimodal histogram after activation, as well as the opposite evolution of glial
and neuronal Vm (glia hyperpolarizes and neuron depolarizes). C1: Similar pattern of activity in the same double intra-
cellular recording during a period of spontaneous activation of the EEG. Periods within squares are expanded at right
(panels 2 and 3). D: Histograms from equivalent time periods before and after activation. (Modified from Seigneur J,
Kroeger D, Nita DA, et al. Cholinergic action on cortical glial cells in vivo. Cereb Cortex. 2006;16:655–668.)
56 Part I ■ Basic Principles

Figure 3.16 2 rhythms are generated in deep layers of cortex. A: Spectrograms of field rhythms generated
in somatosensory cortex slices by 400 nM kainate. Superficial layers generated gamma frequency (30 to
50 Hz) signals (layer II/ III). Deep layers concurrently generated 2 frequency signals (layer V, 20 to 30 Hz).
Layer IV recordings show both frequency bands coexisting. Below each spectrogram are representative field
potential traces (scale bars: 0.2 mV, 100 ms). B: Surgical separation of deep from superficial layers at the layer
IV/ V border abolished neither rhythm. Cartoon illustrates the electrode position for the power spectra taken
from 60-second epochs of field potential data in the superficial layers (a–c, black lines) and deep layers
(d–e, red lines). (Modified from Roopun AK, Middleton SJ, Cunningham MO, et al. A 2-frequency (20–30 Hz)
oscillation in nonsynaptic networks of somatosensory cortex. Proc Natl Acad Sci USA. 2006;103:
15646–15650.)(See color insert.)

(199–203), and sleep (36). Besides these concepts, pioneering
studies in the early 1950s to 1970s have proposed that some of
the slow potential shifts are generated at the interface between
cerebrospinal fluid and blood as a function of the partial pres-
sure of CO2 (PCO2) (204–208). Such potentials were suggested
to originate across the blood–brain barrier (209–212). The
merit, but also the difficulty, of these studies consisted in intro-
ducing among the variables that condition brain phenomena
extraneuronal parameters.
The basic aspects related to EEG potentials situated at the
infraslow limit of the spectrum (DC to 0.01 Hz, although the
upper limit is only informative) rely on the electrical scheme
proposed by Voipio and colleagues (211) (Fig. 3.17A). It proved
to be particularly useful to understand how large DC shifts that
can be recorded on the scalp in response to hypo/hypercapnia
(Fig. 3.17C,D) are not generated by networks of neurons and/or
glia. This is in stark contrast with the prevailing view about the
mechanisms generating ventilation-related DC shifts that
emphasized an almost exclusive role of cortical neurons and their
dendritic tree (201,213). Recent data (212) demonstrate that such
ventilation-induced DC shifts occurred in the absence of any
parallel change in the neuronal or glial membrane potential. No
extracellular, spatial polarity reversal of the DC shift responses
was seen across or within the neocortex and the underlying white
matter. This finding provides further support for the view that
the DC shifts are not caused by cortical current dipoles generated
in response to neuronal and/or glial activity. And finally, the
breakdown of the blood–brain barrier produced itself a DC shift
as mark of the polarizing potential of the barrier. Once the
blood–brain barrier was disrupted, ventilation-induced DC
shifts were abolished (Fig. 3.18). These examples clearly state
that, other than neurons and glia, a series of other parenchymal
elements (blood flow and capillary epithelial cells) participate in
the genesis of potentials that translate into EEG signals.
 Chapter 3 ■ Cellular Substrates of Brain Rhythms 57

Figure 3.17 DC shift potentials generated in the EEG by variations of the systemic CO2. A left: Schematic
drawing of the human head divided into four compartments: brain, blood, the blood–brain or blood–CSF bar-
rier (black double line), and all other tissues. EBB, the electromotive force of the voltage source across the
brain–blood interface; RBB, its internal resistance. This voltage source generates a volume current (lines with
arrowheads) that flows first through RB (the distributed resistance that couples brain potential to the
surrounding extracortical tissue layers) and RS (the distributed resistance of the layers between brain surface
and skin surface pooled together) and gives rise to the voltage drop VDC that can be measured on scalp.
Current returns back to the brain–blood interface through RT1 (resistance of wider tissue pathways below the
level of cranial fossae), RT2 (access resistance to blood), and RT3 (resistance of blood). A right: Simplified
equivalent circuit of the scheme depicted at left. IBB denotes the current that is driven in the circuit by the
brain–blood potential difference (VBB). (Modified from Voipio J, Tallgren P, Heinonen E, et al. Millivolt-scale
DC shifts in the human scalp EEG: evidence for a nonneuronal generator. J Neurophysiol.
2003;89:2208–2214.) B: Scheme with the position of the scalp electrodes with respect to the brain of the cat.
DC shifts induced by hyperventilation (C) and hypoventilation (D) in a cat under ketamine-xylazine anesthe-
sia. Below the EEG signals, variations of the CO2 concentration in the expired air. Hyperventilation induced
positive DC shifts, while hypoventilation was associated with negative DC shifts. (Modified from Nita DA,
Vanhatalo S, Lafortune FD, et al. Nonneuronal origin of CO2-related DC EEG shifts: an in vivo study in the
cat. J Neurophysiol. 2004;92:1011–1022.)
58 Part I ■ Basic Principles

Figure 3.18 DC shifts induced by the disruption of the blood–brain barrier. Scalp DC EEG recording in a cat
under ketamine-xylazine anesthesia. After a control period, a volume of 3 ml saline (SAL) was slowly injected
into the right carotid artery with no effect. Then an identical volume of sodium dehydrocholate (DHC; 17.5% )
was injected in order to break the blood–brain barrier. This induced a clear positive DC shift, with right lat-
eralization. During the period with open blood–brain barrier, the amplitude of the responses to hypo/ hyper-
ventilation (hv/ HV) maneuvers was drastically diminished. The asterisk during the first hypoventilation
maneuver indicates that initially for a period of about 1 minute the respirator was stopped. (Modified from
Nita DA, Vanhatalo S, Lafortune FD, et al. Nonneuronal origin of CO2-related DC EEG shifts: an in vivo study
in the cat. J Neurophysiol. 2004;92:1011–1022.)

REFERENCES 10. Steriade M, Curró Dossi R, Nuñez A. Network modulation of a

1. Walter G. The location of cerebral tumors by electro-
encephalography. Lancet. 1936;8:305–308.
2. IFSECN. A glossary of terms most commonly used by clinical elec-
troencephalographers. Electroencephalogr Clin Neurophysiol.
1974;37:538–548.
3. Creutzfeldt OD, Watanabe S, Lux HD. Relations between EEG phe-
nomena and potentials of single cortical cells. II. Spontaneous and
convulsoid activity. Electroencephalogr. Clin Neurophysiol.
1966;20:19–37.
4. Kellaway P, Gol A, Proler M. Electrical activity of the isolated cerebral
hemisphere and isolated thalamus. Exp Neurol. 1966;14:281–304.
5. Rappelsberger P, Pockberger H, Petsche H. The contribution of the
cortical layers to the generation of the EEG: field potential and
current source density analyses in the rabbit’s visual cortex.
Electroencephalogr. Clin Neurophysiol. 1982;53:254–269.
6. Leresche N, Jassik-Gerschenfeld D, Haby M, et al. Pacemaker-like
and other types of spontaneous membrane potential oscillations
of thalamocortical cells. Neurosci. Lett. 1990;113:72–77.
7. McCormick DA, Pape HC. Properties of a hyperpolarization-
activated cation current and its role in rhythmic oscillation in
thalamic relay neurones. J Physiol. 1990;431:291–318.
8. Leresche N, Lightowler S, Soltesz I, et al. Low-frequency oscillatory
activities intrinsic to rat and cat thalamocortical cells. J Physiol.
1991;441:155–174.
9. Soltesz I, Lightowler S, Leresche N, et al. Two inward currents and
the transformation of low-frequency oscillations of rat and cat
thalamocortical cells. J Physiol. 1991;441:175–197.
slow intrinsic oscillation of cat thalamocortical neurons impli-
cated in sleep delta waves: cortically induced synchronization
and brainstem cholinergic suppression. J Neurosci. 1991;
11:3200–3217.
11. Curró Dossi R, Nuñez A, Steriade M. Electrophysiology of a slow
(0.5–4 Hz) intrinsic oscillation of cat thalamocortical neurones in
vivo. J Physiol. 1992;447:215–234.
12. Nuñez A, Curró Dossi R, Contreras D, et al. Intracellular evidence
for incompatibility between spindle and delta oscillations in thal-
amocortical neurons of cat. Neuroscience. 1992;48:75–85.
13. Nita DA, Steriade M, Amzica F. Hyperpolarisation rectification in
cat lateral geniculate neurons modulated by intact corticothalamic
projections. J Physiol. 2003;552:325–332.
14. Villablanca J. Role of the thalamus in sleep control: Sleep-
wakefulness studies in chronic diencephalic and athalamic cats. In:
Petre-Quadens O, Schlag J, eds. Basic Sleep Mechanisms. New York:
Academic; 1974:51–81.
15. Ball GJ, Gloor P, Schaul N. The cortical electromicrophysiology of
pathological delta waves in the electroencephalogram of cats.
Electroencephalogr. Clin Neurophysiol. 1977;43:346–361.
16. Steriade M, Buzsáki G. Parallel activation of thalamic and cortical
neurons by brainstem and basal forebrain cholinergic systems. In:
Steriade M, Biesold D, eds. Brain Cholinergic Systems. New York:
Oxford University Press; 1990:3–62.
17. Steriade M, Gloor P, Llinás RR, et al. Report of IFCN committee
on basic mechanisms. Basic mechanisms of cerebral rhythmic
activities. Electroencephalogr Clin Neurophysiol. 1990;76:
481–508.

18. Steriade M, Nuñez A, Amzica F. A novel slow (< 1 Hz) oscillation
of neocortical neurons in vivo: depolarizing and hyperpolarizing
components. J Neurosci. 1993;13:3252–3265.
19. Steriade M, Amzica F, Contreras D. Synchronization of fast (30–40
Hz) spontaneous cortical rhythms during brain activation. J
Neurosci. 1996;16:392–417.
20. Achermann P, Borbély AA. Low-frequency (<1 Hz) oscillations in the
human sleep electroencephalogram. Neuroscience. 1997;81:213–222.
21. Amzica F, Steriade M. The K-complex: its slow (<1-Hz) rhythmic-
ity and relation to delta waves. Neurology. 1997;49:952–959.
22. Simon NR, Manshanden I, Lopes da Silva FH. A MEG study of
sleep. Brain Res. 2000;860:64–76.
23. Mölle M, Marshall L, Gais S, et al. Grouping of spindle activity
during slow oscillations in human non-rapid eye movement sleep.
J Neurosci. 2002;22:10941–10947.
24. Massimini M, Rosanova M, Mariotti M. EEG slow (approximately
1 Hz) waves are associated with nonstationarity of thalamo-
cortical sensory processing in the sleeping human. J Neurophysiol.
2003;89:1205–1213.
25. Steriade M, Nuñez A, Amzica F. Intracellular analysis of relations
between the slow (< 1 Hz) neocortical oscillation and other
sleep rhythms of the electroencephalogram. J Neurosci.
1993;13:3266–3283.
26. Timofeev I, Steriade M. Low-frequency rhythms in the thalamus
of intact-cortex and decorticated cats. J Neurophysiol. 1996;76:
4152–4168.
27. Sanchez-Vives MV, McCormick DA. Cellular and network mecha-
nisms of rhythmic recurrent activity in neocortex. Nat Neurosci.
2000;3:1027–1034.
28. Contreras D, Timofeev I, Steriade M. Mechanisms of long-lasting
hyperpolarizations underlying slow sleep oscillations in cat corti-
cothalamic networks. J Physiol. 1996;494(Pt 1):251–264.
29. Massimini M, Amzica F. Extracellular calcium fluctuations and
intracellular potentials in the cortex during the slow sleep oscilla-
tion. J Neurophysiol. 2001;85:1346–1350.
30. Steriade M, Amzica F, Nuñez A. Cholinergic and noradrenergic
modulation of the slow (approximately 0.3 Hz) oscillation in neo-
cortical cells. J Neurophysiol. 1993;70:1385-1400.
31. Contreras D, Steriade M. Cellular basis of EEG slow rhythms: a
study of dynamic corticothalamic relationships. J Neurosci.
1995;15:604–622.
32. Amzica F, Steriade M. Disconnection of intracortical synaptic
linkages disrupts synchronization of a slow oscillation. J Neurosci.
1995;15:4658–4677.
33. Amzica F, Steriade M. Short- and long-range neuronal synchro-
nization of the slow (< 1 Hz) cortical oscillation. J Neurophysiol.
1995;73:20–38.
34. Steriade M, Contreras D, Curró Dossi R, et al. The slow (< 1 Hz)
oscillation in reticular thalamic and thalamocortical neurons: sce-
nario of sleep rhythm generation in interacting thalamic and neo-
cortical networks. J Neurosci. 1993;13:3284–3299.
35. Amzica F, Steriade M. Electrophysiological correlates of sleep delta
waves. Electroencephalogr Clin Neurophysiol. 1998;107:69–83.
36. Amzica F, Neckelmann D. Membrane capacitance of cortical neu-
rons and glia during sleep oscillations and spike-wave seizures. J
Neurophysiol. 1999;82:2731–2746.
37. Amzica F, Massimini M. Glial and neuronal interactions during
slow wave and paroxysmal activities in the neocortex. Cereb
Cortex. 2002;12:1101–1113.
38. Amzica F, Massimini M, Manfridi A. Spatial buffering during slow
and paroxysmal sleep oscillations in cortical networks of glial cells
in vivo. J Neurosci. 2002;22:1042–1053.
Chapter 3 ■ Cellular Substrates of Brain Rhythms 59
39. Bowman CL, Kimelberg HK. Excitatory amino acids directly
depolarize rat brain astrocytes in primary culture. Nature.
1984;311:656–659.
40. Kettenmann H, Schachner M. Pharmacological properties of
gamma-aminobutyric acid-, glutamate-, and aspartate-induced
depolarizations in cultured astrocytes. J Neurosci. 1985;5:
3295–3301.
41. Orkand RK, Nicholls JG, Kuffler SW. Effect of nerve impulses on
the membrane potential of glial cells in the central nervous system
of amphibia. J Neurophysiol. 1966;29:788–806.
42. Kettenmann H, Ransom BR. Electrical coupling between astro-
cytes and between oligodendrocytes studied in mammalian cell
cultures. Glia. 1988;1:64–73.
43. Walz W. Role of glial cells in the regulation of the brain ion
microenvironment. Prog Neurobiol. 1989;33:309–333.
44. Nicholson C. Extracellular space as the pathway for neuron-glial
cell interaction. In: Kettenmann H, Ransom BR, eds. Neuroglia.
New York: Oxford University Press; 1995;387–397.
45. Ballanyi K, Grafe P, ten Bruggencate G. Ion activities and potas-
sium uptake mechanisms of glial cells in guinea-pig olfactory cor-
tex slices. J Physiol. 1987;382:159–174.
46. Kettenmann H. K+ and Cl- uptake by cultured oligodendrocytes.
Can J Physiol Pharmacol. 1987;65:1033–1037.
47. Kocsis JD, Malenka RC, Waxman SG. Effects of extracellular potas-
sium concentration on the excitability of the parallel fibres of the
rat cerebellum. J Physiol. 1983;334:225–244.
48. Rausche G, Igelmund P, Heinemann U. Effects of changes in extra-
cellular potassium, magnesium and calcium concentration on
synaptic transmission in area CA1 and the dentate gyrus of rat
hippocampal slices. Pflugers Arch. 1990;415:588–593.
49. Hille B. Ion Channels of Excitable Membranes. Sunderland:
Sinauer; 2001.
50. Loomis AL, Harvey EN, Hobart G. Distribution of disturbance
patterns in the human electroencephalogram, with special refer-
ence to sleep. J Neurophysiol. 1938;1:413–440.
51. Colrain IM. The K-complex: a 7-decade history. Sleep.
2005;28:255–273.
52. Amzica F, Steriade M. Cellular substrates and laminar profile of
sleep K-complex. Neuroscience. 1998;82:671–686.
53. Steriade M, Oakson G, Ropert N. Firing rates and patterns of mid-
brain reticular neurons during steady and transitional states of the
sleep-waking cycle. Exp Brain Res. 1982;46:37–51.
54. Steriade M, Datta S, Paré D, et al. Neuronal activities in brain-stem
cholinergic nuclei related to tonic activation processes in thalam-
ocortical systems. J Neurosci. 1990;10:2541–2559.
55. Hirsch JC, Fourment A, Marc ME. Sleep-related variations of
membrane potential in the lateral geniculate body relay neurons of
the cat. Brain Res. 1983;259:308–312.
56. Davis H, Davis PA, Loomis AL, et al. Electrical reactions of the
human brain to auditory stimulation during sleep. J Neurophysiol.
1939;6:500–514.
57. Steriade M, Amzica F. Coalescence of sleep rhythms and their
chronology in corticothalamic networks. Sleep Res Online.
1998;1:1–10.
58. Ingvar DH, Sjölund B, Ardo A. Correlation between dominant
EEG frequency, cerebral oxygen uptake and blood flow.
Electroencephalogr Clin Neurophysiol. 1976;41:268–276.
59. Saunders MG, Westmoreland BF. The EEG in evaluation of disor-
ders affecting the brain diffusely. In: Current Practice of Clinical
Electroencephalography, Klass DW, Daly DD, eds. New York: Raven
Press; 1979;343–379.
60. Buzsáki G. Rhythms of the Brain. Oxford University Press; 2006.
60 Part I ■ Basic Principles
61. Brazier MAB. Studies of the EEG activity of limbic structures in
man. Electroencephalogr. Clin Neurophysiol. 1968;25:309–318.
62. Halgren E, Babb TL, Crandall PH. Human hippocampal forma-
tion EEG desynchronizes during attentiveness and movement.
Electroencephalogr Clin Neurophysiol. 1978;44:778–781.
63. Arnolds DE, Lopes da Silva FH, Aitink JW, et al. The spectral prop-
erties of hippocampal EEG related to behaviour in man.
Electroencephalogr. Clin Neurophysiol. 1980;50:324–328.
64. Kahana MJ, Sekuler R, Caplan JB, et al. Human theta oscillations
exhibit task dependence during virtual maze navigation. Nature.
1999;399:781–784.
65. Ekstrom AD, Caplan JB, Ho E, et al. Human hippocampal theta
activity during virtual navigation. Hippocampus. 2005;15:
881–889.
66. Bódizs R, Kántor S, Szabó G, et al. Rhythmic hippocampal slow
oscillation characterizes REM sleep in humans. Hippocampus.
2001;11:747–753.
67. Tesche CD, Karhu J. Theta oscillations index human hippocampal
activation during a working memory task. Proc Natl Acad Sci USA.
2000;97:919–924.
68. Cornwell BR, Johnson LL, Holroyd T, et al. Human hippocampal
and parahippocampal theta during goal-directed spatial naviga-
tion predicts performance on a virtual Morris water maze. J
Neurosci. 2008;28:5983–5990.
69. Stewart M, Fox SE. Hippocampal theta activity in monkeys. Brain
Res. 1991;538:59–63.
70. Green JD, Arduini AA. Hippocampal electrical activity in arousal.
J Neurophysiol. 1954;17:533–557.
71. Buzsáki G. The hippocampo-neocortical dialogue. Cereb Cortex.
1996;6:81–92.
72. Witter MP, Naber PA, van HT, et al. Cortico-hippocampal com-
munication by way of parallel parahippocampal-subicular path-
ways. Hippocampus. 2000;10:398–410.
73. Buzsáki G. Theta oscillations in the hippocampus. Neuron.
2002;33:325–340.
74. Leung LW, Borst JG. Electrical activity of the cingulate cortex. I.
Generating mechanisms and relations to behavior. Brain Res.
1987;407:68–80.
75. Chapman CA, Lacaille JC. Intrinsic theta-frequency membrane
potential oscillations in hippocampal CA1 interneurons of stra-
tum lacunosum-moleculare. J Neurophysiol. 1999;81:1296–1307.
76. Bilkey DK, Heinemann U. Intrinsic theta-frequency membrane
potential oscillations in layer III/V perirhinal cortex neurons of
the rat. Hippocampus. 1999;9:510–518.
77. Petsche H, Stumpf C, Gogolak G. The significance of the rabbit’s
septum as a relay station between the midbrain and the hippocam-
pus. The control of hippocampus arousal activity by septum cells.
Electroencephalogr. Clin Neurophysiol. 1962;14:202–211.
78. Petsche H, Gogolak G, van Zwieten PA. Rhythmicity of septal cell dis-
charges at various levels of reticular excitation. Electroencephalogr.
Clin Neurophysiol. 1965;19:25–33.
79. Freund TF, Antal M. GABA-containing neurons in the septum
control inhibitory interneurons in the hippocampus. Nature.
1988;336:170–173.
80. Vertes RP, Kocsis B. Brainstem-diencephalo-septohippocampal
systems controlling the theta rhythm of the hippocampus.
Neuroscience. 1997;81:893–926.
81. Buzsáki G, Leung LW, Vanderwolf CH. Cellular bases of hip-
pocampal EEG in the behaving rat. Brain Res. 1983;287:
139–171.
82. Mitchell SJ, Ranck JB Jr. Generation of theta rhythm in medial
entorhinal cortex of freely moving rats. Brain Res. 1980;189:49–66.
83. Alonso A, Garcia-Austt E. Neuronal sources of theta rhythm in
the entorhinal cortex of the rat. I. Laminar distribution of theta
field potentials. Exp Brain Res. 1987;67:493–501.
84. Alonso A, Garcia-Austt E. Neuronal sources of theta rhythm in
the entorhinal cortex of the rat. II. Phase relations between unit
discharges and theta field potentials. Exp Brain Res.
1987;67:502–509.
85. Boeijinga PH, Lopes da Silva FH. Differential distribution of beta
and theta EEG activity in the entorhinal cortex of the cat. Brain
Res. 1988;448:272–286.
86. Vanderwolf CH, Buzsáki G, Cain DP, et al. Neocortical and hip-
pocampal electrical activity following decapitation in the rat.
Brain Res. 1988;451:340–344.
87. Vanderwolf CH, Leung LS. Hippocampal rhythmical slow activ-
ity: a brief history and effects entorhinal lesions and phencyci-
dine. In: Seifert W, ed. The Neurobiology of the Hippocampus.
London: Academic Press; 1983:275–302.
88. Montgomery SM, Betancur MI, Buzsáki G. Behavior-dependent
coordination of multiple theta dipoles in the hippocampus. J
Neurosci. 2009;29:1381–1394.
89. Stewart M, Fox SE. Do septal neurons pace the hippocampal
theta rhythm? Trends Neurosci. 1990;13:163–168.
90. Leung LW. Model of gradual phase shift of theta rhythm in the
rat. J Neurophysiol. 1984;52:1051–1065.
91. Chapman CA, Lacaille JC. Cholinergic induction of theta-fre-
quency oscillations in hippocampal inhibitory interneurons
and pacing of pyramidal cell firing. J Neurosci. 1999;19:
8637–8645.
92. Leung LS. Generation of theta and gamma rhythms in the hip-
pocampus. Neurosci. Biobehav Rev. 1998;22:275–290.
93. Buzsáki G, Czopf J, Kondákor I, et al. Laminar distribution of
hippocampal rhythmic slow activity (RSA) in the behaving rat:
current-source density analysis, effects of urethane and atropine.
Brain Res. 1986;365:125–137.
94. Soltesz I, Deschênes M. Low- and high-frequency membrane
potential oscillations during theta activity in CA1 and CA3
pyramidal neurons of the rat hippocampus under ketamine-
xylazine anesthesia. J Neurophysiol. 1993;70:97–116.
95. Somogyi P, Klausberger T. Defined types of cortical interneurone
structure space and spike timing in the hippocampus. J Physiol.
2005;562:9–26.
96. Banks MI, White JA, Pearce RA. Interactions between
distinct GABA(A) circuits in hippocampus. Neuron. 2000;25:
449–457.
97. Wendling F, Bartolomei F, Bellanger JJ, et al. Epileptic fast activ-
ity can be explained by a model of impaired GABAergic dendritic
inhibition. Eur J Neurosci. 2002;15:1499–1508.
98. Fuentealba P, Tomioka R, Dalezios Y, et al. Rhythmically active
enkephalin-expressing GABAergic cells in the CA1 area of the
hippocampus project to the subiculum and preferentially inner-
vate interneurons. J Neurosci. 2008;28:10017–10022.
99. Kamondi A, Acsady L, Wang XJ, et al. Theta oscillations in
somata and dendrites of hippocampal pyramidal cells in vivo:
activity-dependent phase-precession of action potentials.
Hippocampus. 1998;8:244–261.
100. Wulff P, Ponomarenko AA, Bartos M, et al. Hippocampal theta
rhythm and its coupling with gamma oscillations require fast
inhibition onto parvalbumin-positive interneurons. Proc Natl
Acad Sci USA. 2009;106:3561–3566.
101. Klausberger T, Somogyi P. Neuronal diversity and temporal
dynamics: the unity of hippocampal circuit operations. Science.
2008;321:53–57.

102. Sik A, Penttonen M, Ylinen A, et al. Hippocampal CA1 interneu-
rons: an in vivo intracellular labeling study. J Neurosci.
1995;15:6651–6665.
103. Toth K, Freund TF, Miles R. Disinhibition of rat hippocampal
pyramidal cells by GABAergic afferents from the septum. J
Physiol. 1997500(Pt 2):463–474.
104. Rotstein HG, Pervouchine DD, Acker CD, et al. Slow and fast
inhibition and an H-current interact to create a theta rhythm in
a model of CA1 interneuron network. J Neurophysiol.
2005;94:1509–1518.
105. Berger H. Über das Elektroenkephalogramm des Menschen. 1st
report. Arch Psychiat Nervenkr. 1929;87:527–570.
106. Niedermeyer E. The normal EEG of the waking adult. In:
Niedermeyer E,Lopes da Silva FH, eds. Electroencephalography:
Basic Principles, Clinical Applications, and Related Fields.
Baltimore: Lippincott, Williams & Wilkins; 2005:167–192.
107. Adrian ED, Mathews BHC. The interpretation of potential waves
in the cortex. J Physiol. 1934;81:440–471.
108. Creutzfeld O, Grünvald G, Simonova O, et al. Changes of the
basic rhythms of the EEG during the performance of mental and
visuomotor tasks. In: Evans CR, Mulholland TB, eds. Attention in
Neurophysiology. London: Butterworth; 1969;148–168.
109. Ray WJ, Cole HW. EEG alpha activity reflects attentional
demands, and beta activity reflects emotional and cognitive
processes. Science. 1985;228:750–752.
110. Rihs TA, Michel CM, Thut G. Mechanisms of selective inhibition
in visual spatial attention are indexed by alpha-band EEG syn-
chronization. Eur J Neurosci. 2007;25:603–610.
111. Rihs TA, Michel CM, Thut G. A bias for posterior alpha-band
power suppression versus enhancement during shifting versus
maintenance of spatial attention. Neuroimage. 2009;44:190–199.
112. Andersen P, Andersson SA. Physiological Basis of the Alpha
Rhythm. New York: Appleton-Century-Crofts; 1968.
113. Lopes da Silva FH, Van Lierop THMT, Schrijer CFM, et al.
Organization of thalamic and cortical alpha rhythm: spectra and
coherences. Electroencephalogr Clin Neurophysiol.
1973;35:627–639.
114. Lopes da Silva FH, Storm Van Leeuwen W. The cortical alpha
rhythm in dog: depth and surface profile of phase. In: Brazier
MAB, Petsche H, eds. Architecture of the Cerebral Cortex, IBRO
monograph series, Vol. 3. New York: Raven Press; 1978;319–333.
115. Lopes da Silva FH, Vos JE. Mooibroek J, et al. Relative contribu-
tions of intracortical and thalamo-cortical processes in the gen-
eration of alpha rhythms, revealed by partial coherence analysis.
Electroencephalogr Clin Neurophysiol. 1980;50:449–456.
116. Bollimunta A, Chen Y, Schroeder CE, et al. Neuronal mechanisms
of cortical alpha oscillations in awake-behaving macaques. J
Neurosci. 2008;28:9976–9988.
117. Flint AC, Connors BW. Two types of network oscillations in neo-
cortex mediated by distinct glutamate receptor subtypes and
neuronal populations. J Neurophysiol. 199675:951–957.
118. Hughes SW, Lörincz M, Cope DW, et al. Synchronized oscilla-
tions at alpha and theta frequencies in the lateral geniculate
nucleus. Neuron. 2004;42:253–268.
119. Hughes SW, Crunelli V. Thalamic mechanisms of EEG alpha
rhythms and their pathological implications. Neuroscientist.
2005;11:357–372.
120. Lörincz ML, Crunelli V, Hughes SW. Cellular dynamics of
cholinergically induced alpha (8-13 Hz) rhythms in sensory thal-
amic nuclei in vitro. J Neurosci. 2008;28:660–671.
121. Steriade M. Neuronal Substrates of Sleep and Epilepsy. Cambridge:
University Press; 2003.
Chapter 3 ■ Cellular Substrates of Brain Rhythms 61
122. McCormick DA, Prince DA. Actions of acetylcholine in the
guinea-pig and cat medial and lateral geniculate nuclei, in vitro.
J Physiol. 1987;392:147–165.
123. Hu B, Steriade M, Deschênes M. The effects of brainstem peri-
brachial stimulation on neurons of the lateral geniculate nucleus.
Neuroscience. 1989;31:13–24.
124. Hu B, Steriade M, Deschênes M. The effects of brainstem peri-
brachial stimulation on perigeniculate neurons: the blockage of
spindle waves. Neuroscience. 1989;31:1–12.
125. Morison RS, Bassett DL. Electrical activity of the thalamus and
basal ganglia in decorticate cats. J Neurophysiol. 1945;8:09–314.
126. Steriade M, Llinás RR. The functional states of the thalamus and
the associated neuronal interplay. Physiol Rev. 1988;68:649–742.
127. Steriade M. Cellular substrates of brain rhythms. In:
Niedermeyer E, Lopes da Silva FH, eds. Electroencephalography:
Basic Principles, Clinical Applications and Related Fields.
Baltimore: Lippincott Williams & Wilkins; 2004:31–84.
128. Jones EG. The Thalamus. New York: Plenum; 1985.
129. Domich L, Oakson G, Deschênes M, et al. Thalamic and cortical
spindles during early ontogenesis in kittens. Brain Res.
1987;428:140–142.
130. Steriade M, Deschênes M, Domich L, et al. Abolition of spindle
oscillations in thalamic neurons disconnected from nucleus
reticularis thalami. J Neurophysiol. 1985;54:1473–1497.
131. Steriade M, Domich L, Oakson G, et al. The deafferented reticu-
lar thalamic nucleus generates spindle rhythmicity. J
Neurophysiol. 1987;57:260–273.
132. Steriade M, Parent A, Hada J. Thalamic projections of nucleus
reticularis thalami of cat: a study using retrograde transport of
horseradish peroxidase and fluorescent tracers. J Comp Neurol.
1984;229:531–547.
133. Velayos JL, Jimenez-Castellanos J Jr, Reinoso-Suarez F.
Topographical organization of the projections from the reticular
thalamic nucleus to the intralaminar and medial thalamic nuclei
in the cat. J Comp Neurol. 1989;279:457–469.
134. Paré D, Steriade M, Deschênes M, et al. Physiological character-
istics of anterior thalamic nuclei, a group devoid of inputs from
reticular thalamic nucleus. J Neurophysiol. 1987;57:1669–1685.
135. Wang XJ, Rinzel J. Spindle rhythmicity in the reticularis thalami
nucleus: synchronization among mutually inhibitory neurons.
Neuroscience. 1993;53:899–904.
136. Destexhe A, Contreras D, Sejnowski TJ, et al. A model of spindle
rhythmicity in the isolated thalamic reticular nucleus. J
Neurophysiol. 1994;72:803–818.
137. Golomb D, Wang XJ, Rinzel J. Synchronization properties of
spindle oscillations in a thalamic reticular nucleus model. J
Neurophysiol. 1994;72:1109–1126.
138. Deschênes M, Paradis M, Roy JP, et al. Electrophysiology of neu-
rons of lateral thalamic nuclei in cat: resting properties and burst
discharges. J Neurophysiol. 1984;51:1196–1219.
139. Yen CT, Conley M, Hendry SH, et al. The morphology of physi-
ologically identified GABAergic neurons in the somatic sensory
part of the thalamic reticular nucleus in the cat. J Neurosci.
1985;5:2254–2268.
140. Williamson AM, Ohara PT, Ralston DD, et al. Analysis of gamma-
aminobutyric acidergic synaptic contacts in the thalamic reticular
nucleus of the monkey. J Comp Neurol. 1994;349:182–192.
141. Landisman CE, Long MA, Beierlein M, et al. Electrical synapses
in the thalamic reticular nucleus. J Neurosci. 2002;22:1002–1009.
142. Yen CT, Jones EG. Intracellular staining of physiologically identi-
fied neurons and axons in the somatosensory thalamus of the cat.
Brain Res. 1983;280:148–154.
62 Part I ■ Basic Principles
143. Steriade M, Deschênes M. The thalamus as a neuronal oscillator.
Brain Res. 1984;320:1–63.
144. von Krosigk M, Bal T, McCormick DA. Cellular mechanisms of a
synchronized oscillation in the thalamus. Science. 1993;261:
361–364.
145. Bal T, von KM, McCormick DA. Role of the ferret perigeniculate
nucleus in the generation of synchronized oscillations in vitro.
J Physiol. 1995;483(Pt 3):665–685.
146. Bal T, von KM, McCormick DA. Synaptic and membrane mech-
anisms underlying synchronized oscillations in the ferret lateral
geniculate nucleus in vitro. J Physiol. 1995;483(Pt 3):641–663.
147. Contreras D, Destexhe A, Sejnowski TJ, et al. Control of spa-
tiotemporal coherence of a thalamic oscillation by corticothala-
mic feedback. Science. 1996;274:771–774.
148. Contreras D, Destexhe A, Sejnowski TJ, et al. Spatiotemporal
patterns of spindle oscillations in cortex and thalamus. J Neurosci.
1997;17:1179–1196.
149. Jahnsen H, Llinás R. Electrophysiological properties of guinea-pig
thalamic neurones: an in vitro study. J Physiol. 1984;349:205–226.
150. Mulle C, Steriade M, Deschênes M. The effects of QX314 on thal-
amic neurons. Brain Res. 1985;333:350–354.
151. McCormick DA. Functional properties of a slowly inactivating
potassium current in guinea pig dorsal lateral geniculate relay
neurons. J Neurophysiol. 1991;66:1176–1189.
152. Budde T, Mager R, Pape HC. Different types of potassium out-
ward current in relay neurons acutely isolated from the rat lateral
geniculate nucleus. Eur J Neurosci. 1992;4:708–722.
153. Contreras D, Steriade M. Spindle oscillation in cats: the role of
corticothalamic feedback in a thalamically generated rhythm.
J Physiol. 1996;490(Pt 1):159–179.
154. Kim U, Bal T, McCormick DA. Spindle waves are propagating
synchronized oscillations in the ferret LGNd in vitro.
J Neurophysiol. 1995;74:1301–1323.
155. Sanford AJ. A periodic basis for perception and action. In:
Colquhoun WP, ed. Biological Rhythms and Human Perception.
New York: Academic Press; 1971;179–209.
156. Moruzzi G, Magoun HW. Brain stem reticular formation and
activation of the EEG. Electroenceph Clin Neurophysiol. 1949;1:
445–473.
157. McCormick DA, Prince DA. Acetylcholine induces burst firing in
thalamic reticular neurones by activating a potassium conduc-
tance. Nature. 1986;319:402–405.
158. McCormick DA, Wang Z. Serotonin and noradrenaline excite
GABAergic neurones of the guinea-pig and cat nucleus reticu-
laris thalami. J Physiol. 1991;442:235–255.
159. McCormick DA, Pape HC. Noradrenergic and serotonergic mod-
ulation of a hyperpolarization-activated cation current in thala-
mic relay neurones. J Physiol. 1990;431:319–342.
160. Buzsáki G, Bickford RG, Ponomareff G, et al. Nucleus basalis and
thalamic control of neocortical activity in the freely moving rat.
J Neurosci. 1988;8:4007–4026.
161. Bremer F, Stoupel N, Van Reeth PC. Nouvelles recherches sur la
facilitation et l’inhibition des potentiels évoqués corticaux dans
l’éveil réticulaire. Arch Ital Biol. 1960;98:229–247.
162. Lopes da Silva FH, van Rotterdam A, Storm van Leeuwen W, et al.
Dynamic characteristics of visual evoked potentials in the dog. II.
Beta frequency selectivity in evoked potentials and background
activity. Electroencephalogr Clin Neurophysiol. 1970;29:260–268.
163. Freeman WJ, van Dijk BW. Spatial patterns of visual cortical fast
EEG during conditioned reflex in a rhesus monkey. Brain Res.
1987;422:267–276.
164. Murthy VN, Fetz EE. Coherent 25- to 35-Hz oscillations in the
sensorimotor cortex of awake behaving monkeys. Proc Natl Acad
Sci USA. 1992;89:5670–5674.
165. Bouyer JJ, Montaron MF, Vahnee JM, et al. Anatomical localization
of cortical beta rhythms in cat. Neuroscience. 1987;22:863–869.
166. Freeman WJ. Mass Action in the Nervous System. New York:
Academic Press; 1975.
167. Eckhorn R, Bauer R, Jordan W, et al. Coherent oscillations: a mech-
anism of feature linking in the visual cortex? Multiple electrode
and correlation analyses in the cat. Biol Cybern. 1988;60:121–130.
168. Gray CM, Konig P, Engel AK, et al. Oscillatory responses in cat
visual cortex exhibit inter-columnar synchronization which
reflects global stimulus properties. Nature. 1989;338:334–337.
169. Gray CM, Singer W. Stimulus-specific neuronal oscillations in
orientation columns of cat visual cortex. Proc Natl Acad Sci USA.
1989;86:1698–1702.
170. Engel AK, Konig P, Gray CM, et al. Stimulus-dependent neuronal
oscillations in cat visual cortex: inter-columnar interaction as
determined by cross-correlation analysis. Eur J Neurosci.
1990;2:588–606.
171. Gray CM, Engel AK, Konig P, et al. Stimulus-dependent neuronal
oscillations in cat visual cortex: receptive field properties and fea-
ture dependence. Eur J Neurosci. 1990;2:607–619.
172. Rodriguez R, Kallenbach U, Singer W, et al. Short- and long-term
effects of cholinergic modulation on gamma oscillations and
response synchronization in the visual cortex. J Neurosci.
2004;24:10369–10378.
173. Munk MH, Roelfsema PR, Konig P, et al. Role of reticular activa-
tion in the modulation of intracortical synchronization. Science.
1996;272:271–274.
174. Niessing J, Ebisch B, Schmidt KE, et al. Hemodynamic signals
correlate tightly with synchronized gamma oscillations. Science.
2005;309:948–951.
175. von der Malsburg C, Schneider W. A neural cocktail-party
processor. Biol Cybern. 1986;54:29–40.
176. Steriade M, Amzica F. Intracortical and corticothalamic
coherency of fast spontaneous oscillations. Proc Natl Acad Sci
USA. 1996;93:2533–2538.
177. Steriade, M., Contreras, D., Amzica, F., and Timofeev, I.
Synchronization of fast (30–40 Hz) spontaneous oscillations in
intrathalamic and thalamocortical networks. J. Neurosci.
1996;16:2788–2808.
178. Roelfsema PR, Engel AK, Konig P, et al. Visuomotor integration
is associated with zero time-lag synchronization among cortical
areas. Nature. 1997;385:157–161.
179. Fries P, Nikolic D, Singer W. The gamma cycle. Trends Neurosci.
2007;30:309–316.
180. Llinás RR, Grace AA, Yarom Y. In vitro neurons in mammalian
cortical layer 4 exhibit intrinsic oscillatory activity in the 10- to
50-Hz frequency range. Proc Natl Acad Sci USA. 1991;88:
897–901.
181. Nuñez A, Amzica F, Steriade M. Voltage-dependent fast
(20–40 Hz) oscillations in long-axoned neocortical neurons.
Neuroscience. 1992a;51:7–10.
182. Steriade M, Curró Dossi R, Paré D, et al. Fast oscillations (20–40
Hz) in thalamocortical systems and their potentiation by meso-
pontine cholinergic nuclei in the cat. Proc Natl Acad Sci USA.
1991;88:4396–4400.
183. Cunningham ET, LeVay S. Laminar and synaptic organization of
the projection from the thalamic nucleus centralis to primary
visual cortex in the cat. J Comp Neurol. 1986;254:66–77.

184. Chagnac-Amitai Y, Connors BW. Synchronized excitation and
inhibition driven by intrinsically bursting neurons in neocortex.
J Neurophysiol. 1989;62:1149–1162.
185. Hösli L, Hösli E, Della BG, et al. Action of acetylcholine, mus-
carine, nicotine and antagonists on the membrane potential of
astrocytes in cultured rat brainstem and spinal cord. Neurosci
Lett. 1988;92:165–170.
186. Seigneur J, Kroeger D, Nita DA, et al. Cholinergic action on cor-
tical glial cells in vivo. Cereb Cortex. 2006;16:655–668.
187. Roopun AK, Middleton SJ, Cunningham MO, et al. A beta2-
frequency (20–30 Hz) oscillation in nonsynaptic networks of
somatosensory cortex. Proc Natl Acad Sci USA. 2006;103:
15646–15650.
188. Mandema JW, Danhof M. Electroencephalogram effect measures
and relationships between pharmacokinetics and pharmacody-
namics of centrally acting drugs. Clin Pharmacokinet. 1992;23:
191–215.
189. Salenius S, Salmelin R, Neuper C, et al. Human cortical 40 Hz
rhythm is closely related to EMG rhythmicity. Neurosci Lett.
1996;213:75–78.
190. Salenius S, Portin K, Kajola M, et al. Cortical control of human
motoneuron firing during isometric contraction. J Neurophysiol.
1997;77:3401–3405.
191. Gross J, Tass PA, Salenius S, et al. Cortico-muscular synchroniza-
tion during isometric muscle contraction in humans as revealed
by magnetoencephalography. J Physiol. 2000;527(Pt 3):623–631.
192. Hari R. Action-perception connection and the cortical mu
rhythm. Prog Brain Res. 2006;159:253–260.
193. Pfurtscheller G, Neuper C, Andrew C, et al. Foot and hand area
mu rhythms. Int J Psychophysiol. 1997;26:121–135.
194. Chen R, Yaseen Z, Cohen LG, et al. Time course of corticospinal
excitability in reaction time and self-paced movements. Ann
Neurol. 1998;44:317–325.
195. Pfurtscheller G, Neuper C, Kalcher J. 40-Hz oscillations during
motor behavior in man. Neurosci Lett. 1993;164:179–182.
196. Vanhatalo S, Tallgren P, Andersson S, et al. DC-EEG discloses
prominent, very slow activity patterns during sleep in preterm
infants. Clin Neurophysiol. 2002;113:1822–1825.
197. Vanhatalo S, Holmes MD, Tallgren P, et al. Very slow EEG
responses lateralize temporal lobe seizures: an evaluation of non-
invasive DC-EEG. Neurology. 2003;60:1098–1104.
198. Somjen GG. Electrogenesis of sustained potentials. Prog
Neurobiol. 1973;1:201–237.
Chapter 3 ■ Cellular Substrates of Brain Rhythms 63
199. Somjen GG, Trachtenberg M. Neuroglial as generator of extracel-
lular current. In: Speckmann EJ, Caspers H, eds. Origin of
Cerebral Field Potentials. Stuttgart: Thieme; 1979:21–32.
200. Caspers H, Speckmann E-J, Lehmenkuler A. DC potentials of the
cerebral cortex: Seizure activity and changes in gas pressures. Rev
Physiol Biochem Pharmacol. 1987;106:127–178.
201. Heinemann U, Walz W. Contributions of potassium currents and
glia to slow potential shifts (SPSs). In: Laming PR, Sykova E,
Reichenbach A, et al., eds. Glial Cells: Their Role in Behaviour.
Cambridge: Cambridge University Press; 1998:197–209.
202. Amzica F, Steriade M. Neuronal and glial membrane potentials
during sleep and paroxysmal oscillations in the neocortex. J
Neurosci. 2000;20:6648–6665.
203. Tschirgi RD, Taylor JL. Slowly changing bioelectric potentials
associated with the blood-brain barrier. Am J Physiol.
1958;195:7–22.
204. Held D, Fencl V, Pappenheimer JR. Electric potential of cere-
brospinal fluid. J Neurophysiol. 1964;27:942–959.
205. Kjällquist Å, Siesjö BK. Regulation of CSF pH - influence of the
CSF/plasma potential. Scand J Clin Lab Invest Suppl. 1968;102:I:C.
206. Sorensen SC, Severinghaus JW. Effect of cerebral acidosis on the
CSF-blood potential difference. Am J Physiol. 1970;219:68–71.
207. Hornbein TF, Sorensen SC. dc Potential difference between dif-
ferent cerebrospinal fluid sites and blood in dogs. Am J Physiol.
1972;223:415–418.
208. Woody CD, Marshall WH, Besson JM, et al. Brain potential shift
with respiratory acidosis in the cat and monkey. Am J Physiol.
1970;218:275–283.
209. Revest PA, Jones HC, Abbott NJ. The transendothelial DC poten-
tial of rat blood-brain barrier vessels in situ. Adv Exp Med Biol.
1993;331:71–74.
210. Voipio J, Tallgren P, Heinonen E, et al. Millivolt-scale DC shifts in
the human scalp EEG: evidence for a nonneuronal generator.
J Neurophysiol. 2003;89:2208–2214.
211. Nita DA, Vanhatalo S, Lafortune FD, et al. Nonneuronal origin of
CO 2-related DC EEG shifts: an in vivo study in the cat.
J Neurophysiol. 2004;92:1011–1022.
212. Speckmann E-J, Elger CE. Introduction to the neurophysiologi-
cal basis of the EEG and DC potentials. In: Niedermeyer E,Lopes
da Silva R, eds. Electroencephalography: Basic Principles, Clinical
Applications, and Related Fields. Baltimore: Lippincott, Williams
& Wilkins; 2005:17–30.

Dynamics of EEGs as Signals of
Neuronal Populations: Models and
Theoretical Considerations
MODELS OF EEG GENERATION—
INTRODUCTION
In Chapters 2 and 3 the neurophysiological basis of the elec-
troencephalogram (EEG) was discussed with special emphasis
on the phenomena at the cellular level. In Chapter 5, the bio-
physical aspects of the generation of EEG phenomena are con-
sidered mainly in terms of volume conduction, but the
dynamic properties of EEG patterns are only briefly men-
tioned. These dynamic properties, however, are essential for
understanding EEG phenomena. In this chapter, EEG signals
are considered as the result of the dynamic behavior of neu-
ronal populations as revealed by model studies. According to
this perspective, it is necessary to integrate experimental and
theoretical results. The latter must be obtained using models of
neuronal networks and corresponding computer simulations.
The mathematical treatment of such models has been avoided
here; the interested reader is referred to a number of other
publications that are cited for their thorough mathematical
treatment of these areas.
The fundamental assumption on which the following dis-
cussion is based is that EEG signals reflect the dynamics of elec-
trical activity in populations of neurons. A property of such
populations that is of utmost importance for the generation of
EEG signals is the capacity of the neurons to work in synchrony.
This depends on the connectivity between the neurons that
form a network. The classic terminology of Freeman (1) pro-
vides a useful systematization. Populations of neurons with
mutual interactions are called KI sets; the interactions may be
excitatory (KIe) or inhibitory (KIi). Groups of two interacting
populations of neurons are called KII sets; they are formed by
an interaction of a KIe set with a KIt set. Further, KIII sets
formed by the interaction between two KII sets can be defined.
A number of such KIII sets constitute a neural mass that may
occupy a few square millimeters of cortical surface or a few
cubic millimeters of nuclear volume in the brainstem or spinal
cord. Typically, a neural mass consists of approximately 104 to
107 neurons. An example of a KIII set is given in a symbolic way
in Figure 4.1. An essential feature of such a basic module is the
existence of multiple feedback loops. Such a set of neurons can
produce oscillatory phenomena as revealed in EEG signals. The
main parameters that characterize the dynamic behavior of the
set are (a) the synaptic time constants; (b) the length constants,
which define the distance of the interactions between different
neurons; (c) the gain factors, that is, the strength of interac-
CHAPTER
4
FERNANDO H. LOPES DA SILVA
tions, whether by way of chemical synapses or electrical cou-
plings; and (d) the intrinsic ionic and synaptic currents.
There are both inhibitory and excitatory feedback loops. The
terms negative feedback and positive feedback are often used in a
rather loose way to characterize the interactions between popula-
tions of neurons. However, the sign of feedback is defined in an
exact way by comparing the overall gain modulus of the system
with feedback (closed loop gain: |Yo(j)|) and without feedback
(open loop gain: |Y1(j)|). The feedback is said to be negative if
|Yo(j)| |Y1(j)| and positive in the opposite case (2). The exact
sign of feedback in the interaction between neuron populations is
in most cases not known. It therefore seems preferable to use the
terms excitatory and inhibitory feedback to define the main synap-
tic type of interaction between neuron populations. This will be
considered from the reference point of the “main” population,
which usually is the population that receives the input from
another source and/or sends its output to another structure.
In this chapter, models of the generation of characteristic
types of EEGs, including the spatial spread of rhythmic activity
over the cortex, and nonlinear dynamics as a new approach for
understanding EEG phenomena are presented.
The processes of generation of EEG signals in a large number
of neurons forming a neural mass are complex. Therefore, it is
useful to use both analytic and synthetic approaches to under-
stand such processes. It is essential to use both approaches in a
dialectic way. The experimental data, obtained through histolog-
ical and physiological analytic methods, must be put together just
as pieces of a puzzle are assembled. Very often the relationship
between the different pieces is still unclear. It is in this context
that a synthetic approach may be most helpful. The point is then
to put together the available data in such a way as to form the
most likely pattern. In this way, a model of a neurophysiological
system can be constructed. Such a model can help to advance
knowledge of the system in two respects. First, it provides the
possibility for testing the influence of different types of inputs or
of changing some of the properties of the constituting elements
upon the output of the system. Second, it allows the formulation
of hypotheses concerning new elementary properties, relation-
ships, and overall behavior; it may thus predict new properties of
the system, raise new questions, and suggest new experiments to
explore these hypotheses. In this way, the model can constitute a
bridge between experimentation and theory. As Harmon (3)
stated long ago, it is in suggesting functional relationships
between the activity of single neurons and the behavior of groups
of neurons that theoretical neurophysiology may be most potent.
65
66 Part I ■ Basic Principles
Figure 4.1 Lumped circuit diagram of the sets of neurons in the olfac-
tory bulb (OB), the anterior olfactory nucleus (AON), and the prepyri-
form cortex (PC). PON, primary olfactory nerve; LOT, lateral olfactory
tract; EC, external capsule; R, receptors of the nasal mucosa; gl,
glomeruli; P, periglomerular neuron; M, mitral neuron; G, granule cell;
E, superficial pyramidal cells of AON; I, inhibitory neuron; A, superficial
pyramidal neuron; B, short local interneuron; C, deep pyramidal neuron;
, excitation; , inhibition; L, latency. (Adapted from Freeman WJ.
Analytic techniques used in the search for the physiological basis of the
EEG. In: Gevins A, Remond A, eds. Handbook of Electroencephalography
and Clinical Neurophysiology, vol. 1: Methods of Analysis of Brain
Electrical and Magnetic Signals. Amsterdam: Elsevier; 1986.)
With these general objectives in mind, this chapter considers a
few models of the generation of EEG phenomena that have been
elaborated in the past decades. We focus here on models of EEG
rhythmic activities both at the macroscopic and the microscopic
level. The former are simplified models in which detailed mech-
anisms are kept to a minimum and the main properties of sub-
populations of neurons are lumped together. These models can,
however, be easily adapted to include more specific neuronal
properties and more complex connectivity. Taking into consider-
ation that in this chapter we are particularly interested in
dynamic models of neuronal networks that are responsible for
the generation of the main properties of EEG/MEG
(electro/magnetoencephalographic) signals, we pay special atten-
tion to macroscopic models. It should be emphasized that the
brain is a very complex system, and even one cortical column
consists of a large number of elements, neurons, ionic channels,
synapses, neurotransmitter systems, and glial cells that require
enormous amount of parameters and state variables for a proper
description. The dynamics of EEG/MEG signals, however, can be
accounted for within the framework of systems operating within
spaces with much lower dimensions. This is the realm of macro-
scopic models, namely neural mass models (NMMs) that use
mean field concepts, where the properties of subpopulations of
neurons are lumped together based on the fact that many neu-
rons within a given neuronal system share basic properties.
Nonetheless we also consider, albeit in abbreviated form, some
results of studies using microscopic models that are also relevant
for understanding some characteristics of EEG/MEG signals. In
this respect the models developed by Edelman and collaborators
(4) are good examples of models that combine detailed proper-
ties of neurons and synapses within large-scale systems. Although
these models have been introduced mainly to study functional
connectivity at different scales of complexity with respect to neu-
rocognitive processes, their capacity to generate brain rhythms
and EEG-like phenomena should also be stressed. It is to be
expected that the applications of this new generation of large-
scale computer models of complex neuronal networks will grow
fast in the coming years, since these kinds of models can now be
realized by virtue of the current availability of powerful com-
puter systems. An ambitious project along this research line is the
Blue Brain project that has the aim of creating a synthetic mam-
malian brain by reverse-engineering, using IBM’s Blue Gene
supercomputer, in order to investigate the brain’s architectural
and functional principles underlying cognitive processes (5).
Another new trend is the development of Topological Models
of neuronal networks with their patterns of connectivity that
can be studied using the mathematical theoretical approach of
“small-world” graphs. These are mathematical graphs that con-
sist of nodes that can be reached from every other by a small
number of steps, which have been applied in a wide variety of
fields, from gene to social networks. Watts and Strogatz (6)
classified such graphs according to two independent structural
features, namely the clustering coefficient and average node-to-
node path length. This approach is in essence an analytical tool
that can extract interesting parameters to characterize func-
tional connectivity within networks of cortical neurons and
conditions determining neuronal synchrony (7,177) and to ana-
lyze fMRI and MEG/EEG data (8), but it has not addressed
explicitly the question of the generation of EEG/MEG activities.
MACROSCOPIC MODELS
OF EEG GENERATION
Basic Principles: The Wilson–Cowan Approach
At the macroscopic level we describe NMMs that are based on
the mean-field approximation approach. Under this heading we
consider two classes of models: (i) models that account for
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
time-dependent properties of neural populations and (ii) models
that include, in addition, space-dependent properties. In gen-
eral terms this approach was introduced in a systematic way in
the Neurosciences by Wilson and Cowan (9) who proposed a
theoretical framework that emphasizes not the properties of
individual cells but rather those of neuronal populations. A
basic assumption is that cortical neurons form a dense network
consisting of, at least, two subpopulations that are intercon-
nected either directly or by way of interneurons. The basic vari-
able is the proportion of neurons of each subpopulation that
are active at a given time, such that the relevant variable is not
the single spike but the average spike frequency. The dynamics
of the neural mass depends on the interaction of excitatory and
inhibitory neuronal subpopulations. Accordingly the model
requires the use of two variables E(t) and I(t), describing the
activity of the excitatory and inhibitory subpopulations, respec-
tively, to characterize the state of the whole population.
Based on these assumptions the general Wilson and Cowan
equations (9) governing the dynamics of a localized population
of neurons were derived:
t models and showed that they may produce in-phase oscillations.
E(t r) c1 E(t¿)dt¿d #Se
t r
t
e a (t t¿) [c1E(t¿) c2I(t¿)
q at the same time a decrease (event-related desynchronization or
and
t EEG/MEG power within the alpha frequency range (mu rhythm)
I(t¿)dt¿d #Si
I(t r¿) c1
t r¿
t
e a (t t¿) [c3E(t¿) c4I(t¿)
q mic activities within the alpha frequency range, several
This formalism describes the activities of the excitatory E(t)
and inhibitory I(t) subpopulations; c1 and c2, and c3 and c4, are
the connectivity constants representing the average number of
excitatory and inhibitory synapses per cell for each subpopula-
tion, is the decay time constant, Se(x) and Si (x) are the
response functions that give the expected proportion of cells in
a subpopulation that would respond to a given level of excita-
tion and are represented by sigmoid functions; P(t) and Q(t)
are the external inputs to the excitatory and the inhibitory sub-
populations, respectively; r is the refractory period.
Neural Mass Models of EEG Rhythms
Based on the theoretical formalism introduced by Wilson and
Cowan (9), Lopes da Silva et al. (10) described an NMM with the
aim of simulating the conditions under which alpha rhythms
occur. The model consisted of lumped subpopulations of inter-
connected excitatory and inhibitory neurons with properties of
thalamocortical cells and a special subpopulation of interneurons.
The mathematical analysis was initially carried out for a linearized
approximation of the neuronal network in order to express the
transfer function of the network and the power spectrum of the
67
average membrane potential. The latter was interpreted as the
local EEG signal, which in this case consisted of a predominant
alpha rhythm. Later the analysis was extended to take into account
a first approximation of the nonlinear behavior (11). In the sec-
tion of alpha rhythms we describe these models in more detail.
More recently these kinds of models were extended to account
for EEG frequency components other than the alpha rhythms.
In this context the model study of Jansen and Rit (12) marks an
important advance. These authors introduced a lumped-parame-
ter model of a cortical column consisting of a population of
pyramidal cells interconnected with populations of excitatory and
inhibitory interneurons, which displays five kinds of simulated
EEG outputs when fed with random noise. These different EEG
outputs are the following: low amplitude noise for high values of
the inhibitory feedback, slow periodic activity similar to that
obtained in comatose patients, typical alpha rhythmic activity,
noisy alpha activity, and low-amplitude high-frequency activity
similar to beta activity for high values of the excitatory feedback. In
order to explore how different cortical columns interact, Jansen
and Rit investigated the behavior of two identical single-column
Another case where NMMs of interacting neuronal popula-
tions helped to clarify the origin of a specific EEG/MEG phenom-
enon was studied by Suffczynski et al. (13). This phenomenon is
the so-called focal ERD–surround ERS, described in detail in
P(t¿) ] dt¿f Chapter 45, where a given event (e.g., finger movement) can cause
ERD) and an increase (event-related synchronization or ERS) of
depending on the site over the scalp. The analysis of coupled
NMMs representing thalamocortical systems responsible for the
hand and for the foot movements, respectively, was able to reveal
how this reciprocal “ERD–ERS” phenomenon takes place.
Q(t¿) ] dt¿f . Based on the results obtained with NMMs generating rhyth-
researchers extended these models to account for different EEG
frequency bands besides the alpha rhythm. In this respect,
David and Friston (14) described a lumped model that gener-
ates various rhythmic activities ranging from delta to gamma,
according to the kinetics of the populations included in the
model. These authors also investigated the dynamics of the
interaction between different cortical areas by coupling two
lumped models. Depending on coupling strength, direction of
coupling (unidirectional or reciprocal coupling), and propaga-
tion delays, they showed that the oscillation frequency depends
on the delay and that a reciprocal coupling is characterized by
oscillatory signals in phase or in antiphase. This latter result
reproduces the experimental finding of zero-lag synchroniza-
tion among remote cortical areas as found experimentally by
Roelfsema et al. (15) and Chawla et al. (16).
With the aim of investigating how different cortical regions
can display complex EEG frequency spectra and several pat-
terns of functional connectivity, Zavaglia et al. (17,18) also
developed interconnected lumped models. They showed that
three populations, consisting of four subpopulations (pyrami-
dal neurons, excitatory, slow, and fast inhibitory interneurons),
arranged in parallel are sufficient to account for complex EEG
68 Part I ■ Basic Principles

spectra with dominant frequency ranging form theta to alpha
and gamma. This model is akin to the comprehensive NMM
introduced by Wendling et al. (19) to account for the generation
of different kinds of neural activities and the corresponding
EEG signals, including epileptic activity as described in more
detail in the section dedicated to models of epileptic phenom-
ena generated by neural populations. This model illustrates very
clearly the main features of the currently used NMMs (Fig. 4.2).
Neural Mass Models w ith Spatial-Temporal
Features—Continuous Neural Field Models
In addition to the lumped NMMs described above, where spatial
aspects are not explicitly simulated, there are also models that
include spatial properties and consist of spatiotemporal chains of
different subpopulations of neurons. An example is the model of
van Rotterdam et al. (20) that consists of a series of pyramidal
cells and interneurons interconnected by means of recurrent col-

Figure 4.2 (a): Neural mass model consisting of a subpopulation of principal cells (pyramidal cells) that project to and receive feedback from sub-
populations of local interneurons. Feedback to pyramidal cells is either excitatory (recurrent excitation) or inhibitory: dendritic-projecting interneu-
rons with slow synaptic kinetics—IPSP (slow)—shown in (c)—and somatic-projecting interneurons (gray hexagon) with faster synaptic
kinetics—IPSP (fast)—shown in (c). (b): Correspon ding block diagram representing for each subpopulation, the average pulse density of afferent
action potentials that is transformed into an average inhibitory or excitatory postsynaptic membrane potential using a linear dynamic transfer func-
tion with impulse response h EXC(t), h SDI(t), and h FSI(t) while this potential is converted into an average pulse density of action potentials fired by
the neurons by way of a static nonlinear function (asymmetric sigmoid curve, S(v)). Interactions between main cells and interneurons are repre-
sented by seven connectivity constants (C1 to C7), which account for the average number of synaptic contacts. One of the model outputs represents
the summated average postsynaptic potentials on pyramidal cells. It simulates an EEG signal. The three main parameters of the model respectively
correspond to the average excitatory synaptic gain (EXC), to the average slow inhibitory synaptic gain (SDI), and to the average fast inhibitory synap-
tic gain (FSI). (Adapted from Wendling F, Hernandez A, Bellanger JJ, et al. Interictal to ictal transition in human temporal lobe epilepsy: insights
from a computational model of intracerebral EEG. J Clin Neurophysiol. 2005;22(5):343–356.)
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
laterals and inhibitory fibers with connectivity functions
assumed to be homogenous while the strength of the connectiv-
ity was an exponentially decreasing function of distance. This
model was constructed to study the factors that determine the
propagation of alpha waves along a cortical sheet. Assuming a
continuous chain of neurons the model’s output, u(x,t), is the
space-averaged membrane potential density of the main cells. To
study propagation properties in the neuronal chain an expression
has been derived as a linear inhomogenous partial differential
equation by means of which one can determine group velocity of
alpha waves and dispersion properties. Alternatively the neuronal
chain may be conceived as a two-dimensional system with a spe-
cific transfer function relating a random input p(x,t) to the out-
put u(x,t). The model results were able to emulate experimental
data, namely the propagation velocity that was found experimen-
tally in the canine cortex to be about 0.3 m/sec (21) based on
parameters taken from physiology and histology, as explained
further in more detail in this chapter.
A series of other models, akin to that of van Rotterdam et al.
(20), where the cortical neuronal populations are modeled as
forming a continuum have been made, including models incor-
porating stochastic features called continuous neural field models.
In this chapter we focus on those aspects that are specifically rele-
vant for EEG/MEG generation, while the more general aspects of
cortical processing of information with respect to neurocognitive
functions lie outside the scope of this text. With respect to these
continuous neural field models, in general, the reader may be
referred to comprehensive reviews of Harrison et al. (22), Deco et
al. (23), and Jirsa (24). Examples of continuous neural field models
are those of Jirsa (24,25) and of the Australian group (26–29). Jirsa
(24) gives a very thorough account of how a set of state equations
with parameter values based on physiology can account for the
dynamics of cortical neuronal populations at different spatial
scales: microscopic, mesoscopic, and macroscopic. At microscopic
level, fast synaptic feedbacks appear to be crucial to the stability of
the excited cortex. At mesoscopic scale the model shows how fields
of synchrony can be established, and at macroscopic scale it shows
how traveling waves may be generated in the cortex. The last two
levels are particularly relevant with respect to the dynamics of
EEG/MEG generation. With respect to these models the activity
state of a single neuron can be described by its membrane poten-
tial, V, the capacitive current, I, or the time since the last action
potential, T. Each of these variables can be represented by a
dimension in the phase space of the neuron’s activity state.
Accordingly this phase space would be three-dimensional and the
state of a neuron would correspond to a point v = {V,I,T} or par-
ticle in phase space. Using a general formalism the activity state of
a population of neurons at any time, t, can be described in phase
space by a density p(v,t). As the state of each neuron evolves the
ensemble density, p(v,t) evolves until it reaches some steady state.
The evolution of the ensemble density dynamics can be described
using the formalism of the Fokker–Planck equation. The latter
describes in general terms the time evolution of the probability
density function of the position of a particle in a given multidi-
mensional space. The interested reader may find a thorough treat-
ment of the applications of this formalism in the study of
continuous field models by Deco et al. (23) and Harrison et al. (22).
69
Neural Mass Models and Hemodynamic Signals
NMMs are also being applied to make a bridge between the
activity of neuronal networks as reflected in EEG signals and
hemodynamic signals as measured by way of functional MRI, in
order to understand better the working of cortical networks
underlying cognitive processes. This constitutes a central feature
of dynamic causal modeling (DCM) (30) that has been applied
not only to fMRI signals but also to electrophysiological signals
such as event-related potentials (ERPs) (31).
Valdés-Sosa et al. (32) proposed an approach where the
summed postsynaptic potentials generated by an NMM of a cor-
tical area are used in the forward mode to estimate the primary
current density distribution of a cortical area (Fig. 4.3). This is
then transformed into the signal recorded at EEG sensors by
applying a linear spatial convolution with the EEG lead field, that
is, the function that relates any source in the brain to the corre-
sponding measurements at the EEG sensors. Simultaneously the
summed postsynaptic potentials generated by the NMM are trans-
formed into a local Vasomotor feed forward signal that is then fur-
ther transformed by way of a temporal convolution with the
hemodynamic response function (HRF) into a simulated BOLD
signal. In this way the information obtained from the EEG signal
and from the corresponding fMRI BOLD signal recorded from the
same region of interest can be fused (Fig. 4.3). We should note that
although this approach opens interesting novel perspectives and
Forward Models for fusion of EEG/MEG & BOLD
signals based on a common Neural Mass Model
Neural Mass Model
generates an ensemble
of postsynaptic potentials
Transfer function Transfer function
Estimated cortical Vasomotor feedforward
primary current density signal
EEG/MEG BOLD signal
Figure 4.3 The output of a neural mass model in the form of summed
postsynaptic potentials can be used to estimate the primary current
density distribution of a cortical area that is transformed into the
EEG/ MEG signal. Simultaneously the summed postsynaptic potentials
generated are transformed into a local Vasomotor feed-forward signal
that is then further transformed by way of a temporal convolution with
the hemodynamic response function into a simulated BOLD signal. In
this way EEG/ MEG information can be fused with the corresponding
fMRI BOLD signal recorded from the same region of interest. (Inspired
from Valdes-Sosa PA, Sanchez-Bornot JM, Sotero RC, et al. Model
driven EEG/ fMRI fusion of brain oscillations. Hum Brain Mapp.
2009;30(9):2701–2721.)
70 Part I ■ Basic Principles
merits being further elaborated, it is based on concepts that are still
insufficiently established. Namely more research is needed to bet-
ter understand how neural activities, generated by neural popula-
tions with complex geometrical configurations, are transformed
into cortical current distributions on the one hand, and how they
are associated with BOLD signals on the other.
MODELS OF SPECIFIC EEG ACTIVITIES
Here we describe a number of models that account for the gen-
eration of specific types of EEG activities, namely the following:
• Models describing the generation of the EEG of olfactory
areas of the brain as proposed extensively by Freeman (1) and
later further elaborated (33,34) that may be considered para-
digmatic of neural mass modeling approaches;
• Models of alpha rhythms of the thalamus and cortex, since
these are one of the most conspicuous features of the human
EEG, as proposed by Lopes da Silva et al. (10,11) and elabo-
rated in more detail by van Rotterdam et al. (20), and
extended later by Jansen and Rit (12), Nunez (35), Wright and
Liley (36), Stam et al. (37), Valdes et al. (38), Robinson et al.
(39), Suffczynski et al. (13), David and Friston (14), and
Zavaglia et al. (17,18);
• Detailed models of membrane and synaptic properties of
thalamic cells and circuits responsible for the generation of
7- to 14-Hz spindle rhythmicity that occurs at the onset and
in the light stages of sleep, based on the experimental findings
of Steriade et al. (40–48);
• Models of epileptiform EEG, namely of absence seizures, that
display complex nonlinear dynamics, namely bistable behav-
ior, where abrupt transitions occur between two states—nor-
mal and paroxysmal (49,50)—and of epileptic seizures of the
limbic cortex (19,51), as well as detailed models of the gener-
ation of epileptiform transients by Traub (52), Traub and
Wong (53), Traub et al. (54,55), and Traub and Miles (56);
• Models of beta and gamma rhythms developed by Traub et al.
(57) with realistic simulations of synapses (55,57–60). We do
not include here models of hippocampal theta rhythms since
the basic physiology of these rhythmic activities has already
been described in Chapter 3.
Model of Olfactory Area EEG
The main neuronal populations of the olfactory brain struc-
tures that are of importance for an understanding of the gener-
ation of the olfactory EEG can be summarized as follows: in the
bulb, the mitral and tufted cell (M) population, which is excita-
tory, and the granule cell (G) population, which is inhibitory; in
the PC, the surface pyramidal population, which is excitatory
(E), and the basket cell population, which is inhibitory (I). The
interaction between the M and the granule neuron populations
forms an inhibitory feedback loop. The interaction between the
surface pyramidal and the basket cell populations in the PC also
forms an inhibitory feedback loop (Fig. 4.1).
As can be seen in Figure 4.1, there are also mutual excitatory
interactions between M neurons in the bulb and between
pyramidal A neurons in the PC, as well as mutual inhibitory
interactions between granule cells in the bulb and basket cells in
the prepyriform cortex (PC). Excitation of the M cells by pri-
mary olfactory nerve (PON) inputs leads to excitation of the
granule (G) neurons and to feedback inhibition of the M neu-
rons. In consequence, the G neurons are disexcited and the M
neurons are disinhibited. In this way, the M neurons may
become excited again, and the whole process may continue in
an oscillatory mode. The basic oscillation is at 40 Hz in cat and
60 to 80 Hz in rabbit (61). A similar type of inhibitory feedback
loop occurs in the PC, and the EEG of this cortical structure
resembles that of the olfactory bulb (OB) in frequency content.
In this model and similar ones, the action potentials are con-
verted into graded currents at the somadendritic synapses caus-
ing the postsynaptic potentials. In a first approximation, these
postsynaptic potentials may be considered to sum at the soma.
At the trigger zone, the axon hillock, these potentials are trans-
formed into spike trains depending on a threshold.
The postsynaptic potentials may be considered as the wave
(W) mode of operation of the neural set; the action potentials
represent the pulse (P) mode. This is, of course, a simplification
of processes taking place at the membrane, but at the level of gen-
eralization at which NMMs of the EEG are constructed, these are
the main features that must be considered. The sensitivity of the
threshold process in the bulb and cortex is controlled by centrifu-
gal influences that originate in the forebrain and brainstem,
among others from the locus coeruleus (62). The conversion
between the postsynaptic potential (W mode) and the pulse rate
(P mode) may be given in terms of the sigmoid curves shown in
Figure 4.4 for a single neuron and for the population or ensem-
ble. The wave–pulse conversion can be considered as a static sig-
moidal nonlinearity at the single neuron level. However, over
small signal ranges this relation may be assumed to be linearized,
and therefore it may be represented by a simple coefficient. To
simulate the local dynamics of the EEG, delays due to synaptic
and conduction processes may be accounted for by a stage of
integration. Along these lines, the dynamics of these structures
may be represented by a mathematical model consisting of cou-
pled differential equations that incorporate static nonlinear func-
tions (for a mathematical treatment see Refs. 1, 34, and 63). The
main input to the populations of the OB comes from the olfac-
tory mucosa and is modulated by the respiration rate. A study of
the behavior of the KII set described above shows that this set has
multiple stable states. Two are of particular interest because they
represent the most common situations. The first is a quasiequi-
librium state that is maintained by a steady input of low intensity;
the second is the so-called carrier state (34), characterized by a
limit cycle with quasisinusoidal EEG activity at 40 Hz (cat) or 60
to 80 Hz (rabbit) that occurs during inspiration. This nonlinear
oscillation represents a limit cycle because the feedback gain is
displaced to the excitatory side of the resting value with an
increase in pulse input (either of sensory origin or depending on
arousal level) that may drive the KII set out of the equilibrium
state. Certain drugs such as carbachol or picrotoxin can induce a
limit cycle state of long duration (34). A later section shall con-
sider the relevance of nonlinear dynamics such as the theory of
deterministic chaos for understanding EEG generation. Here it is
sufficient to observe that the EEG signals simulated by the set of
nonlinear coupled differential equations that describe the KII set
approximate the experimentally recorded EEGs.
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
Figure 4.4 Conversion operations between the wave mode (W) and
the pulse mode (P) for a neuron and an ensemble of neurons. At the
neuronal level both operations are time-varying (t); the P-W conversion
is nonlinear; the W-P conversion is linear between threshold and catho-
dal block. At the ensemble level, the P-W conversion is restricted to a
small linear range and W-P conversion is static and nonlinear. The
derivative of this curve defines the gain; this may change ( o) depend-
ing on arousal level. The P-W conversion can vary to represent synap-
tic changes occurring with learning. (Adapted from Freeman WJ.
Dynamics of image formation by nerve cell assemblies. In: Basar E,
Flohr H, Haken H, et al., eds. Synergetics of the Brain. Berlin: Springer-
Verlag; 1983:102–121.)
Alpha Rhythm Models: Spatial
and Temporal Characteristics
A current view in neurophysiology is that the origin of certain
rhythmic electrical activity seen in the EEG, such as that char-
acteristic of some thalamic nuclei and cortical areas after
administration of a barbiturate (64,65) and possibly also that
occurring spontaneously, such as sleep spindles and alpha
rhythms (66,67) is to be found in populations of neurons char-
acterized by the interaction between excitatory and inhibitory
neuronal populations. Different types of models have been
developed to account for the generation of EEG rhythmic activ-
ities within the alpha frequency range. Some of these models
operate at the microscopic level of the single neuron or the
small local network (68,69), others at the macroscopic level
with the aim of accounting for global properties of EEG signals
recorded from the scalp (not limited to alpha rhythms) (35).
Recently attempts have been made to combine the local and the
global levels within a synthetic theoretical framework, among
others by Nunez who developed a thoughtful “marriage of local
and global theories” (Chapter 11 in Ref. 35). A particular inter-
esting discussion of this issue is the paper of Wright and Liley
(36) followed by a wide-open peer commentary.
At the cellular level, an important issue that has been repea-
tedly discussed in studies of the mechanisms underlying EEG
71
rhythmic activities is the question of whether such rhythms are
caused by single cells with pacemaker properties. A significant
advance in this discussion was achieved when it was demon-
strated by Jahnsen and Llinás (70,71) that some types of thalamic
neurons display oscillatory behavior in vitro, even after blockage
of synaptic transmission. Most neurons studied in this way tend
to generate oscillations in the frequency range of 6 to 10 Hz.
These intrinsic oscillatory properties of some neurons are most
likely of importance in shaping the rhythmic behavior of the net-
works to which they belong. However, these properties may not
be sufficient to account for the network rhythmic behavior. One
main argument supports this statement: the disconnection of
cortically projecting thalamic cells from their inputs, which arise
in the reticular (RE) nucleus of the thalamus, abolishes spindles
in thalamocortical relay (TCR) neurons (72). However, single
TCR neurons may preserve their intrinsic membrane oscillatory
properties in vitro (70,71). Interestingly, the capacity of both
TCR–RE circuits and the isolated RE nucleus, in vivo, to generate
spindle oscillations contrasts with the fact that the RE nucleus
isolated in vitro does not show spontaneous oscillations (73–78).
This is likely because the input conditions of the RE neurons in
vitro are different from the situation in vivo because the neurons
lack neuromodulation influences, including cholinergic, nora-
drenergic, and serotoninergic inputs. In modeling studies, the
influence of neuromodulators on RE neurons membrane poten-
tial was simulated (68), namely noradrenaline and serotonin, by
blocking leak K currents and thus depolarizing the RE neurons
to the same level as observed in vivo. These studies showed that
waxing–waning oscillations could occur in the RE network, sim-
ilar to those seen in vivo. This finding provides clear evidence that
the oscillatory behavior of these neuronal populations of RE neu-
rons depends on specific levels of the neuronal membrane poten-
tial that is affected by several inputs, both at short- and
long-range. Among the latter, the neuromodulatory inputs aris-
ing from the raphe nucleus (serotoninergic), the locus coeruleus
(noradrenergic), as well as glutamatergic projections form fore-
brain areas are particularly important. Among the former it is
interesting to note that gap junctions have been shown to exist
between RE nucleus (79) and thus may also participate in the
modulation of their oscillatory properties.
To understand how oscillations may be generated in thala-
mic nuclei, we must consider the basic structure of these neu-
ronal networks. In the thalamic relay nuclei where this type of
rhythmic activity can be recorded, namely the relay nuclei of
the thalamus, three main types of neurons can be distinguished
(see also Chapter 3 for a discussion of the physiology of these
cellular elements): the TCR neurons, whose axons project to the
cortex; the RE nucleus neurons that contribute to the inhibitory
feedback control of the former; and the local, intrinsic, neu-
rons. The RE forms a thin sheet of neurons that surround the
anterior and lateral side of the thalamus and interact synapti-
cally with the TCR neurons. The axons of the RE cells have as
main target the TCR cells. Their dendrites make synaptic con-
tacts with the axons of the TCR cells. Thus the TCR and the RE
neurons are interconnected by means of a feedback loop as
schematically shown in Figure 4.5. Both the RE and the local-
circuit neurons contain aminobutyric acid (GABA) as trans-
mitter. This implies that the feedback of the RE on the TCR
72 Part I ■ Basic Principles
Figure 4.5 Simplified scheme of a thalamic neural network where the
main features are presented. The thalamocortical relay (TCR), the retic-
ular thalamic (RE), and the local-circuit interneurons (IN) are indicated,
along with projections from a cholinergic population (ACh). The synap-
tic triad formed by the dendrites of TCR and IN neurons with the spe-
cific afferent (spec. aff.) fibers is also indicated. Three types of synapse
are shown: excitatory, aminobutyric acid (GABA)ergic (inhibitory), and
cholinergic. The effect of the latter on RE neurons is assumed to be an
increase in potassium conductance (gk), whereas on the TCR neurons
the effect is the opposite. (From Lopes da Silva FH. Neural mechanisms
underlying brain waves: from neural membranes to networks.
Electroencephalogr Clin Neurophysiol. 1991;79:81–93, with permission.)
neurons is GABAergic and thus inhibitory. The RE neurons
receive fast GABAA-mediated inhibitory postsynaptic potentials
(IPSPs) from neighboring RE neurons; in this way these neu-
rons form a chain of cells interconnected by inhibitory
synapses. In addition they receive fast glutamatergic (AMPA
type) excitatory postsynaptic potential (EPSPs) from TCR neu-
rons. In turn the TCR neurons receive both fast GABAA and
slow GABAB IPSPs from RE neurons (74,75). In a thalamic
nucleus, a number of feedback circuits between TCR and RE
neurons can be distinguished.
In the context of this discussion, it is important to note that
the TCR neurons can work essentially according to two distinct
modes: either (a) as relay cells that simply depolarize and can
produce single spikes in response to an adequate input volley or
(b) as oscillatory cells producing bursts of high-frequency
spikes that are repeated in a rhythmic fashion. The resting
membrane potential of a TCR neuron determines whether such
a neuron will be active in the relay or in the oscillatory mode.
To understand this question, it is necessary to take into account
the intrinsic membrane properties of these cells, namely their
main ionic conductances and the corresponding dynamics. It
should be noted that these properties became known mainly
due to microphysiological investigations in thalamic slices, in
vitro, and in isolated brain preparations by the groups of Llinás
and collaborators (70,71). For a comprehensive account of
these membrane properties, the reader is referred to the origi-
nal publications and to the excellent reviews by Steriade and
Llinás (80), Steriade et al. (66,67), McCormick (81),
McCormick and Bal (82), Destexhe et al. (69), Destexhe and
Sejnowski (83), (84), Lörincz et al. (85), and to Chapter 3.
A basic question is whether the different types of oscillations
can be found in vitro in the absence of external inputs, since
this would give support to the idea that these thalamic cells may
support pacemaker rhythmic activity. Jahnsen and Llinás (70)
provided an initial answer to this question by showing that the
oscillatory activity at about 10 Hz may be observed in the
absence of external drive in vitro. However, we must also con-
sider what is the contribution of the feedback circuits between
TCR and RE neurons to the generation of EEG oscillations
in vivo.
In this context the most direct experimental evidence per-
tains to the generation of spindle oscillations occurring during
the early stages of sleep. Spindles appear in the EEG of most
mammals as oscillations with a frequency in the range from 7
to 14 Hz, which form bursts lasting for 1.5 to 2 sec that recur
periodically with an irregular periodicity of about 5 to 10 sec.
This type of spindle can also be seen in animals under barbitu-
rate anesthesia. It was shown by Steriade et al. (72) that no thal-
amic nucleus is capable of generating spindle oscillations after
disconnection from the RE nucleus. Therefore, the existence of
feedback circuits including the RE nucleus is necessary for spin-
dling to occur under normal conditions in vivo. Nonetheless
there are other mechanisms that may be responsible for alpha
rhythmic generation at least in brain slices in vitro but have not
been incorporated in explicit models as described by Lörincz et
al. (85) and in Chapter 3.
In conclusion, under the normal conditions of an intact
brain, both synaptic interactions and input signals play an
essential role in setting the conditions necessary for oscillatory
behavior to occur. The frequency of this oscillation depends on
the intrinsic membrane properties on the membrane potential
of the individual neurons and on the strength of the synaptic
interactions.
Simplified Distributed Network Model
A simplified model has been proposed (10,11) that incorpo-
rates the limited physiological and histological data available at
that time; these assumptions were sufficient to explain the gen-
eration of the alpha rhythm, that is, the EEG stochastic signal
that lies in the frequency range of 8 to 13 Hz. This appears
mainly at eye closure and can be recorded best from the poste-
rior regions of the scalp in humans. A similar type of activity
has also been recorded at the visual cortex and some thalamic
nuclei in dogs (86) and cats (87,88). The model in its most sim-
ple form consists of a set of simulated neurons, thalamocortical
relay (TCR) cells, and interneurons (INs). The latter may be
assumed to be the RE cells as shown in Figure 4.5, although the
IN cells shown in this figure may have a complementary effect.
For simplicity in the model, the cells providing feedback are
denoted, in the following, as INs. In contrast with the lumped
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
models, some examples of which have been given for the olfac-
tory system, this model is a distributed one in the sense that
each neuron is modeled individually; it occupies a specific posi-
tion within a matrix of neurons. Such a distributed model
offers some special possibilities for theoretical studies that are
not easily available in lumped models.
Without entering into great detail about the alpha rhythm
distributed model, it is of interest to point out some of its
characteristics. First, the structure of the distributed model
initially used (10) consisted of 144 TCR neurons and 36 INs;
the 32 TCR neurons that surrounded one IN were responsible
for its excitation, while each IN was responsible for the inhi-
bition of the 12 TCRs around it. The excitation and inhibition
were represented by time functions that were an approxima-
tion of the waveforms of EPSPs and IPSPs, respectively. The
matrix of neurons should be viewed as lying at the surface of
a torus. Furthermore, each neuron fired whenever the mem-
brane potential exceeded a simple threshold; a short refrac-
Figure 4.6 Results obtained with the distributed model of alpha
rhythm. Upper panel: Input density of the input to all TCR neurons.
Middle panel: Wave mode summed membrane potentials of all TCR
neurons. Lower panel: Pulse mode: impulse density of all TCR neurons.
Right: The amplitude distributions and the power spectra of the three
signals are shown. Note the apparent waxing and waning of the simu-
lated alpha rhythm and the dominant alpha rhythm with a second har-
monic. (Adapted from Lopes da Silva FH, Hoeks A, Smits H, et al.
Model of brain rhythmic activity. Kybernetik. 1974;15:27–37.)
73
tory period then followed. A basic characteristic of this model
was that the input received by each TCR neuron was in the
form of a series of pulses (action potentials) that had a
Poisson distribution; the inputs to the individual TCR neu-
rons were uncorrelated. The model’s output signal was the
sum of the membrane potential fluctuations of all TCR neu-
rons. This output signal simulated the general spectral prop-
erties of an alpha rhythm. This result is illustrated in Figures
4.6 through 4.8.
In the distributed model, it is also easy to investigate the
result of changing the range of excitatory and inhibitory influ-
ence of, respectively, TCRs and INs. In the case of 4.7, there
were 12 TCR neurons that were inhibited by one IN. This is the
common situation; the effects of reducing this number to four
and of increasing it to 21 are shown in Figures 4.7 and 4.8.
Reduction of the inhibitory area of influence of an IN results
in a decrease of the spectral peak amplitude (alpha frequency)
and an increase in bandwidth; an increase in the inhibitory
area has the opposite effects. These are some of the phenom-
ena that can be studied using a distributed neuronal model of
rhythmic activity. However, this type of model has one main
disadvantage: it is difficult to treat analytically. Therefore, it
was necessary to develop a lumped model that would take into
account the main characteristics of the distributed model. A
lumped model of the same type as those of Freeman, previ-
ously described, permits a more general treatment of the acti-
vity of neuronal populations.
The Neural Mass (Lumped) Alpha Rhythm Model
The development of the lumped model in analytical terms
included the step of translating the properties of the discrete
Figure 4.7 Output signals of the distributed alpha model: Vm (t) repre-
sents the summed membrane potentials of all TCR neurons and E(t) rep-
resents the impulse density of the same neurons. The power spectrum
of Vm (t) is shown on the left below. In this case, a smaller inhibitory area
than in the case of Figure 4.8 has been simulated. Here one interneuron
(IN) inhibits only four TCR neurons. (Adapted from Lopes da Silva FH,
van Rotterdam A, Barts P, et al. Models of neuronal populations: the
basic mechanisms of rhythmicity. In: Corner MA, Swaab DF, eds.
Perspectives of Brain Research, vol. 45. Prog Brain Res. 1976:281–308.)
74 Part I ■ Basic Principles
Figure 4.8 Similar to Figure 4.7 but for a model with a larger
inhibitory area. Here 1 IN inhibits 21 TCR neurons. Note that the inten-
sity of the alpha peak is larger and the bandwidth is smaller in compar-
ison with the model shown in Figure 4.7. (Adapted from Lopes da Silva
FH, van Rotterdam A, Barts P, et al. Models of neuronal populations:
the basic mechanisms of rhythmicity. In: Corner MA, Swaab DF, eds.
Perspectives of Brain Research. vol. 45. Prog Brain Res. 1976;281–308.)
matrix of neurons into a system such as that depicted in
Figure 4.9. Essentially, the neural mass (here thalamic nucleus),
the EEG of which is of interest, forms a KII set, which is consti-
tuted by the interaction of an excitatory (M) and an inhibitory
(INIF) population (KIe and KIi sets). In addition, the scheme of
Figure 4.9 shows a second KIe (INEF) set that provides excita-
tory feedback and an additional KIe set that may receive a direct
input (P(t)) from a source situated outside the neural mass and
in this way lead to feed-forward inhibition. To obtain an ana-
lytic simplification of the distributed model, it is useful to con-
sider the general formalism of Wilson and Cowan (9) as
described in the introduction to this chapter.
Zetterberg (89) has made use of this formalism to simplify
the distributed model of alpha rhythm. Considering the simpli-
fied block diagram of Figure 4.10, where the EPSPs and IPSPs
are represented by he(t) and hi(t), respectively, and the sigmoid
wave pulse transfer functions are represented by fE(V) and fi (V),
expressions for the transfer function of the simulated neuronal
set can be obtained (for the mathematical treatment, see Ref. 11).
One of the outputs of such a neuronal set represents the EEG
signal of the population, assuming that the latter is determined
by the somadendritic membrane potential of the main cells.
The corresponding spectrum is shown in Figure 4.11 for differ-
ent values of K, which represents a measure of the gain factor of
the neuronal set. The spectrum R(f) indeed shows a peak
around 10 Hz, as was expected. The bandwidth and the peak
frequency both depend on the coefficient K. Up to a point, a larger
K will cause an increase in the frequency of the rhythmic activ-
ity, but too large a K will cause instability. K is also proportional
to the derivatives 1/ael and 1/ail of the sigmoid functions in their
working point, which in turn is determined by the mean input
pulse density P(t). The peak frequency will increase and the
Figure 4.9 Block diagram of the lumped model for rhythmic activity
with a possibility of modulation of inhibitory feedback and excitatory
feedback. The cartel consists of four populations. M, main cells with
input and output impulse density, respectively, P(t) and E(t), and mem-
brane potential, Vm (t); spike generation nonlinearity sigmoid function.
INIF, inhibitory population with membrane potential Vi1(t) and modulat-
ing inputs: excitatory by means of M1(t) and inhibitory by means of
M2(t), spike generation with sigmoid. INFF, excitatory population driven
by the main input P(t) inhibiting the inhibitory population INIF; INEF,
excitatory population feeding back on to the main population, spike gen-
eration sigmoid. The interconnectivity constants giving the number of
cells of one population projecting to one cell of the other population are
C1 and C2 for the inhibitory feedback and C4 and C5 for the excitatory
feedback. C3 represents the number of cells of the INFF population
excited by one input fiber (feed-forward). (Adapted from Lopes da Silva
FH, van Rotterdam A, Barts P, et al. Models of neuronal populations: the
basic mechanisms of rhythmicity. In: Corner MA, Swaab DF, eds.
Perspectives of Brain Research. vol. 45. Prog. Brain Res. 1976:281–308.)
Figure 4.10 Block diagram of the lumped model for rhythmic activity of
simplified alpha rhythm model. The thalamocortical relay (TCR) neurons
are represented by two linear systems having impulse responses simulat-
ing an EPSP (h c(t)) and an IPSP (h i(t)), respectively, together with the
static nonlinearity (fE(V)) representing the spike-generating process. The
interneurons (IN cells) are represented by one linear system (h e(t)) and a
nonlinearity (fI(V)). C1 represents the number of INs to which one TCR
neuron projects; C2 represents the number of TCRs to which one IN proj-
ects. (From Lopes da Silva FH, Hoeks A, Smits H, et al. Model of brain
rhythmic activity. Kybernetik. 1974;15:27–37, with permission.)
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
Figure 4.11 Power spectra of the output of the lumped alpha rhythm
model shown in Figure 4.10 for different values of the feedback gain
(K), which is mainly determined by the coupling constants representing
the synaptic interactions within the neuronal set. As the synaptic inter-
actions become stronger and K increases, the spectrum acquires a clear
selectivity at the alpha frequency. (Adapted from Lopes da Silva FH,
Hocks A, Smits H, et al. Model of brain rhythmic activity. Kybernetik.
1974;15:27–37.)
bandwidth will decrease with an increase of the coupling
constants, C2 and C1, that is, with an increase in synaptic inter-
actions within the neuronal cartel.
The linear analysis of the preceding section gives only a first
approximation to the spectral characteristics of the alpha
rhythm, both experimentally measured and simulated. The spec-
trum of the experimentally measured alpha rhythm also presents
a clear component at the second harmonic of the peak frequency
(Fig. 4.12). It can be demonstrated by means of bispectral analy-
sis that this component is indeed a second harmonic of the peak
frequency. This fact implies that static nonlinear properties of the
network have also to be taken into account.
Continuous Neural Field Model of Alpha Rhythm
The spatial properties of alpha rhythms are less well under-
stood, although it has been found experimentally that, in the
visual cortex of the dog, small cortical areas appear to act as
“epicenters” from which alpha rhythm activity “spreads” in dif-
ferent directions (21,179). The spreading has been measured in
75
Figure 4.12 An epoch of alpha rhythm (A) experimentally recorded
from the visual cortex of the dog (technique as used in Ref. 86). The
power spectrum (B) and the bicoherence (C) were obtained by ensem-
ble averaging. Note the peak in the power spectrum at about 10 Hz and
the smaller peak at the second harmonic frequency, which are related
to each other as the bicoherence shows. A: Experimentally recorded
alpha rhythm (arbitrary scale). B: Power spectrum (arbitrary scale). C:
Bicoherence limits 99% at 0.119, 95% at 0.096, five lines from 0.18
until 0.26. (Adapted from Lopes da Silva FH, van Rotterdam A, Barts P,
et al. Models of neuronal populations: the basic mechanisms of rhyth-
micity. In: Corner MA, Swaab DF, eds. Perspectives of Brain Research.
vol. 45. Prog. Brain Res. 1976:281–308.)
terms of phase differences in rhythmic components at different
locations on the cortical surface.
One hypothesis is that the propagation of alpha activity
depends on spatial properties of the cortical neuronal net-
works; the strength of inhibition is a function of distance
depending on the length of interneuronal connections. This
hypothesis has been realized by way of a “spatially distributed
model” of neurons and interneurons (20), which is an exten-
sion of the lumped parameter model described by Lopes da
Silva et al. (10). In this spatially distributed model, it is assumed
that two populations of neurons (main cells and local interneu-
rons) are arranged in a one-dimensional row and are intercon-
nected by means of recurrent collaterals and inhibitory fibers.
Such a neuronal chain is assumed to have the following proper-
ties: (a) the transfer function of input-output spike densities are
second-order linear functions with a longer time constant for
inhibition than for excitation; (b) the interconnectivity properties
76 Part I ■ Basic Principles

Figure 4.13 Transfer properties of the spatially distributed model or

neuronal chain. Left: The time courses of the simulated IPSP and EPSP Figure 4.14 Schematic drawing of a section of the spatially distributed
are shown with the corresponding equations. Right: The spatial connec- model (neuronal chain). For explanation see text. (Adapted with per-
tivity functions are shown. Ce(x) and Di(x) describe the homogeneous mission from van Rotterdam A, Lopes da Silva FH, Van den Ende J, et
interconnectivity functions for the excitatory and the inhibitory connec- al. A model of the spatial-temporal characteristics of the alpha rhythm.
tions, which depend only on distance x. The values of the characteris- Bull Math Biol. 1982;44(2):283–305.)
tic lengths correspond to l/ c and l/ d. The numerical values of c and d
in the figure are only indicative. (Adapted from van Rotterdam A, Lopes
da Silva FH, Van den Ende J, et al. A model of the spatial-temporal char- model may be written as a linear inhomogeneous partial differ-
acteristics of the alpha rhythm. Bull Math Biol. 1982;44(2):283–305.) ential equation, which is a rather complicated type of wave
equation. By assuming a wave packet existing over a restricted
length somewhere in the neuronal chain at some instant of
are homogeneous in space, that is, the strength of inhibition is time, one can try to determine group velocity and dispersion
only a function of distance between the interconnected neu- properties. Alternatively, the neuronal chain may be conceived
rons; and (c) the strength of interconnectivity decreases expo- as a two-dimensional linear system with a specified transfer
nentially as a function of distance, as shown in Figure 4.13. function relating the input p(x,t) to the output u(x,t). The
To simulate alpha activity as a function of both space and response to an arbitrary input can then be obtained in the time
time, the lumped model must be extended in such a way that it domain by means of a two-dimensional convolution or in the
can account for spatial properties. This can be done as follows:
the spatially distributed model consists of a chain of neurons
and interneurons interconnected by means of spatially distrib-
uted excitatory collaterals and inhibitory fibers from the
interneurons. The interconnections are described by Ce(k,l),
denoting the number of excitatory synapses from the kth main
cell projecting on the lth interneuron, and Ci(k,l), denoting the
number of inhibitory synapses from the kth interneuron on the
lth main cell. Figure 4.14 shows a schematic drawing of such a
neuronal chain, where T k is the kth main cell and Ik is the kth
local interneuron. The main cells receive input along afferent
fibers from outside the chain. Figure 4.15 shows the correspon-
ding block diagram. The time-dependent behavior of the mem-
brane potential U(k,t) of the kth cell is determined by the
transfer functions he(t) and hi(t) of the lumped model. Nle and
Nli denote the static nonlinearities that relate membrane poten-
tials to impulse densities of main cells and interneurons,
respectively. These static nonlinearities are assumed to have the
same characteristics as fe(V e) and fi(V i) in the lumped model
and hence can be approximated as linear gain factors in the Figure 4.15 Block diagram of a section of the neural chain model
neuronal chain. Finally, P(k,t) denotes the input impulse den- showing the main excitatory and inhibitory interconnections as
sity on the main cell k. explained in text. (Adapted with permission from van Rotterdam A,
To study the propagation properties in the neuronal chain, Lopes da Silva FH, Van den Ende J, et al. A model of the spatial-tempo-
various approaches can be used. In the first place, the expres- ral characteristics of the alpha rhythm. Bull Ma th Biol.
sion describing the spatial and temporal characteristics of the 1982;44(2):283–305.)
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
two-dimensional frequency domain by multiplication of the
spectra of the transfer function and the input function (see
mathematical treatment of van Rotterdam et al. (20)).
It is reasonable to assume that, in order to simulate alpha
rhythm conditions, the neuronal chain must be excited ran-
domly in space as well as in time. Due to the linear character of
the model, however, a simple local harmonic excitation of the
chain already reveals the propagation properties of the neu-
ronal chain.
If the neuronal chain is locally excited at x = 0 with a har-
monic component of frequency 0, the model response with
the same time characteristics has a local component at x = 0 and
two components that spread in two directions along the chain.
The amplitudes of these components decay exponentially pro-
portional to exp[ Re{ki( 0)}|x|] and the phase shifts that
reflect the propagation properties of the harmonic component
equal exp[ Im{ki( 0)}|x|] radians when measured at a distance
x in relation to x = 0. Both phase shift and decay are thus
dependent on the frequency of excitation.
The results of the evaluation of the roots of the equation
describing the output of the chain as a function of space and
time (20) show that the rhythmic components existing in the
chain in the alpha band are less damped than the components
outside the alpha band. Furthermore, it follows that the damp-
ing in the alpha band is mainly determined by the interconnec-
tivity function with the largest characteristic length.
The model just described schematically has characteristics
that are in qualitative agreement with experimental data on
alpha rhythmic activity obtained from the visual cortex of dog
(21). This can be examined in relation first to coherence analy-
sis and second to experimental phase measurements:
1. Coherence analysis of EEG signals recorded from chronically
implanted subdural electrodes lying directly on the marginal
gyrus revealed a higher correlation between EEG compo-
nents in the alpha band than between components outside
that band (90,91). These experimental data support the con-
clusion from the model study that alpha rhythms are the
least damped in the chain.
2. Phase shifts at the alpha frequency (~12 Hz) measured on
the marginal gyrus in four dogs reflected phase velocities
that were about 0.3 m/sec (21). Using the model, the charac-
teristic length corresponding to this phase velocity and fre-
quency was computed. The order of magnitude of the
characteristic lengths 1/c and 1/d of the connectivity func-
tions Ce(x) and Di(x) (Fig. 4.13) was estimated and found to
be within the range of 160 to 330 m. The characteristic
lengths correspond to the distance at which the influence of
a neuron on a neighbor is reduced to 37% of the starting
value and not to the whole length of the interconnecting
fibers. The figures that were calculated using the model are
in the same range as the dimensions of the basic cortical
space module, which is considered to be a vertical cylinder of
200/300 m diameter (92–95). Different systems of intra-
cortical fibers, which involve much larger distances than the
diameter of a column (at least distances one order of magni-
tude larger), assure the interconnectivity of several columns.
77
One of these systems consists of tangential intracortical col-
laterals of the pyramidal cells. These may spread over dis-
tances of about 3 mm; they form the “pyramid collateral
spread module” (95). Another system consists of long sur-
face parallel terminal axons that may extend over distances
of about 6 to 8 mm; this forms the “surface parallel intracor-
tical system” (95). These systems of overlapping intracortical
connections will lead to the establishment of cooperative
action between large populations of cortical neurons and
thus to the appearance of dynamic patterns over the cortex
(96). In addition to the corticocortical systems described
above, one must also take into consideration the role of local
interneurons; Tombol (97) described, among others, basket
cells in the cat cortex with vertical axons branching horizon-
tally over 500 to 1,000 m, with terminals surrounding and
making contact with pyramidal cell somata; these basket
cells are probably inhibitory interneurons. The collaterals
belonging to the pyramid collateral spread module system of
Szentagothai, which is most likely excitatory, and the axons
of the inhibitory interneurons can be identified as the exci-
tatory collaterals and the inhibitory fibers modeled in the
model neuronal chain; because this system of interconnec-
tions ranges over distance in the order of magnitude of hun-
dreds of micrometers, the estimated characteristic lengths
computed from the experimentally measured alpha rhythm
frequency and phase velocity with the help of the neuronal
chain model are in agreement with neuroanatomical data.
Finally, note that the neuronal chain can be easily general-
ized to a neuronal network extending in two or three dimen-
sions. A more general approach that accounts for the
generation of alpha rhythms and its spread in the cortex was
proposed more recently by Liley et al. (98–100), who consid-
ered the cortex as an excitable spatial continuum of recipro-
cally connected excitatory and inhibitory neurons interacting
by way of short-range (intracortical) and long-range (cortico-
cortical) connections.
An essential feature of this general model is that IPSPs are
described by third-order differential equations, since according
to these authors lower orders were not able to support wide-
spread activity in the alpha band. This theoretical approach
showed that the strength and the form of synaptic interactions
between inhibitory interneurons constitute the most relevant
determinants of the frequency and damping of alpha band
oscillations. This feature derives from the assumption that local
inhibitory–inhibitory loop delays are longer than the corre-
sponding local and long-range excitatory–excitatory loop
delays. This theory has been presented in the form of a set of
nonlinear differential equations that can be solved by numeri-
cal methods, but most important predictions of the theory can
be obtained by a simplified set of linear equations. In this way
the input to the cortex can be assumed to be indistinguishable
from band-limited white noise, while the cortex operates as a
noise filter that passes preferentially frequencies around 8 to 13
Hz. The filter can be described by a transfer function, the prop-
erties of which are determined by both the physiological state
and the anatomical structure of the cortex.
78 Part I ■ Basic Principles

MODELS OF OSCILLATIONS AT THE SINGLE

CELL AND LOCAL THALAMIC NETWORK
LEVELS
In the last decades, important advances were obtained with
respect to the physiology of single cells of the thalamus and to
their patterns of interaction at the level of local networks,
including the corresponding intrinsic membrane and synaptic
properties. Interesting models incorporating these properties
were proposed, among others by Wang (44), Wang et al. (48),
Golomb et al. (42,43), Destexhe et al. (69), and have been
reviewed extensively by Destexhe and Sejnowski (83). These
models are mainly based on the experimental results of Steriade
and colleagues (101) and also of McCormick and collaborators
(74,75). The latter results were obtained in the ferret slice
preparation that facilitates carrying out controlled experi-
ments, although under circumstances that differ from the con-
ditions in the intact brain. Nevertheless, in these slices it was
possible to show (74,75) the occurrence of spindles with main
frequency between 5 and 9 Hz, during which the RE neurons
tend to fire synchronously with almost every cycle of local
rhythmic spindle. In contrast, TCR neurons tend to fire bursts
also phase-locked to the local rhythm, but not at every cycle
since they may skip one, two, or even more cycles. The modula-
tion of this activity by blocking GABAA receptors results in a
decrease of the main frequency to 2 to 4 Hz, whereas the rhyth-
mic pattern is hardly changed by blocking GABAB receptors. In
addition, the spindle bursts appear to propagate in the slice, as
traveling fronts, at a velocity of about 1 mm/sec.
As indicated above, the RE and the TCR neurons can exhibit Figure 4.16 Synaptic architecture of the model of Golomb et al. (43).
oscillations of the membrane potential, which depend strongly A: Reticular neurons (RE) receive thalamocortical (TC) excitation via
on the state of the potassium leak conductance of the cells (gkl). glutamatergic (AMPA) synapses and intra-RE inhibition via GABAA and
In the model of Golomb et al. (43) both RE and TCR neurons GABAB synapses from other RE cells. B: This model has a one-dimen-
are at rest, that is, they do not display oscillatory behavior, for sional architecture. The RE-to-TC coupling strength decays with dis-
too small or too large values of gkl and they oscillate at interme- tance. The typical decay length RT is called the synaptic footprint length.
diary levels. In this model the TCR neurons elicit EPSPs on RE The same applies to the TC-to-RE coupling (TR) and to the RE-to-RE
neurons by way of glutamatergic (AMPA) synapses, whereas RE coupling (RR). C: A typical footprint shape of the synaptic coupling with
neurons elicit both GABAA- and GABAB-mediated IPSPs on distance; the decay length (either TR or RT or RR) is the distance for
TCR neurons. Due to the existence of a low-voltage activated which the synaptic coupling reaches 1/ e of its maximal value. (Adapted
Ca current (T current) the TCR neurons can elicit a rebound from Golomb D, Wang XJ, Rinzel J. Propagation of spindle waves in a
burst after a hyperpolarization due to the GABAergic IPSP, thalamic slice model. J Neurophysiol. 1996;75:750–769.)
whether or not associated with a Ca-dependent K current
(afterhyperpolarization [AHP] current). In the model simula-
tions if the neurons were not in intrinsic oscillatory mode, the
network would be quiescent. However, the introduction of a oscillations, whereas an up-regulation of Ih elicited refractori-
few oscillating TCR neurons was sufficient to recruit the other ness of the neurons, leading to extinction of the oscillations. In
neurons into a rhythmic spindle. The oscillation was critically this model, the responses of the neuronal network to different
dependent on the existence of a cationic current activated by stimulus patterns were not reported, although this is most
hyperpolarization, the Ih or “sag” current, in these neurons (44). important to understand the behavior of the network of the
Using this model it was possible to investigate the patterns of intact brain under physiological conditions.
propagation of the oscillations throughout the network. The
neurons in the network were connected by synaptic couplings, MODELS OF EPILEPTIFORM ACTIVITY
the strength of which decays with spatial distance, as shown in
Figure 4.16. An oscillation started typically at a focal site and Here we consider those models by means of which one attempts
propagated as a wavefront that recruited successively more neu- to explain how EEG phenomena characteristic of epileptiform
rons into the oscillatory mode. The Ih current modulated the seizures may emerge from the normal ongoing activity. Most
occurrence of these oscillations: a reduction of Ih led to sustained theoretical studies have addressed in general terms the question
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
of how oscillations may take place in a network of neurons,
assuming that such oscillations are akin to seizure activity. A
general way to investigate this problem using a model of such a
network is to analyze the trajectories in the phase plane whose
coordinates are the potential and its first derivative. A nonlinear
system that goes into stable oscillation even after the input
stimulus has ceased will produce a closed trajectory in the phase
plane; such a closed trajectory is called a limit cycle. Working
from the model initially described by Lopes da Silva et al.
(10,11), Zetterberg et al. (102) have shown that a slightly mod-
ified model may generate limit cycle oscillations that can simu-
late some forms of epileptogenic activity recorded from the
cortex. Freeman noted that his KII model was not initially
designed to simulate epileptiform activity. However, by using a
different nonlinear gain curve, Babloyantz and Kaczmarek
(103) have generated limit cycle oscillations that are thought to
resemble epileptiform activity. In such models, however, a more
or less precise fit between simulated and recorded EEG activity
was not attempted. More recently, models that attempt a precise
waveform fit have been introduced. Such a model was proposed
by Traub (52) and Traub and Wong (104) and is based on
experimental data at the cellular level concerning the mem-
brane mechanisms responsible for epileptogenesis (53) together
with the properties of synaptic interactions in hippocampal
neurons.
We should note that in recent years the interest for NMMs
with respect to epileptic phenomena has steadily increased, as
illustrated below, most likely due to the fact that epileptic phe-
nomena involve brain systems with complex dynamics that are
difficult to comprehend exclusively on the basis of knowledge at
the cellular level.
Cellular and Network Models
of Epileptiform Activity
Observations at the cellular level obtained in the hippocampal
slice preparation have yielded a number of new facts concern-
ing the origin of an epileptogenic focus. It has been shown that
pyramidal cells of the hippocampus (cornus ammonis or CA3)
have active membrane regions, the so-called soma and den-
dritic hot spots that have voltage-dependent ion conductances
(g) for Na , K , and Ca2 and a calcium-activated K conduc-
tance; furthermore, a voltage-dependent inactivation of gK and
a partial inactivation of gCa by the cytosolic accumulation of
Ca2 ions has been assumed (52). Incorporating all these
properties, Traub (52) has made a multicompartmental model
by means of which spontaneous bursts in CA3 cells can be gen-
erated, as shown in Figure 4.17. Using these elementary cellular
components (105), Traub et al. (54) constructed a network of
CA3 neurons interconnected by excitatory synaptic connec-
tions. The generation of epileptiform transients induced by the
administration of antagonists of the neurotransmitter GABA
was the first problem considered. The conditions for the pro-
duction of a chain reaction of excitation that spreads over the
whole network were found using the simulations. The most
important of these conditions was that the excitatory intercon-
nectivity should be sufficiently dense and the synaptic interac-
tions of sufficient strength. A good fit between simulated and
79
Figure 4.17 Simulation of a bursting neuron showing an epileptiform
discharge. Each simulated neuron had a number of compartments and
ion conductances, as explained in the text. The intracellular burst
model is shown above a recorded burst. The network contains 100 neu-
rons interconnected in a random synaptic network. Vertical calibration,
25 mV by 25 msec (model) and 40 msec (experimental). (Adapted from
Traub RD, Wong RKS. Cellular mechanism of neuronal synchronization
in epilepsy. Science. 1982;216:745–747.)
experimentally recorded epileptiform transients was obtained,
as shown in Figure 4.17. A second problem analyzed was how
sustained afterdischarges can be produced in such a model. It
was found that for the occurrence of repetitive afterdischarges,
some form of pacing inhibitory process was needed. This was
accomplished by introducing two additions in the model: (a) a
slow voltage-dependent K conductance (M current) and (b)
an axonal refractoriness that causes intermittent conduction
from one to the next depolarized cell, located at the axonal ini-
tial segment, at branch points, or at presynaptic terminals.
Furthermore, a number of loops allowing reentrant paths
within the neuronal population are of importance for the main-
tenance of a series of afterdischarges. Using this kind of models,
in a series of studies Traub and collaborators (57,106) investi-
gated the role of a number of factors in inducing epileptiform
oscillations similar to those observed in brain slices, in vitro,
namely synaptic mediated activation of N-methyl- D-aspartate
(NMDA) dendritic synapses and intrinsic voltage-dependent
conductances. The possible importance of the firing of ectopic
presynaptic action potentials in eliciting the synchronization of
neurons during epileptiform seizure-like activity was suggested
(57). Another factor that may contribute to epileptiform activity
has been put forward on the basis of models and in vitro physi-
ological studies of Traub’s group (55): gap junctions located
between the axons of principal hippocampal neurons would
play a role in the generation of very fast (about 200 Hz) oscilla-
tions. These model simulations have allowed making the predic-
tion that action potentials can cross from axon to axon via gap
junctions. Axonal network oscillations would, in turn, induce
oscillatory activity in larger neuronal networks, by a variety of
mechanisms, and could underlie in vivo ripples (to ~200 Hz in
hippocampal EEG) to drive gamma (30 to 70 Hz) oscillations
that appear in the presence of carbachol and to initiate certain
types of ictal discharges. A comprehensive review of the work of
Traub and collaborators can be found in Traub et al. (107).
80 Part I ■ Basic Principles
Along the same line high-frequency ripples may be determined
by high-frequency inhibitory activity, the breakdown of which
may enable seizure activity to propagate from an initial focal
area (108,109).
In addition, model studies have also addressed the possibil-
ity that changes in extracellular ionic concentrations (increase
of K and/or decrease of Ca2 ) may be important for the gen-
eration of epileptiform seizure activity. Somjen and collabora-
tors (110–112) used computer models to better understand
how transmembrane ionic and water movements may affect
neuronal excitability and observed that the model produced
epileptiform afterdischarges if during stimulation interstitial
K concentration was sufficiently enhanced so that the persist-
ent Na (slowly inactivating) current was activated. In order for
firing of a neuron to be maintained after the end of a stimulus,
the [K o]/[K i] ratio had to be sufficiently large to maintain
the neuronal membrane depolarized so that the inward cur-
rents could persist, which activated a series of action potentials.
In the course of time the persistent Na current was inactivated
and the ion pumps, associated with the buffering of K ions by
glia cells, restored the normal levels of K extracellular concen-
tration. These model studies strongly suggest that the elevation
of extracellular K is not the initial stimulus for a seizure, but it
may cause the transition from interictal to ictal discharges
(113). The initial stimulus starting the seizure may require
synaptic interactions among a population of interconnected
neurons.
In contrast with the models described above, which are
directed to the simulation of epileptiform activity as seen at the
cellular level, especially in in vitro preparations, other models
are based on a neural mass approach using mean-field concepts.
A paramount example of these is the model study of Wendling
et al. (19) that focuses on high-frequency EEG activities
(gamma band) that constitute one of the most characteristic
EEG patterns in focal seizures of human epilepsy, especially
those that have an epileptogenic zone in the hippocampus and
associated limbic cortical areas (114,115). This model starts
from the hypothesis that networks of inhibitory interneurons
with different properties are essential for the generation of
gamma frequency oscillations (116). As already pointed out
above, this lumped model includes different kinds of neuronal
populations, in particular two kinds of GABAergic neurons:
GABAA fast and GABAA slow populations, where the former
cause IPSPs with short time constants on the soma of pyrami-
dal neurons, while the latter cause IPSPs with slow time con-
stants on the dendrites of these neurons. Furthermore as
demonstrated experimentally by Banks et al. (117), both classes
interact: GABAA slow cells inhibit not only pyramidal cells but
also GABAA fast interneurons. The model is based on the
important experimental observation that a nonuniform alter-
ation of GABAergic inhibition underlies the generation of
epileptiform activity in hippocampus and associated areas of
the limbic brain, namely a reduced dendritic inhibition and
increased somatic inhibition (118). The simulations show that
strikingly realistic activities are produced by the model when
compared to real EEG signals recorded with intracerebral elec-
trodes in patients with temporal lobe epilepsy. They show,
further, that the transition between interictal and fast ictal
activity is explained, in the model, by the impairment of den-
dritic inhibition (Fig. 4.18).
Also using NMMs Kramer et al. (119–121) constructed
models of human cortical networks that can generate seizure-
like activity, as the result of increased connectivity between
excitatory and inhibitory cell populations or of decreased con-
nectivity within either excitatory or inhibitory cell populations.
These studies also allowed to test the possibility of controlling
seizures (120) by means of feedback controllers of the model
dynamics.
In contrast with NMMs where the topology of the network
is not taken into account or is much simplified, topological mod-
els there have been developed, that is, graphical representation
models where topological details of the neuronal networks are
explicitly included. These models permit to simulate the effects
of histological changes that may occur in an epileptogenic tis-
sue, such as the death of certain cellular types and the sprout-
ing of new axonal or dendritic connections. Thus Lytton et al.
(122) developed a network model of the dentate gyrus of the
hippocampus including granule cells, inhibitory aspiny
interneurons, and excitatory mossy cells. According to this
model mossy fibers sprouting and disinhibition were necessary
for repetitive granule cell firing, alike to EEG seizure activity, to
occur. More elaborate topological models of similar kind were
developed by Santhakumar et al. (123) and Dyhrfjeld-Johnsen
et al. (124). The former showed that if sprouting of mossy fibers
exceeds a certain value, seizure activity can occur regardless of
the number of hilar cells lost. The latter model study, including
more than one million cells, simulated the topology of the den-
tate gyrus. This can display globally and locally connected
(“small world”) architecture. The effect of changes of network
topology during epileptogenesis was modeled on the one hand
by removing long-distance projecting hilar cells which causes a
reduction in the connectivity and, on the other, by sprouting of
granule cell axons which has the opposite effect of increasing
local connectivity. The model simulations demonstrated that
the survival of even a small fraction of hilar cells was enough to
sustain hyperexcitability in the network. This is consistent with
experimental data since in animal models of hippocampal scle-
rosis the loss of hilar cells is never 100%; in our study (125) it
was about 66% to 74% in rats with a progressive form of epilep-
tic seizures. Also in brain tissue extracted from patients with
hippocampal sclerosis about 20% of mossy cells have been esti-
mated to survive (126), which corresponds closely with the
finding in the model of Dyhrfjeld-Johnsen et al. (124) that
maximal epileptiform activity was seen at a level of about 80%
sclerosis.
In general we may state that a variety of changes at the cellu-
lar level may lead to the occurrence of epileptiform seizure
activity. In fact there is no unique process that can account for
the generation of epileptiform activity (127). A generalized
concept is that seizures occur due to a shift from a dynamical
unbalance between excitatory and inhibitory processes with a
predominance of the former in strongly coupled networks. Van
Drongelen et al. (128) showed, however, in a detailed computa-
tional model that seizure-like activity may occur when the
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
excitatory synapses are weakened. The model predictions were
tested experimentally and the authors showed that a pharmaco-
logical reduction of excitatory synaptic transmission could lead
to repetitive network bursting.
Models of Thalamic Spike-and-Wave Seizures
Thalamic neurons can display different modes of activity
including rhythmic bursting that are akin to spike-and-wave
(SW) discharges characteristic of generalized epilepsies. In this
81
Figure 4.18 A: A real depth-EEG signal
recorded in human hippocampus at the begin-
ning of a partial temporal lobe seizure with
intracerebral electrodes (SEEG). Four phases (a-1
to a-4) are distinguished according to the pseudo-
stationary nature of the activities reflected by the
signal: normal background activity (a-1), dis-
charge of rhythmic spikes (a-2), low-voltage
rapid discharge (a-3), and high-amplitude quasi-
sinusoidal slowing down activity (a-4). The nor-
malized power spectral density (PSD) shows that
part of the activity corresponding to phase a-3
belongs to the gamma band (55 to 60 Hz). B: A
candidate path on the activity map, for a value of
excitation (A = 5) and different values of slow
(B) and fast (G) inhibition, showing the transi-
tions between different types of activity observed
in the real signal (type 1 → type 3 → type 4 →
type 6). C: Slow and fast inhibitory synaptic gain
profiles defined by the candidate path and used
in the model to simulate a timeseries signal with
transitions in dynamics. D: Simulated EEG signal
obtained with the model when the synaptic gains
vary as a function of time and PSDs computed on
the four periods of 10 seconds (b-1 to b-4).
(Adapted from Wendling F, Bartolomei F,
Bellanger JJ, et al. Epileptic fast activity can be
explained by a model of impaired GABAergic
dendritic inhibition. Eur J Neurosci.
2002;15(9):1499–1508.)
respect the single neuron model of a thalamic relay neuron of
Wang (44), which shows both 10- and 3-Hz bursting dynamical
modes, is particularly instructive. This model demonstrates
clearly how distinct oscillatory modes depend on different bal-
ances of sets of ion currents. In this model the emergence of
3-Hz bursts, which resemble the 3-Hz oscillations seen in gen-
eralized epilepsies, depends on the value of the hyperpolarization-
activated cationic current, Ih the so-called sag current, although
the T-type Ca current also plays an important role. According
82 Part I ■ Basic Principles
to this model the 10-Hz oscillatory mode is maintained, at the
level of a single neuron, by the dynamics of the low-voltage-
activated Ca current, IT. Under normal conditions the neuronal
membrane does not hyperpolarize sufficiently to activate the Ih
current.
Indeed if the neuronal membrane would be deeply hyperpo-
larized, it would stay in such a condition in case only IT would
be active, because the latter current is not sufficient to overcome
a strong hyperpolarization. However, the presence of Ih allows
the neuron to escape this strong hyperpolarized state. Thus the
transition between the 10- and the 3-Hz modes depends on the
level of hyperpolarization. This single neuron model is com-
pletely deterministic and it may display “strange attractors” typ-
ical of a chaotic dynamic state.
A model of a thalamic neuronal network that is closer to the
reality of epileptic discharges than the single neuron model
described above was developed to account for SW seizures at
the level of a neuronal population. The basic physiological phe-
nomenon simulated was the finding that during SW discharges
the RE neurons discharge with long spike bursts riding on a
depolarization, whereas the TCR neurons are either entrained
into the seizure oscillation or are quiescent (129). The spatial
model of these neuronal populations predicts that in a center of
seizure activity there will be intense RE activity and TCR quies-
cence and that this focus will be surrounded by an area where
the TCR neurons will be less hyperpolarized such that low-
threshold oscillatory spikes will occur. This surrounding area
will be the forefront of a wave of propagating seizure activity.
The model shows that the complex nonlinear dynamics of the
neuronal network depends critically on the low-voltage-acti-
vated (LVA) Ca-current IT. In this context it is interesting to
note that in the genetic model of absence epilepsy GAERS
(genetic absence epilepsy rats of Strasbourg), the presence of a
significantly enhanced LVA current in thalamic neurons was
demonstrated by Tsakiridou et al. (130). It is also noteworthy
that drugs, like ethosuximide, that depress IT should suppress
seizure initiation. It should be noted, however, that more recent
studies showed that ethosuximide acts also on the noninactivat-
ing Na current and on a Ca2 -activated K current (131). The
level of hyperpolarization of TCR cells, which appears to be a
crucial determinant of the dynamic behavior of the neuronal
population, depends on synaptic mechanisms. The model of
Destexhe et al. (69) goes further than previous models and
shows that the transition from 9–11 Hz to 3 Hz oscillations
appears to depend on GABAB synapses. Relatively weak RE
activity would mainly elicit GABAA responses, whereas stronger
activity would recruit GABAB synapses and this would cause the
quiescent mode of the TCR neurons. Indeed, GABAB presynap-
tic inhibition is conspicuous between RE neurons, such that an
enhancement of the latter would result in a decrease of intra-RE
inhibition and, thus in an increase of inhibition of TCR neu-
rons (quiescent state). Thus GABAergic drugs would aggravate
the tendency for seizure oscillations to occur.
The models discussed above have given insight into some
basic neuronal mechanisms of SW discharges but do not
address specifically the most essential issue of this type of
epileptic activity—that a given thalamocortical loop can display
both kinds of activity without specific adjustments of parame-
ters being expressly made. Indeed the essence of epilepsy is that
a patient displays (long) periods of normal EEG activity (i.e.,
nonepileptiform) intermingled with epileptiform paroxysmal
activity only occasionally. Thus, it is essential to understand the
mechanisms responsible for transitions from normal activity to
paroxysmal SW discharges. This aspect of the dynamical
process responsible for epilepsy was addressed in a computa-
tional model of absence epilepsy (50). In this model the behav-
ior of populations of interacting neurons integrating neuronal
and network properties was simulated, as a more extended and
elaborated version of a previous model of the alpha rhythm
(132). This approach facilitated investigating system dynamics
at the macroscopic level, that is, at the level where electric brain
signals such as local field potentials or EEG are recorded. The
model may exhibit two qualitatively different types of behavior,
such as seen in experimental animals with genetically deter-
mined absence-like seizure (WAG/Rij rats and GAERS ani-
mals). The output signal may display a waxing and waning
“spindle-like” oscillation having a spectrum with a peak at
approximately 11 Hz or a high-amplitude “seizure-like” oscilla-
tion at a frequency around 9 Hz, as seen in these rats during
absence seizures (Fig. 4.19). The former behavior corresponds
to the normal ongoing activity, while the latter corresponds to
the paroxysmal epileptiform activity. Thus, for a set of neuronal
parameters, the model is in a “bistable regime” where it may
generate both normal and paroxysmal oscillations and sponta-
neous transitions between these two types of behavior. This
model facilitated making a number of important predictions
that can be tested experimentally, both in animals and in
Figure 4.19 Upper panel: 20 seconds of a simulation with the occur-
rence of a spontaneous paroxysmal episode. Lower panels: Average
power spectra of signals simulated by the model and recorded from a
WAG/ Rij rat with absence seizures. Left-hand side: The spectra of nor-
mal activity. Right-hand side: The spectra of paroxysmal activity.
(Adapted from Suffczynski P, Kalitzin S, Lopes da Silva FH. Dynamics
of non-convulsive epileptic phenomena modeled by a bistable neuronal
network. Neuroscience. 2004;126(2):467–484.)
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
humans: (a) The transitions between the normal and the parox-
ysmal states can emerge spontaneously and are not induced
necessarily by any parameter changes. (b) Probabilities of tran-
sitions between normal neuronal activity and paroxysmal oscil-
lations depend on a number of model parameters; therefore,
these probabilities can be controlled in a nonunique way; in
more general terms these probabilities may be described by
gamma distributions, of which the exponential distribution is a
special case. Interestingly, exponential shape of histograms of
durations of ictal events has also been found in other types of
human epilepsy. (c) Paroxysmal oscillations can be annihilated
by a well-timed pulse, such that, under given conditions, this
kind of seizures may be aborted. The model facilitates investi-
gating these conditions such that the possibility of aborting
seizures may be tested experimentally. Another model of
absence seizures gives special attention to the time delays of
thalamocortical and corticothalamic pathways on the frequency
of SW discharges (133). Although these time delays likely con-
tribute to the frequency of the seizure oscillations, it is difficult
to deduce from this model what is the role of these delays in this
process, since this model does not include other important cel-
lular elements that may also affect the oscillation frequency,
such as GABA synaptic processes with different time constants
and calcium T-channels with particular kinetics. Further it does
not account for nonstationary changes in apparent time delays
between cortex and thalamus that have been shown experimen-
tally to occur during one episode of SW discharge (134). For a
comprehensive review of computer modeling in epilepsy the
reader is referred to Lytton (135).
LARGE-SCALE THALAMOCORTICAL MODELS
As stated in the introductory remarks brain rhythms and EEG-
like phenomena have also been investigated in large-scale mod-
els including detailed neuronal properties by the group of
Edelman et al. In the context of the present discussion it is par-
ticularly interesting to note that in a large-scale thalamocortical
model rhythmic activities can emerge characterized by frequen-
cies typical of EEG/MEG rhythms, such as delta, alpha, beta,
and gamma, although these activities were not explicitly built in
the neuronal elements (4).
In this model it was shown that rhythms at specific fre-
quencies occur in different cortical areas, for example, a beta
rhythm at about 20 Hz in areas corresponding to motor and
somatomotor cortex, while the microcircuitry of different
cortical areas was the same. Therefore these authors conclude
that these differences in frequency of oscillations should be
determined by differences in white-matter anatomical con-
nectivity (the latter was modeled based on diffusion tensor
imaging data of the human brain) between and within corti-
cal areas. Multiple propagating waves were also generated by
the model within limited cortical regions with typical propa-
gation velocities around 0.1 m/sec. We noted above that the
propagation speed of alpha waves on the surface of the dog’s
cortex was in the range of 0.3 m/sec also within limited corti-
cal areas, and the model of van Rotterdam et al. (20) revealed
that this propagation is mainly determined by the intercon-
83
nectivity function among chains of cortical neurons; although
these values are in the same order of magnitude, many factors
may account for the difference between them, such as the dif-
ferences in anatomical connectivity between human and
canine’s cortex.
MODELS OF EEG GENERATION INVOLVING
CORTICAL AND SUBCORTICAL AREAS
The models presented in the previous section describe only a
rather small number of EEG phenomena. There are many EEG
patterns that have not yet been modeled in terms of sets of neu-
ronal networks although some NMMs have been shown to gen-
erate EEG-like oscillations at different frequencies. Moreover,
most models have been limited to simulations of the behavior
of local neural networks. This chapter has already considered a
model that involved a KIII set including networks of different
areas. However, several EEG phenomena have started to be ana-
lyzed in these terms. For example, sleep EEG patterns are
known in some detail in terms of cellular processes in a num-
ber of brainstem and cortical regions (136–138; see also
Chapter 3), and a model describing many essential features of
slow-wave sleep and activated states in the thalamocortical sys-
tem as well as the transition between them was described (139).
Specific models that address the question of how pathological
slow oscillations (140) may be generated have not yet been
developed. It is to be expected that efforts in this direction will
lead to a better understanding of such EEG phenomena by
combining physiological data at the cellular level with mathe-
matical modeling.
Another important theoretical aspect of the generation of
EEG phenomena is the question of relating the temporal and
spatial aspects of such phenomena. Most models are concerned
with local networks, but not with the question of the coupling
between relatively distant sets of neurons (e.g., between the
thalamus and the cortex). However, a model aimed at analyzing
such a problem concerning the spatial organization of alpha
rhythms has been introduced by Lopes da Silva et al. (90,91). A
number of experimental studies of alpha rhythm in the dog
have revealed that intracortical coherences are in general larger
than any thalamocortical coherence measured in the same ani-
mal. This led investigators to question the amount of influence
of thalamic sources on the large intracortical coherences. This
amounts to choosing between two models: either thalamocor-
tical relations are determined by point-to-point projections or
each thalamic population may project to several cortical areas.
Using partial coherence analysis (91), it has been shown that a
model based on point-to-point projections cannot be accepted
to explain the organization of thalamocortical alpha rhythms.
Furthermore, it was found that there are widespread influences
of populations of the dorsal thalamus on intracortical coher-
ences. Notwithstanding these influences, the residual intracor-
tical coherences were so large that other factors must be
assumed to explain this intracortical coherent activity; these
factors may have as an anatomical substrate inter- and intra-
cortical systems of fibers responsible for the spatial distribution
of alpha rhythms within the cortex. Indeed, if such systems did
84 Part I ■ Basic Principles
not exist, investigators would be left with a patchy nonoverlap-
ping assembly of cortical modules with little intracortical
coherence; it would be almost impossible in such a case to
record an EEG at a relatively large distance from the cortex.
A similar problem has been mathematically elaborated by
Nunez (141), who developed a model, also based on the exis-
tence of interconnections between different cortical modules,
to account for the global wave pattern of the EEG at the scalp.
In this way, Nunez takes into account that traveling EEG waves
may coalesce to form standing waves because, he assumes, the
cortex forms a closed surface. In his view, there is a fixed rela-
tionship between the nondimensional temporal and spatial fre-
quency of EEG given by the dispersion relation. This relation
relates the temporal frequency with the characteristic velocity
of propagation of activity in associative cortical fibers and the
so-called wave number.
Nunez (142) discusses the theory proposed by Lopes da Silva
et al. (91) and van Rotterdam et al. (20), which he calls the
“local theory” of the electrocorticogram, in connection with his
“global theory,” which predicts traveling and standing waves
with wavelengths up to the cortical circumference having phase
velocities in the 10 m/sec range. According to the local theory,
alpha activity spreads away from multiple epicenters with a pre-
ferred wave number range for alpha activity and with phase
velocities of about 0.3 m/sec. It is likely that a combined
local/global theory is needed. Indeed, as Nunez points out
(35,142), the simultaneous existence of both short- and long-
wavelength phenomena in the same frequency range can occur.
Some EEG phenomena may be dominated by short-range
intracortical interactions, while others are dominated by long-
range corticocortical interactions. The human scalp EEG results
probably from both types of activity. In the combined local-
global theory of Nunez (35), the input to the cortex consists of
the modulating afferent input from subcortical areas and inputs
from other cortical regions, whether distant or neighboring,
through corticocortical fibers. In this model, it is assumed that
the local networks include inhibitory and excitatory feedback
circuits, whereas the global corticocortical interactions are exci-
tatory. Nunez and collaborators (143,144) proposed a theoreti-
cal framework supporting experimental measures of dynamic
properties of human EEG. According to this model both glob-
ally coherent and locally dominated behavior can occur within
the alpha band, depending on narrow band frequency, spatial
measurement scale, and brain state. At the same time, alpha and
theta phase locking between cortical regions during certain
behavioral states can occur.
MODELS OF CORTICAL BETA/ GAMMA
RHYTHMS
In the cortex fast rhythms have been identified, both in the EEG
and at the level of single neuronal activity, that are usually
called beta or gamma rhythms, depending on frequency range.
It has been found experimentally that fast frequency rhythms
(33–45 Hz) occurring in the frontoparietal areas may occupy a
restricted cortical area in the awake cat (145). These thalamo-
cortical beta rhythms are characteristic of attentive immobility.
Beta rhythms with a different frequency (~28 Hz) have been
shown to be rather localized to occipital areas in dogs engaged
in a visual discrimination task (146). Particularly in the visual
cortex, oscillations in the gamma frequency range ( 30 Hz) at
the single neuronal level (147) have been recorded (reviewed in
Ref. 148). In this way, synchronization occurs between neurons
that represent related features of an object that should be inte-
grated for the generation of a coherent visual percept.
Furthermore, it was shown that neurons in other areas of the
cerebral cortex may also synchronize their activity with those of
the visual cortex, remarkably without time lags; this synchrony
was particularly strong between areas subserving related func-
tions (15).
A large number of studies have demonstrated the existence
of high-frequency rhythmic activities in the human EEG under
different circumstances, particularly in relation with motor and
cognitive functions (149–152). A review of clinically relevant
beta rhythms is given in Chapter 37 and of the influence of
drugs on these EEG activities in Chapter 43. Model studies of
these EEG fast rhythms are being developed. The demonstra-
tion of these high-frequency activities in animals, particularly
in the hippocampus and associated areas, has inspired a num-
ber of model studies that try to account for the generation of
these activities, integrating experimental observations and
computer simulations. Such a model to account for the occur-
rence of fast oscillations in a neuronal network was proposed by
Traub et al. (57) and Jefferys et al. (58,59). The model consists
of a chain of interconnected excitatory pyramidal and
inhibitory interneurons. According to this model, when the
excitatory drive on the interneurons reaches a level sufficient to
induce pairs of action potentials at short intervals (“spike dou-
blets”) the network generates oscillations in the gamma fre-
quency range on a millisecond time scale from one end of the
chain to the other. It should be noted, however, that neuronal
oscillations in a similar frequency range have been found in
other areas of the brain, mainly in olfactory bulb and cortex (1,
33,153,154), but also in the hippocampus and other limbic
areas (for review, see Ref. 155). These rhythmic activities have
been modeled by means of cortical neuronal networks with
recurrent inhibitory connections, and it was shown that such
models, with the relevant physiological parameters are capable
of accounting for the 20- to 50-Hz gamma rhythms (see discus-
sion on this topic in Refs. 58, 59, 155, and 156).
Based on experimental observations in brain slice prepara-
tions of the hippocampus or the neocortex, in vitro, where
oscillations are induced by electrical stimulation and/or by
exogenous drug application, a series of models was created as
reviewed by Whittington et al. (157). These models demon-
strate that fast gamma rhythms can be generated by the mech-
anism of mutual inhibition in populations of interconnected
inhibitory interneurons, which can be studied independently by
pharmacological manipulations. Despite the artificial condi-
tions of these experiments, these physiological data have led to
interesting theoretical generalizations (60), as for instance the
possible role of electrical coupling via gap junctions in the gen-
eration of very fast oscillations both under normal and epilep-
tic conditions (158,159).
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
Recently, a generalized model that accounts for changes in
rhythmic activities with behavioral state, from the slower alpha
to the faster gamma and beta as arousal increases, was proposed
(160). This model is especially interesting because the same set
of underlying intrinsic and network mechanisms appears to be
responsible for transitions between the different types of
rhythms. The effects of neuromodulation on ionic conduc-
tances were investigated in a cortical circuit model, consisting
of synaptically coupled excitatory and inhibitory neurons that
support rhythmic activity in the alpha, beta, and gamma
ranges. It was shown that the effects of neuromodulation on
ionic conductances are, by themselves, sufficient to induce tran-
sitions between synchronous gamma and beta rhythms and
asynchronous alpha rhythms.
It should be noted that despite extensive search, no evidence
was found by Menon et al. (161) for globally correlated activity
in the human EEG, either in narrow (i.e., 35–45 Hz) or in broad
(i.e., 20–50 Hz) frequency bands. Instead, spatially and tempo-
rally intermittent synchronization was observed between pairs
of electrodes in 1-cm 2 regions with high variability. The distri-
bution of correlation coefficients differed substantially from
background levels at interelectrode distances of 1 and 1.4 cm
but not 2 cm or more. These findings suggest that the surface
diameters of domains of spatially correlated activity within the
gamma band in human EEG are limited to less than 2 cm. Thus,
if such gamma band spatial patterns exist in the human brain,
no existing technology would be capable of measuring them at
the scalp, and subdural electrode arrays for cortical surface
recording would have to have spacings under 5 mm. This means
that on the cortex there are well-characterized spatial domains
of rhythmic activities that do not spread all over the cortical
surface.
RELEVANCE OF NONLINEAR DYNAMIC
SYSTEMS FOR EEG GENERATION
The models of neuronal networks presented above possess non-
linear wave-pulse functions. In a number of studies of such net-
works, a simplification has been introduced; it was assumed
that this nonlinear characteristic could be approximated by a
linearized version, at least under those conditions in which the
input signals are limited to a small range of amplitude.
However, such simplification is not necessary, and it may lead
one to ignore essential nonlinear properties of the system, such
as the generation of higher harmonics (Fig. 4.12). Indeed, it is
possible to account for the behavior of neuronal networks
assuming that these networks can be described by a number of
coupled nonlinear differential equations as a function of time
and space. In general, such a set of nonlinear differential equa-
tions may be considered to provide the basis for understanding
the generation of all types of EEG activity.
A number of investigators, such as Freeman (34,178) and
Babloyantz (162), started to explore the possibilities given by
the modern mathematical field of research of nonlinear
dynamic systems in order to obtain a better understanding of
EEG phenomena. A relevant aspect for this discussion is that all
nonlinear dynamic systems with more than two degrees of
85
freedom can display unpredictable behavior over long time
scales; in other words, they can display what mathematicians
call chaotic behavior (163). Such systems can have multiple sta-
ble states governed by “equilibrium,” “limit cycle,” or “chaotic”
(“strange”) attractors. Whether the system will find itself in one
or another stable state depends on the input conditions and on
the system’s parameters.
These mathematical notions can be illustrated using the
example of the KIII set defined by Freeman (34). Such a set of
neurons (Fig. 4.1) may have a least four equilibrium states.
There is a rest state, in which the nonlinear pulse-wave function
is set to zero and corresponds to the state during deep anesthe-
sia. Another is a state in which a near sinusoidal limit cycle
occurs, such as is produced in the olfactory bulb and cortex by
drugs and by hyperthermia (1). Still another state is character-
ized by recurrent limit cycle behavior in which the olfactory
bulb depends on the level of the normal respiratory rhythm. If
slightly structurally changed, a KIII set can be led to another
state in which chaotic activity is generated. In such a case, the
output consists of long stretches, or bursts, with unpredictable
amplitudes and phases that resemble the fluctuations of the
EEG at rest. A similar behavior can be obtained by giving a low-
level random noise input to the KII set. This fact illustrates a
general question with which one is faced when modeling the
EEG: Is the ongoing EEG a manifestation of a chaotic attractor
or of an equilibrium state under random perturbation? The
mathematical techniques with which to characterize chaotic
behavior of nonlinear dynamic systems and to differentiate it
from random activity are described in detail in a number of
specialized publications (163–167). Furthermore, the problems
involved in applying these mathematical tools of nonlinear
dynamical analysis to EEG signals are discussed in a number of
publications (168–174).
In general, one can state that it has been difficult to find
unambiguous evidence for the existence of nonlinear determin-
ism in EEG signals except for those recorded during epileptic
seizures (175). The latter observations have led to the concept
that a given neuronal network may present essentially different
modes of activity, that is, it may display different dynamics,
depending on specific conditions. For some values of control
parameters, a neuronal network may change its dynamical
mode of activity; for example, instead of a random noisy state,
a limit cycle or even a chaotic attractor may occur. Under these
circumstances we say that a bifurcation has taken place. Thus a
bifurcation represents a qualitative change and depends on a set
of control parameters that define the operating regimen of the
system. We hypothesized (127) that the main difference
between a normal and an epileptic brain is that the operating
regimen of a neuronal network in the epileptic brain is much
closer to a bifurcation point than in the normal brain, leading
to a low-dimensional chaotic mode of behavior. This means
that compared to the normal brain, where the operating regi-
men is situated far from such a bifurcation point, in an epilep-
tic neuronal network the distance between operating and
bifurcation points is so small that the system may easily switch
from a stable equilibrium to a chaotic attractor, even in
response to a very weak stimulus. The latter may pass undetected.
86 Part I ■ Basic Principles
Accordingly, in the case of epilepsy the basic question is which
factors are responsible for the change in the operating point of
the involved neuronal networks (176). It should be added that
this nonlinear dynamical behavior, which we formulate here at
the macroscopic level, has also been identified at the neuronal
level, since near the transition from subthreshold to bursting
oscillations in thalamic neurons an apparently chaotic oscilla-
tion was observed. This was made evident in the model studies
of Wang (44,48).
CONCLUSION
The complexity of EEG phenomena calls for the use of mathe-
matical models and computer simulations in order to under-
stand the underlying processes of generation. In the last few
years, a number of problems in electroencephalography have
been tackled in this way. Such theoretical studies are justified if
they are combined with experimental investigations. In this
way, one may not only obtain new insights about the generation
of EEG patterns, but also formulate hypotheses to be tested
under experimental conditions.
In the course of the last decades, a shift of attention from
models describing the behavior of local neuronal networks in
the temporal domain toward models that take into account the
spatial properties of complex networks has occurred.
Furthermore, an evolution can be seen in the techniques used.
In the early phase, the approach was based mainly on the use of
linear systems analysis, but more recently new mathematical
tools for the analysis of nonlinear dynamic systems have been
introduced in this domain of theoretical neurophysiology.
Moreover, the availability of powerful computer tools opens
new possibilities for modeling both complex membrane phe-
nomena and network properties. The analysis of the dynamics
of such networks is a challenge for the theoretical neurophysi-
ologist interested in understanding EEG phenomena.
REFERENCES
1. Freeman WJ. Mass Action in the Nervous System. New York:
Academic Press; 1975.
2. Hammond PH. Feedback Theory and Its Applications. London:
English University Press; 1958.
3. Harmon LD. Problems in neural modeling. In: Reiss RF, ed. Neural
Theory and Modeling, Proceedings of the 1962 OGAI Symposium.
Stanford, CA: Stanford University Press; 1964.
4. Izhikevich EM, Edelman GM. Large-scale model of mammalian
thalamocortical systems. Proc Natl Acad Sci. 2008;105(9):
3593–3598.
5. Markram H. The blue brain project. Nat Rev Neurosci. 2006;7(2):
153–160.
6. Watts DJ, Strogatz SH. Collective dynamics of “small-world” net-
works. Nature. 1998;393(6684):440–442.
7. Yu S, Huang D, Singer W, et al. A small world of neuronal syn-
chrony. Cereb Cortex. 2008;18(12):2891–2901.
8. Stam CJ, Reijneveld JC. Graph theoretical analysis of complex net-
works in the brain. Nonlinear Biomed Phys. 2007;1(1):3.
9. Wilson HR, Cowan JD. Excitatory and inhibitory interaction in
localized populations of model neurons. Biophys J. 1972;12:1–23.
10. Lopes da Silva FH, Hoeks A, Smits H, et al. Model of brain rhyth-
mic activity. Kybernetik. 1974;15:27–37.
11. Lopes da Silva FH, van Rotterdam A, Barts P, et al. Models of neu-
ronal populations: the basic mechanisms of rhythmicity. In:
Corner MA, Swaab DF, eds. Perspectives of Brain Research, vol. 45.
Prog Brain Res. 1976:281–308.
12. Jansen BH, Rit VG. Electroencephalogram and visual evoked
potential generation in a mathematical model of coupled cortical
columns. Biol Cybern. 1995;73:357–366.
13. Suffczynski P, Kalitizin S, Pfurtscheller G, et al. Computational model
of thalamocortical networks: dynamical control of alpha rhythms in
relation to focal attention. Int J Psychophysiol. 2001;43(1):25–40.
14. David O, Friston KJ. A neural mass model for MEG/EEG: coupling
and neuronal dynamics. NeuroImage. 2003;20:1743–1755.
15. Roelfsema PR, Engel AK, Koenig P, et al. Visuomotor integration is
associated with zero timelag synchronization among cortical areas.
Nature. 1997;385:157–161.
16. Chawla D, Friston KJ, Lumer ED. Zero-lag synchronous dynamics
in triplets of interconnected cortical areas. Neural Netw.
2001;14:727–735.
17. Zavaglia M, Astolfi L, Babiloni F, et al. A neural mass model for the
simulation of cortical activity estimated from high resolution EEG
during cognitive or motor tasks. J Neurosci Methods. 2006;
157(2):317–329.
18. Zavaglia M, Astolfi L, Babiloni F, et al. The effect of connectivity on
EEG rhythms, power spectral density and coherence among cou-
pled neural populations: analysis with a neural mass model. IEEE
Trans Biomed Eng. 2008;55(1):69–77.
19. Wendling F, Bartolomei F, Bellanger JJ, et al. Epileptic fast activity
can be explained by a model of impaired GABAergic dendritic
inhibition. Eur J Neurosci. 2002;15(9):1499–1508.
20. van Rotterdam A, Lopes da Silva FH, Van den Ende J, et al. A model
of the spatial-temporal characteristics of the alpha rhythm. Bull
Math Biol. 1982;44(2):283–305.
21. Lopes da Silva FH, Storm van Leeuwen W. The cortical alpha
rhythm in dog: the depth and surface profile of phase. In: Brazier
MAB, Petsche H, eds. Architectonics of the Cerebral Cortex. New
York: Raven Press; 1978:319–333.
22. Harrison LM, David O, Friston KJ. Stochastic models of neuronal
dynamics. Phil Trans Biol Sci. 2005;360(1457):1075–1091.
23. Deco G, Jirsa VK, Robinson PA, et al. The dynamic brain: from
spiking neurons to neural masses and cortical fields. PLoS Comput.
Biol. 2008;4(8):e1000092. doi:10.1371/journal.pcbi.1000092.
24. Jirsa VK. Neural field dynamics with local and global connectivity
and time delay. Philos Transact. Series A, Math Phys Eng Sci.
2009;367(1891):1131–1143.
25. Jirsa VK, Haken H. Field theory of electromagnetic brain activity.
Phys Rev Lett. 1996;77:960–963.
26. Wright JJ, Robinson PA, Rennie CJ, et al. Toward an integrated
continuum model of cerebral dynamics: the cerebral rhythms,
synchronous oscillation and cortical stability. Biosystems.
2001;63(1–3):71–88.
27. Wright JJ, Rennie CJ, Lees GJ, et al. Simulated electrocortical activity
at microscopic, mesoscopic, and global scales. Neuropsychopharma-
cology. 2003;28(suppl 1):S80–SA93.
28. Rennie CJ, Robinson PA, Wright JJ. Unified neurophysical model
of EEG spectra and evoked potentials. Biol Cybern. 2002;86(6):
457–471.
29. Robinson PA, Whitehouse RW, Rennie CJ. Nonuniform corti-
cothalamic continuum model of electroencephalographic spectra
with application to split-alpha peaks. Phys Rev E Stat Nonlin Soft
Matter Phys. 2003;68(2 pt 1):021922.
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
30. Friston KJ, Harrison L, Penny W. Dynamic causal modeling.
NeuroImage. 2003;19:1273–1302.
31. Stephan KE, Harrison LM, Kiebel SJ, et al. Dynamic causal models
of neural system dynamics: current state and future extensions.
J Biosci. 2007;32(1):129–144.
32. Valdes-Sosa PA, Sanchez-Bornot JM, Sotero RC, et al. Model
driven EEG/fMRI fusion of brain oscillations. Hum Brain Mapp.
2009;30(9):2701–2721.
33. Freeman WJ. Nonlinear dynamics of paleocortex manifested in the
olfactory EEG. Biol Cybern. 1979;35:21–34.
34. Freeman WJ. Analytic techniques used in the search for the physi-
ological basis of the EEG. In: Gevins A, Remond A, eds. Handbook
of Electroencephalography and Clinical Neurophysiology, vol. 1:
Methods of Analysis of Brain Electrical and Magnetic Signals.
Amsterdam: Elsevier; 1986.
35. Nunez PL, ed. Neocortical Dynamics and Human EEG Rhythms.
New York: Oxford University Press; 1995.
36. Wright JJ, Liley DTJ. Dynamics of the brain at global and micro-
scopic scales: neural networks and the EEG. Behav Brain Sci.
1996;19:285–320.
37. Stam CJ, Pijn JP, Suffczynski P, et al. Dynamics of the human alpha
rhythm: evidence for non-linearity? Clin Neurophysiol. 1999;
110(10):1801–1813.
38. Valdes PA, Jimenez JC, Riera J, et al. Nonlinear EEG analysis based
on a neural mass model. Biol Cybern. 1999;81(5–6):415–424.
39. Robinson PA, Loxley PN, O’Connor SC, et al. Modal analysis of
corticothalamic dynamics, electroencephalographic spectra, and
evoked potentials. Phys Rev E Stat Nonlin Soft Matter Phys.
2001;63(4 pt 1):041909.
40. Golomb D, Rinzel J. Dynamics of globally coupled inhibitory neu-
rons with heterogeneity. Physiol Rev E. 1993;48:4810–4814.
41. Golomb D, Rinzel J. Clustering in globally coupled inhibitory neu-
rons. Physica D. 1994;72:259–282.
42. Golomb D, Wang XJ, Rinzel J. Synchronization properties of spin-
dle oscillations in a thalamic reticular nucleus model.
J Neurophysiol. 1994;72:1109–1126.
43. Golomb D, Wang XJ, Rinzel J. Propagation of spindle waves in a
thalamic slice model. J Neurophysiol. 1996;75:750–769.
44. Wang XJ. Multiple dynamical modes of thalamic relay neurons:
rhythmic bursting and intermittent phase-locking. Neuroscience.
1994;59:21–31.
45. Wang XJ, Rinzel J. Alternating and synchronous rhythms in recip-
rocally inhibitory model neurons. Neural Comp. 1992;4:84–97.
46. Wang XJ, Rinzel J. Spindle rhythmicity in the reticularis thalami
nucleus: synchronization among mutually inhibitory neurons.
Neuroscience. 1993;53:899–904.
47. Wang XJ, Rinzel J, Rogawski MA. A model of the T-type calcium
current and the low threshold spike in thalamic neurons.
J Neurophysiol. 1991;66:839–850.
48. Wang XJ, Golomb D, Rinzel J. Emergent spindle oscillations and
intermittent burst firing in a thalamic model: specific neuronal
mechanisms. Proc Natl Acad Sci USA. 1995;92:5577–5581.
49. Suffczynski P, Kalitzin S, Pfurtscheller G, et al. Computational
model of thalamo-cortical networks: dynamical control of alpha
rhythms in relation to focal attention. Int J Psychophysiol. 2001;
43(1):25–40.
50. Suffczynski P, Kalitzin S, Lopes da Silva FH. Dynamics of non-con-
vulsive epileptic phenomena modeled by a bistable neuronal net-
work. Neuroscience. 2004;126(2):467–484.
51. Wendling F, Hernandez A, Bellanger JJ, et al. Interictal to ictal transi-
tion in human temporal lobe epilepsy: insights from a computational
model of intracerebral EEG. J Clin Neurophysiol. 2005;22(5):343–56.
87
52. Traub RD. Simulation of intrinsic bursting in CA3 hippocampal
neurons. Neuroscience. 1982;7:1233–1242.
53. Traub RD, Wong RKS. Synaptic mechanisms underlying interictal
spike initiation in a hippocampal network. Neurology. 1983;33:
258–266.
54. Traub RD, Knowles WD, Miles R, et al. Synchronized afterdis-
charges in the hippocampus: simulation studies of the cellular
mechanism. Neuroscience. 1984;12:1191–1200.
55. Traub RD, Draguhn A, Whittington MA, et al. Axonal gap junc-
tions between principal neurons: a novel source of network oscil-
lations, and perhaps epileptogenesis. Rev Neurosci. 2002;13
(1):1–30.
56. Traub RD, Miles R. Neuronal Networks of the Hippocampus.
Cambridge: Cambridge University Press; 1991.
57. Traub RD, Whitington MA, Colling SB, et al. Analysis of gamma
rhythms in the rat hippocampus in vitro and in vivo. J Physiol
(Lond.) 1996;493:471–484.
58. Jefferys JG, Traub RD, Whittington MA. Neuronal networks for
induced “40 Hz” rhythms. Trends Neurosci. 1996;19:202–208.
59. Jefferys JG, Traub RD, Whittington MA. Reply. Trends Neurosci.
1996;19:469–470.
60. Whittington M, Traub RD. Inhibitory interneurons and network
oscillations in vitro. Trends Neurosci. 2003;26(12):676–682.
61. Freeman WJ. Linear analysis of the dynamics of neural masses.
Annu Rev Biophys Bioeng. 1972;1:225–256.
62. Gray CM, Freeman WJ, Skinner JE. Associative changes in the spa-
tial amplitude patterns of rabbit olfactory EEG are norepinephrine
dependent. Neurosci Abstr. 1984;10(36,2):121.
63. Freeman WJ. A proposed name for aperiodic brain activity: sto-
chastic chaos. Neural Netw. 2000;13(1):11–13.
64. Andersen P, Andersson SA. Physiological Basis of Alpha Rhythm.
New York: Appleton-Century-Crofts; 1968.
65. Andersen P, Andersson SA, Lomo T. Nature of thalamocortical
relations during spontaneous barbiturate spindle activity. J
Physiol. 1967;192:283–307.
66. Steriade M, Jones EG, Llinás RR. Thalamic Oscillations and
Signaling. Neuroscience Institute Publications. New York: John
Wiley & Sons; 1990.
67. Steriade M, Gloor P, Llinás RR, et al. Report of the IFCN commit-
tee on basic mechanisms. Electroencephalogr Clin Neurophysiol.
1990;76:481–509.
68. Destexhe A, Contreras D, Sejnowski TJ, et al. Modeling the control
of reticular thalamic oscillations by neuromodulators. Neuroreport.
1994;5:2217–2220.
69. Destexhe A, Bal T, McCormick DA, et al. Ionic mechanisms under-
lying synchronized oscillations and propagating waves in a model
of ferret thalamic slices. J Neurophysiol. 1996;76:2049–2070.
70. Jahnsen H, Llinás R. Electrophysiological properties of guinea-pig
thalamic neurones: an in vitro study. J Physiol (Lond.) 1984;349:
205–226.
71. Jahnsen H, Llinás R. Ionic basis for the electro-responsiveness and
oscillatory properties of guinea-pig thalamic neurons in vitro.
J Physiol (Lond.) 1984;349:227–247.
72. Steriade M, Deschènes M, Domich L, et al. Abolition of spindle
oscillations in thalamic neurons disconnected from nucleus retic-
ularis thalami. J Neurophysiol. 1985;54:1473–1497.
73. Avanzini G, de Curtis M, Panzica F, et al. Intrinsic properties of
nucleus reticularis thalami neurons of the rat studied in vitro.
J Physiol. 1989;416:111–122.
74. Bal T, von Krosigk M, McCormick DA. Synaptic and membrane
mechanisms underlying synchronized oscillations in the ferret
LGNd in vitro. J Physiol (Lond.) 1995;483:641–663.
88 Part I ■ Basic Principles
75. Bal T, von Krosigk M, McCormick DA. Role of the ferret perigenic-
ulate nucleus in the generation of synchronized oscillations in
vitro. J Physiol (Lond.) 1995;483:665–685.
76. Huguenard JR, Prince DA. Intrathalamic rhythmicity studied in
vitro: nominal T-current modulations causes robust anti-oscillatory
effects. J Neurosci. 1994;14:5485–5502.
77. Von Krosigk M, Bal T, McCormick DA. Cellular mechanisms of
a synchronized oscillation in the thalamus. Science. 1993;261:
361–364.
78. Warren RA, Agmon A, Jones EG. Oscillatory synaptic interactions
between ventroposterior and reticular neurons in mouse thalamus
in vitro. J Neurophysiol. 1994;72:1993–2003.
79. Landisman CE, Long MA, Beierlein M, et al. Electrical synapses in
the thalamic reticular nucleus. J Neurosci. 2002;22:1002–1009.
80. Steriade M, Llinás RR. The functional states of the thalamus and
the associated neuronal interplay. Physiol Rev. 1988;68:649–742.
81. McCormick DA. Neurotransmitter actions in the thalamus and
cerebral cortex and their role in neuromodulation of thalamocor-
tical activity. Prog Neurobiol. 1992;39:337–388.
82. McCormick DA, Bal T. Sensory gating mechanisms of the thala-
mus. Curr Opin Neurobiol. 1994;4:550–556.
83. Destexhe A, Sejnowski TJ. Interactions between membrane con-
ductances underlying thalamocortical slow-wave oscillations.
Physiol Rev. 2003;83(4):1401–1453.
84. Hughes SW, Crunelli V. Thalamic mechanisms of EEG alpha
rhythms and their pathological implications. Neuroscientist. 2005;
11(4):357–372.
85. Lörincz ML, Crunelli V, Hughes SW. Cellular dynamics of cholin-
ergically induced alpha (8–13 Hz) rhythms in sensory thalamic
nuclei in vitro. J Neurosci. 2008;28(3):660–671.
86. Lopes da Silva FH, van Lierop THMT, Schrijer CF, et al. Organization
of the thalamic and cortical alpha rhythms: spectra and coherences.
Electroencephalogr Clin Neurophysiol. 1973;35:627–639.
87. Chatile M, Milleret C, Buser P, et al. A 10 Hz “alpha-like” rhythm
in the visual cortex of the waking cat. Electroencephalogr Clin
Neurophysiol. 1992;83:217–230.
88. Lanoir J. Étude électrocorticographique de la veille et du sommeil
chez le chat, organisation du cycle nycthéméral, rôle du thalamus.
Thesis Doctor of Science, Centre Régional de recherche et de doc-
umentation Pédagogiques, Marseille, France, 1972.
89. Zetterberg LH. Stochastic activity in a population of neurons—A
system analysis approach. Rep Inst Med Physics, TNO, Utrecht
1973;1:53.
90. Lopes da Silva FH, Vos JE, Mooibroek J, et al. Partial coherence
analysis of thalamic and cortical alpha rhythms in dog—a contri-
bution towards a general model of the cortical organization of
rhythmic activity. In: Pfurtscheller G, Buser P, Lopes da Silva FH,
eds. Rhythmic EEG Activities and Cortical Functioning. Amsterdam:
Elsevier; 1980:33–59.
91. Lopes da Silva FH, Vos JE, Mooibroek J, et al. Relative contribu-
tions of intracortical and thalamocortical processes in the genera-
tion of alpha rhythms, revealed by partial coherence analysis.
Electroencephalogr Clin Neurophysiol. 1980;50:449–456.
92. Braitenberg V. On the Texture of Brains. New York: Springer-Verlag;
1977.
93. Chow KL, Leimann AL. Aspects of the structural and functional
organization of the neocortex. Neurosci Res Prog Bull. 1970;8(2):
157–183.
94. Szentagothai J. Specificity versus (quasi) randomness in cortical
connectivity. In: Brazier MAB, Petsche H, eds. Architectonics of the
Cortex. New York: Raven Press; 1978:77–97.
95. Szentagothai J. The local neuronal apparatus of the cerebral
cortex. In: Buser P, Rougeaul-Buser A, eds. Cerebral Correlates of
Conscious Experience, Inserm Symposium no. 6. Amsterdam:
Elsevier; 1978:131–138.
96. Katchalsky AK, Rowland V, Blumenthal R. Dynamic patterns of
brain cell assemblies. Neurosci Res Prog Bull. 1974;12(1):1–187.
97. Tombol T. Comparative data on the Golgi architecture of
interneurons of different conical areas in cat and rabbit. In:
Brazier MAB, Petsche H, eds. Architectonics of the Cerebral Cortex.
New York: Raven Press; 1978:59–76.
98. Liley DTJ. A continuum model of the mammalian alpha rhythm.
In: Lopes da Silva F, Cadusch JC, Borges de Almeida L, eds.
Spatiotemporal Models in Biological and Artificial Systems.
Amsterdam: IOS Press; 1997:89–96.
99. Liley DTJ, Cadusch PC, Wright JJ. A continuum theory of elec-
tro-cortical activity. Neurocomputing. 1999;26–27:795–800.
100. Liley DTJ, Cadusch PJ, Dafilis MP. A spatially continuous mean
field theory of electrocortical activity. Network Comp Neural Sys.
2002;13:67–113.
101. Contreras D, Steriade M. Spindle oscillation in cats: the role of
corticothalamic feedback in a thalamically generated rhythm. J
Physiol (Lond.) 1996;490:159–179.
102. Zetterberg LH, Kristiansson L, Mossberg K. Performance of a
model for a local neuron population. Biol Cybern. 1978;31:15–26.
103. Babloyantz A, Kaczmarek LK. Self-organization in biological sys-
tems with multiple cellular contacts. Bull Math Biol. 1979;41:
193–201.
104. Traub RD, Wong RKS. Synchronized burst discharge in disinhib-
ited hippocampal slice. II. Model of cellular mechanism. J
Neurophysiol. 1983;49:442–458.
105. Traub RD, Wong RKS. Cellular mechanism of neuronal synchro-
nization in epilepsy. Science. 1982;216:745–747.
106. Traub RD, Jefferys JG, Whittington MA. Functionally relevant
and functionally disruptive (epileptic) synchronized oscillations
in brain slices. Adv Neurol. 1999;79:709–724.
107. Traub RD, Bibbig A, LeBeau FE, et al. Cellular mechanisms of
neuronal population oscillations in the hippocampus in vitro.
Annu Rev Neurosci. 2004;27:247–278.
108. Trevelyan AJ, Sussillo D, Watson BO, et al. Modular propagation
of epileptiform activity: evidence for an inhibitory veto in neo-
cortex. J Neurosci. 2006;26(48):12447–12455.
109. Trevelyan AJ, Sussillo D, Yuste R. Feedforward inhibition con-
tributes to the control of epileptiform propagation speed. J
Neurosci. 2007;27(13):3383–3387.
110. Kager H, Wadman WJ, Somjen GG. Simulated seizures and spread-
ing depression in a neuron model incorporating interstitial space
and ion concentrations. J Neurophysiol. 2000;84(1):495–512.
111. Kager H, Wadman WJ, Somjen GG. Conditions for the triggering
of spreading depression studied with computer simulations. J
Neurophysiol. 2002;88(5):2700–2712.
112. Kager H, Wadman WJ, Somjen GG. Seizure-like afterdischarges
simulated in a model neuron. J Comput Neurosci. 2007;22(2):
105–128.
113. Borck C, Jefferys JGR. Seizure-like events in disinhibited ventral
slices of adult rat hippocampus. J Neurophysiol. 1999;82:
2130–2142.
114. Allen PJ, Fish DR, Smith SJ. Very high-frequency rhythmic activ-
ity during SEEG suppression in frontal lobe epilepsy.
Electroencephalogr Clin Neurophysiol. 1992;82(2):155–162.
115. Bragin A, Engel J Jr, Wilson CL, et al. Hippocampal and entorhinal
cortex high-frequency oscillations (100–500 Hz) in human epilep-
 Chapter 4 ■ Dynamics of EEGs as Signals of Neuronal Populations: Models and Theoretical Considerations
tic brain and in kainic acid-treated rats with chronic seizures.
Epilepsia. 1999;40(2): 127–137.
116. White JA, Banks MI, Pearce RA, et al. Networks of interneurons
with fast and slow gamma-aminobutyric acid type A (GABA-A)
kinetics provide substrate for mixed gamma-theta rhythm. Proc
Natl Acad Sci USA. 2000;97:8128–8133.
117. Banks MI, White JA, Pearce RA. Interactions between distinct
GABA(A) circuits in hippocampus. Neuron. 2000;25(2):
449–457.
118. Cossart R, Dinocourt C, Hirsch JC, et al. Dendritic but not
somatic GABAergic inhibition is decreased in experimental
epilepsy. Nat Neurosci. 2001;4(1):52–62.
119. Kramer MA, Kirsch HE, Szeri AJ. Pathological pattern formation
and cortical propagation of epileptic seizures. J R Soc Interface.
2005;2:113–127.
120. Kramer MA, Lopour BA, Kirsch HE, et al. Bifurcation control of
a seizing human cortex. Phys Rev E Stat Nonlin Soft Matter Phys.
2006;73(4 pt 1):041928.
121. Kramer MA, Szeri AJ, Sleigh JW, et al. Mechanisms of seizure
propagation in a cortical model. J Comput Neurosci. 2007;22(1):
63–80.
122. Lytton WW, Hellman KM, Sutula TP. Computer models of hip-
pocampal circuit changes of the kindling model of epilepsy. Artif
Intell Med. 1998;13(1–2):81–97.
123. Santhakumar V, Aradi I, Soltesz I. Role of mossy fiber sprouting
and mossy cell loss in hyperexcitability: a network model of the
dentate gyrus incorporating cell types and axonal topography. J
Neurophysiol. 2005;93(1):437–453.
124. Dyhrfjeld-Johnsen J, Santhakumar V, Morgan RJ, et al.
Topological determinants of epileptogenesis in large-scale struc-
tural and functional models of the dentate gyrus derived from
experimental data. J Neurophysiol. 2007;97:1566–1587.
125. Gorter JA, van Vliet EA, Aronica E, et al. Progression of sponta-
neous seizures after status epilepticus is associated with mossy
fiber sprouting and extensive bilateral loss of hilar parvalbumin
and somatostatin-immunoreactive neurons. Eur J Neurosci.
2001;13:657–669.
126. Blumcke I, Suter B, Behle K, et al. Loss of hilar mossy cells in
Ammon’s horn sclerosis. Epilepsia. 2000;41(suppl. 6):S174–S180.
127. Lopes da Silva F, Blanes W, Kalitzin SN, et al. Dynamical diseases
of brain systems: different routes to epileptic seizures. IEEE Trans
Biomed Eng. 2003;50:540–549.
128. Van Drongelen W, Lee HC, Hereld M, et al. Emergent epilepti-
form activity in neural networks with weak excitatory synapses.
IEEE Trans Neural Syst Rehabil Eng. 2005;13(2):236–241.
129. Lytton WW, Confreras D, Destexhe A, et al. Dynamic interactions
determine partial thalamic quiescence in a computer network
model of spoke-and-wave seizures. J Neurophysiol. 1997;77:
1676–1696.
130. Tsakiridou E, Bertollini L, de Curtis M, et al. Selective increase in
T-type calcium conductance of reticular thalamic neurons in a
rat model of absence epilepsy. J Neurosci. 1995;15:3310–3317.
131. Leresche N, Parri HR, Erdemli G, et al. On the action of the
antiabsence drug ethosuximide in the rat and cat thalamus. J
Neurosci. 1998;18(13):4842–4853.
132. Suffczynski P, Pijn JPM, Pfurtscheller G, et al. Event-related dynam-
ics of alpha band rhythms: a neuronal network model of focal
ERD/surround ERS. In: Pfurtscheller G, Lopes da Silva F, eds. Event
Related Desynchronization, Handbook of Electroencephalography and
Clinical Neurophysiology, Revised Series vol. 6. Amsterdam: Elsevier
Science; 1999:67–85.
89
133. Roberts JA, Robinson PA. Modeling absence seizure dynamics: impli-
cations for basic mechanisms and measurement of thalamocortical
and corticothalamic latencies. J Theor Biol. 2008;253:180–201.
134. Meeren HK, Pijn JP, Van Luijtelaar EL, et al. Cortical focus drives
widespread corticothalamic networks during spontaneous
absence seizures in rats. J Neurosci. 2002;22(4):1480–1495.
135. Lytton WW. Computer modelling of epilepsy. Nat Rev Neurosci.
2008;9:626–637.
136. Steriade M. Mechanisms underlying cortical activation: neuronal
organization and properties of the midbrain reticular core and
intralaminar thalamic nuclei. In: Pompeiano O, Ajmone-Marsan
C, eds. Brain Mechanisms and Perceptual Awareness. New York:
Raven Press; 1981:327–377.
137. Steriade M, Nunez A, Amzica F. A novel slow ( 1 Hz) oscillation
of neocortical neurons in vivo: depolarizing and hyperpolarizing
components. J Neurosci. 1993;13:3252–3265.
138. Steriade M, Nunez A, Amzica F. Intracellular analysis of relations
between the slow ( 1 Hz) neocortical oscillation and other sleep
rhythms of the electroencephalogram. Neuroscience. 1993;13:
3266–3283.
139. Bazhenov M. Timofeev I, Steriade M, et al. Model of thalamocor-
tical slow-wave sleep oscillations and transitions to activated
states. J Neurosci. 2002;22(19):8691–8704.
140. Ball GJ, Gloor R, Schaul N. The cortical electromicrophysiology
of pathological delta waves in the electroencephalogram of cats.
Electroencephalogr Clin Neurophysiol. 1977;43:346–361.
141. Nunez PL. Electric Field of the Brain. New York: Oxford
University Press; 1981.
142. Nunez PL. Generation of human EEG by a combination of long
and short range neocortical interactions. Brain Topogr. 1989;1:
199–215.
143. Nunez PL. Toward a quantitative description of large-scale neo-
cortical dynamic function and EEG. Behav Brain Sci. 2000;23(3):
371–398; discussion 399–437.
144. Nunez PL, Wingeier BM, Silberstein RB. Spatial-temporal struc-
tures of human alpha rhythms: theory, microcurrent sources,
multiscale measurements, and global binding of local networks.
Hum Brain Mapp. 2001;13:125–164.
145. Montaron MF, Bouyer JJ, Rougeul A, et al. Ventral mesencephalic
tegmentum (VMT) controls electrocortical beta rhythms and
associated alternative behavior in the cat. Behav Brain Res.
1982;6:129–145.
146. Lopes da Silva FH, van Rotterdam A, Storm van Leeuwen W, et
al. Dynamic characteristics of visual EP’s in the dog. II. Beta fre-
quency selectivity in evoked potentials and background activity.
Electroencephalogr Clin Neurophysiol. 1970;29: 260–268.
147. Gray C, Konig P, Engel A, et al. Oscillatory responses in cat visual
cortex exhibit intercolumnar synchronization which reflects
global stimulus properties. Nature. 1988;338:334–337.
148. Singer W, Gray CH. Visual feature integration and the temporal
correlation hypothesis. Annu Rev Neurosci. 1995;18:555–586.
149. Freeman WJ, Burke BC, Holmes MD. Aperiodic phase resetting
in scalp EEG of beta-gamma oscillations by state transitions at
alpha-theta rates. Hum Brain Mapp. 2003;19(4):248–272.
150. Howard MW, Rizzuto DS, Caplan JB, et al. Gamma oscillations
correlate with working memory load in humans. Cerebral Cortex.
2003;13:1369–1374.
151. Pfurtscheller G, Graimann B, Huggins JE, et al. Spatiotemporal
patterns of beta desynchronization and gamma synchronization
in corticographic data during self-paced movement. Clin
Neurophysiol. 2003;114(7):1226–1236.
90 Part I ■ Basic Principles
152. Sederberg PB, Kahana MJ, Howard MW, et al. Theta and gamma
oscillations during encoding predict subsequent recall. J
Neurosci. 2003;23:10809–10814.
153. Boeijinga PH, Lopes da Silva FH. Differential distribution of beta
and theta EEG activity in the entorhinal cortex of the cat. Brain
Res. 1988;448:272–286.
154. Kay LM, Lancaster LR, Freeman WJ. Reafference and attractors in
the olfactory system during odor recognition. Int J Neural Syst.
1996;7:489–495.
155. Leung LS. Recurrent inhibition model of hippocampal CA in
vivo (letter to editor). Trends Neurosci. 1996;19:468–469.
156. Freeman WJ. Feedback models of gamma rhythms (letter to edi-
tor). Trends Neurosci. 1996;19:468.
157. Whittington MA, Traub RD, Kopell N, et al. Inhibition-based
rhythms: experimental and mathematical observations on net-
work dynamics. Int J Psychophysiol. 2000;38(3):315–336.
158. Perez Velasquez JLP, Carlen P. Gap junctions, synchrony and
seizures. Trends Neurosci. 2000;23:68–74.
159. Traub RD, Whittington MA, Buhl EH, et al. A possible role for gap
junctions in generation of very fast EEG oscillations preceding the
onset of, and perhaps initiating, seizures. Epilepsia. 2001;42:153–170.
160. Pinto DJ, Jones SR, Kaper TJ, et al. Analysis of state-dependent
transitions in frequency and long-distance coordination in a
model oscillatory cortical circuit. J Comput Neurosci. 2003;15(2):
283–298.
161. Menon V, Freeman WJ, Cutillo BA, et al. Spatio-temporal corre-
lations in human gamma band electrocorticograms. Electroen-
cephalogr Clin Neurophysiol. 1996;98:89–102.
162. Babloyantz A. Strange attractors in the dynamics of brain activ-
ity. In: Haken H, ed. Complex Systems Operational Approaches.
Berlin: Springer-Verlag; 1985:116–123.
163. Schuster HG. Deterministic Chaos. Weinheim: PhysikVerlag; 1984.
164. Grassberger P, Procaccia I. Measuring the strangeness of strange
attractors. Physica. 1983;9D:189–208.
165. Grassberger P, Schreiber T, Schaffrath C. Chaos. Int J Bifurc.
1991;1:521–547.
166. Ott E. Chaos in Dynamical Systems. New York: Cambridge
University Press; 1993.
167. Rapp PE, Albano AM, Schmah TI, et al. Filtered noise can mimic
low-dimensional chaos. Phys. Rev. E. 1993;47:2289–2297.
168. Basar E. Chaotic dynamics and resonance phenomena in brain
function: progress, perspectives, and thoughts. In: Basar E, Bullock
TH, eds. Chaos in Brain Function. Berlin: Springer; 1990:1–31.
169. Casdagli MC, Iasemidis LO, Savit RS, et al. Nonlinearity in inva-
sive EEG recordings from patients with temporal lobe epilepsy.
Electroencephalogr Clin Neurophysiol. 1997;102:98–105.
170. Elbert T, Ray WJ, Kowalik JE, et al. Chaos and physiology: determin-
istic chaos in excitable cell assemblies. Physiol Rev. 1994;74:1–47.
171. Jansen BH. Is it and so what? A critical review of EEG chaos. In:
Duke DW, Pritchard WS, ed. Measuring Chaos in the Human
Brain. Singapore: World Scientific; 1991:49–82.
172. Pijn JP, Velis DN, Lopes da Silva FH. Chaos or noise in EEG sig-
nals: dependence on state and brain site. Electroencephalogr Clin
Neurophysiol. 1991;79:371–381.
173. Theiler J, Rapp PE. Reexamination of the evidence for lowdimen-
sional nonlinear structure in the human electroencephalogram.
Electroencephalogr Clin Neurophysiol. 1996;99:213–222.
174. Varela F, Lachaux JP, Rodriguez E, et al. The brainweb: phase syn-
chronization and large-scale integration. Nat Rev Neurosci.
2001;2(4):229–239.
175. Pijn JP, Velis DN, van der Heyden M, et al. Nonlinear dynamics
of epileptic seizures on basis of intracranial EEG recordings.
Brain Topogr. 1997;9(4):249.
176. Lopes da Silva FH, Pijn JP. Epilepsy: network models of genera-
tion. In: Arbib MA, ed. The Handbook of Brain Theory and Neural
Networks. Cambridge, MA: A Bradford Book, MIT Press;
1995:367–369.
177. Breakspear M. “Dynamic” connectivity in neural systems: theoret-
ical and empirical considerations. Neuroinformatics. 2004;2(2):
205–226
178. Lopes da Silva FH. Neural mechanisms underlying brain waves:
from neural membranes to networks. Electroencephalogr Clin
Neurophysiol. 1991;79:81–93.
179. Freeman WJ. Dynamics of image formation by nerve cell assem-
blies. In: Basar E, Flohr H, Haken H, et al., eds. Synergetics of the
Brain. Berlin: Springer-Verlag; 1983:102–121.

Biophysical Aspects of EEG and
Magnetoencephalogram Generation
FERNANDO H. LOPES DA SILVA WITH AN APPENDIX BY AB VAN ROTTERDAM
T
he electrical activity of the brain consists of ionic cur-
rents generated by biochemical sources at the cellular
level. These ionic currents cause electric and magnetic
fields that can be measured in the brain and surrounding tis-
sues. The behavior of these fields can be predicted because they
obey physical laws. This chapter is an introduction to the bio-
physical aspects of the generation of electroencephalogram
(EEG) signals. It is convenient to consider the generation of
EEG signals in biophysical terms because this is the exact way to
determine the potential distribution at the scalp given a set of
intracerebral current sources, that is, the so-called forward
problem of electroencephalography. An understanding of this
problem is necessary to discuss the inverse problem that consti-
tutes the main concern of clinical electroencephalography,
which is to determine the intracerebral sources given a meas-
ured potential distribution at the scalp. The inverse problem
has no unique solution, as shown long ago by Helmholtz (1);
therefore, it is essential to understand the forward problem so
that constraints of the inverse problem may be well examined.
This chapter treats the biophysical basis of the generation of
EEG potentials nonmathematically so that the general reader
may follow it more easily; in the theoretical appendix, a mathe-
matical treatment of the biophysical aspects is provided.
SPACE-DEPENDENT PROPERTIES
The electrical potential at a point in the brain in principle can
be computed if the microscopic cellular sources are known. In
expression 5.8 of the Appendix, the field potential is given as a
function of intracerebral current sources; in expression 5.10, it
is given in terms of the membrane potential.
In general terms, the field potential of a population of neu-
rons equals the sum of the field potentials of the individual
neurons. To understand EEG phenomena, the activity of popu-
lations of neurons must always be considered. EEG phenomena
can be measured only at a considerable distance from the
source if the responsible neurons are regularly arranged and
activated in a more or less synchronous way. A typical regular
arrangement is the palisade, in which the neurons are distrib-
uted with the main axes of the dendritic trees parallel to each
other and perpendicular to the cortical surface. When in such a
population (e.g., the pyramidal neurons of layers IV and V of the
cortex) the neurons are more or less simultaneously activated
by way of synapses lying at the proximal dendrites, extracellular
currents will flow; their longitudinal components (i.e., parallel
to the main axes of the neurons) will add, whereas their trans-
verse components will tend to cancel out. The result is a
CHAPTER
5
laminar current along the main axes of the neurons. The net
membrane current that results from the activation at the level
of the synapse can be either a positive or a negative ionic cur-
rent directed to the inside of the cell. Because there is no accu-
mulation of charge anywhere in the medium (see also
Appendix, expression 5.1), the injected current at the synaptic
level is compensated by other currents flowing in the medium,
as shown in Figure 5.1. At the level of the synapse and in the case
of an excitatory postsynaptic potential (EPSP), the synaptic cur-
rent is carried by positive ions; in the case of an inhibitory
postsynaptic potential (IPSP), the corresponding current is
carried by negative ions. Because the direction of a current is
defined by the direction along which the positive charge is
Figure 5.1 Current flow patterns in and around an idealized neuron
owing to synaptic activation. E, current flow caused by activation of a
synapse at the level of the apical dendrite resulting in depolarization of
the membrane and flow of a net positive current. This current causes a
sink at the site of the synapse. The extracellularly measured excitatory
postsynaptic potential (EPSP) is drawn at the left; it has a negative
polarity at the level of the synapse. At the soma there exists a distrib-
uted passive source resulting in an extracellular potential of positive
polarity. I, current flow caused by activation of a synapse at the level of
the soma resulting in hyperpolarization of the membrane and flow of a
net negative current. This results in an active source at the level of the
soma and passive sinks at the basal and distal dendrites. The extracel-
lularly measured inhibitory postsynaptic potentials (IPSPs) at the soma
and dendritic level are drawn. Note that an inhibition at the soma gen-
erates about the same extracellular potential field as an excitation at the
apical or distal dendrites.
91
92 Part I ■ Basic Principles

transported, the ionic current is directed to the intracellular

medium with an EPSP; with an IPSP, it is directed to the extra-
cellular medium. Therefore, at the level of the synapse, there is
an active sink in the case of a positive ionic current (EPSP) or
an active source in the case of a negative ionic current (IPSP).
The extracellular potential at the sink is negative; at the source,
it is positive, as can be seen from expression 5.8. Along the cell
and at a distance from the synaptic level, there exists a distrib-
uted passive source in the case of an EPSP or a distributed pas-
sive sink in the case of an IPSP.
In addition to postsynaptic potentials, other relatively slow
variations of membrane potential, such as those associated with
depolarizing or hyperpolarizing afterpotentials and dendritic
events as calcium action potentials, may also be the sources of
extracellularly measurable potentials.
At a macroscopic level, it may therefore be stated that the
potential field generated by a synchronously activated palisade
of neurons behaves like that of a dipole layer. This is no more
than a rough model of reality. It may also be considered that an
active sink (for instance, at the level of the cell somata in the
cortex) is flanked by two passive sources, one lying more super-
ficially and the other deeper than the sink; in such a case, the
potential field should behave more like a quadrupole layer. This
is, however, an unnecessary complication because in general the
two sides of the quadrupole are rather asymmetrical; one is
usually dominant. This is to be expected because it can be seen
from histological sections of the cortex that there is a clear Figure 5.2 Examples of closed, open, and open-closed fields for differ-
asymmetry of the pyramidal cells along a direction perpendicu- ent types of neuron pools in the central nervous system. Left: The pop-
lar to the cortex. These cells present a morphological polariza- ulations are drawn. Right: Neurons representing the simultaneously
tion in the vertical direction because they are characterized by a activated pools are shown together with the arrows representing the
rather long vertically directed apical dendrite that ramifies in lines of flow of current at the instant when the impulses have evaded
the most superficial layers of the cortex and by basal dendrites the cell bodies. The zero isopotential lines (0) are also indicated.
distributed around the soma; the axon runs vertically down- A: Oculomotor nucleus represented schematically by only one neuron
ward to the white matter. For the sake of simplicity, it therefore with dendrites oriented radially outward. The isopotential lines are cir-
may be assumed that most potential fields generated in pal- cles; the currents flow entirely within the nucleus, resulting in closed
isades of neurons can be mimicked by simple dipole layers. field (all points outside the nucleus remain at zero potential).
Lorente de No (2) named this type of potential field the “open B: Superior olive represented by neurons each having a single dendrite
field” (Fig. 5.2C) (3), in contrast to those fields generated by oriented radially inward. The currents result in a closed field.
neurons with dendritic arborizations radially distributed about C: Accessory olive represented by only one neuron with a single long
the soma; according to his description, these would generate dendrite. The arrangement of sources and sinks in this structure permits
“closed fields” as shown in Figure 5.2A,B. In the latter case, the the spread of current in the volume of the brain and thus results in an
configuration of the radially oriented neurons is such that each open field. D: Two structures mixed together, generating an open-closed
neuron may be considered as generating a central source (or field. (Modified from Lorente de No R Action potential of the motoneu-
sink, depending on the type of synaptic activity) surrounded by rons of the hypoglossus nucleus. J Cell Comp Physiol. 1947;29: 207–287;
spherically distributed sinks (or sources). The field potential in Hubbard, J I, Llinas, R Quastel D M J Electrophysiological Analysis of
this case is equivalent to that of a distribution of radially ori- Synaptic Transmission. London: Edward Arnold Ltd; 1969.)
ented dipoles at the surface of a sphere. It can be seen intuitively
that, under these circumstances, not only the tangential but also
the radial components of current cancel each other; this can
also be proved by the volume conduction theory (Fig. 5.3) (4). variation caused by an action potential generates a field that is
Postsynaptic potentials in neuronal populations with an equivalent to that of a single dipole perpendicular to the mem-
appropriate spatial organization can be sources of field poten- brane because the piece of membrane that is depolarized at any
tials that can be measured at a distance and therefore are also instant of time is small. In contrast, that of an electrotonically
sources of EEG signals. Can other types of membrane poten- conducted postsynaptic potential extends, at any moment of
tials as action potentials also contribute to those EEG signals? time, over a larger portion of the membrane; thus, it generates
This is not generally the case for two reasons (see also a field that corresponds rather to that of a dipole layer with
Appendix). The first reason is that the membrane potential dipoles perpendicular to the membrane surface. The latter
 Chapter 5 ■ Biophysical Aspects of EEG and Magnetoencephalogram Generation 93

Figure 5.3 Equipotential plots and depth profiles for open cortices. Intersection of base lines of the depth profiles is at
the center of the sphere. Tick marks represent the region of cells. Note that the 90-degree profile of the hemisphere and
45-degree profiles of the cap and punctured sphere run along the open edge of the cortex. (Adapted from Klee M, Rall W.
Computed potentials of cortically arranged populations of neurons. J Neurophysiol. 1977;40:647–666.)

attenuates with distance less rapidly than the former, as can be
seen by comparing expressions 5.12 and 5.15 of the Appendix;
this is illustrated in Figure 5.10B and has also been discussed by
Humphrey (5). These properties have been analyzed more rig-
orously by Pettersen and Einevoll (6) using realistic computer
models of pyramidal and stellate neurons with different mor-
phologies. These authors demonstrated that the neural mor-
phology and passive electrical parameters influence the width
and amplitude of extracellular spikes. These modeling studies
revealed that in all neurons there is a low-pass filtering effect, that
is, the spike-width increases with increasing distance from soma.
These authors showed quantitatively that the spike-amplitude
decays with distance r depending on dendritic morphology,
according to the rule 1/rn with 1 n 2 close to the soma and
n 2 far away along the dendritic tree. These studies have also
implications for understanding the relation between the extent of
fields produced by multiunit activity (MUA) and by synaptic
potentials, which constitute the so-called local field potentials
(LFPs or local EEG), since they substantiate why MUAs are more
confined around the neural sources than the LFPs generated by
synaptic currents. In the latter case the dipole layer lengths are
much longer than the typical action-potential current dipoles,
and thus the attenuation with distance is weaker than for the
MUAs. The second reason that action potentials do not con-
tribute to EEG signals is that the latter, owing to their short dura-
tion (1 to 2 msec), tend to overlap much less than do postsynaptic
potentials (EPSP and IPSP), which last longer (10 to 250 msec).
In those cases in which the action potentials occur simultane-
ously in clusters (e.g., when a group of fibers is excited by a short
stimulus), the field of these action potentials may also be
recorded at relatively large distances in the form of what is gen-
erally called a “compound action potential.”
Quantitative measurements in vitro and computational neu-
ronal models used to infer contributions of neuronal populations
to MEG and EEG signals.
Studies using in vitro preparations allow the simultaneous
measurement of electric and magnetic fields generated by neu-
ronal activity. A calculation of the magnetic field around an
axon was performed by Swinney and Wikswo (7) using a prepa-
ration where an isolated axon was kept in a bath with saline, in
vitro, and was stimulated electrically. These authors found that
the magnetic field resulted mainly from the intracellular cur-
rent flowing inside the axon along the axial direction, whereas
the contribution of the return passive current that flows in the
medium was two orders of magnitude smaller, and that of the
transmembrane current was negligible.
More recently, Murakami and Okada (8) applied the same
kind of approach to the neocortex. This study is particularly
relevant because it has yielded some results that may help to
interpret EEG and magnetoencephalogram (MEG) recordings
from the scalp. These authors made a computer model, based
on that proposed by Mainen and Sejnowski (9), of the four
main types of cortical neurons taking into account their real-
istic shapes. Each neuron is described as a three-dimensional
94 Part I ■ Basic Principles
compartmental model, where each compartment has its typical
geometric dimensions, passive electrical properties (membrane
capacitance and resistance, intracellular resistance), and five
voltage-dependent ionic conductances. Neuronal activity was
obtained by stimulating each neuron with an intracellular cur-
rent injected at the soma. The intracellular current is repre-
sented by a vector quantity, the magnitude of which is the
current dipole moment and is represented by the scalar quan-
tity Q in [pA m].
The population Q for a group of pyramidal neurons was cal-
culated by computing the sum of the dot product of each cur-
rent dipole Qk and the unit vector orthogonal to the cortical
surface. The same was done for the stellate cells, with a modifi-
cation due to the fact that these neurons are differently ori-
ented. The magnetic field B and electrical potential V at the
scalp can be expressed as functions of the intracellular currents,
called primary currents, and the return currents that flow in the
space around the sources. In the model the magnetic field B is
assumed to be generated by the tangential component of Q,
with respect to the inner skull surface, and the electric field is
due to both the tangential and radial components. This implies
that B is mainly caused by neuronal activity in sulci, since Q is
directed perpendicularly to the pial surface, whereas both radial
and tangential components contribute to the electric field. This
investigation led to another result of practical interest, namely it
gave some direct indications about how to interpret the magni-
tude of the magnetic fields recorded in human. According to the
model the overall magnitude of Q for the activity of pyramidal
neurons of layers V and II/III is in the order of 0.29 to 0.90 pA
m. Apparently this value deviates appreciably from that esti-
mated previously for a cell by Hämäläinen et al. (10) that was
just 0.02 pA m. That larger value, however, is of the same order
of magnitude as those estimated for hippocampal pyramidal
neurons (0.2 pA m per cell in CA1 and 0.17 pA m per cell in
CA3, (11,12)). Murakami and Okada (8) point out that assum-
ing a Q of 0.2 pA m per cortical pyramidal neuron, a population
of 50,000 synchronously active cells would generate a field with
a magnitude of 10 nA m, which corresponds precisely to the
smallest value measurable from the human cortex according to
Hämäläinen et al. (10). Assuming that a cortical minicolumn
with a diameter of 40 m contains 100 cells (13), the cortical
surface that would correspond to 50,000 cells should form a
patch with about 0.63 mm 2 area. If this cortical patch would
have a circular form its diameter would be about 0.9 mm.
In order to estimate the number of cortical cells correspon-
ding to a cortical column, the synchronous activity of which is
responsible for the EEG/MEG signal measured, one has to
know, at least approximately, the density and the spatial distri-
bution of pyramidal and other cell types within the cortex.
TIME-DEPENDENT PROPERTIES
We have emphasized above the importance of synchrony of
neuronal activity for the generation of EEG signals. We consider
here the factors that cause this synchrony. The main factor is of
a structural nature. Neuronal masses are, in general, organized
as combinations of interlocked excitatory and inhibitory
populations (14–18); the interlocking takes place by way of
recurrent collaterals forming different types of synapses that
possibly include dendrodendritic synapses and even gap junc-
tions (19).
These neural masses act as macroscopic sources of EEG sig-
nals. For example, in the case of the generation of alpha
rhythms, it has been shown that such a macroscopic source is
located within the cortex and behaves like a dipole layer
directed perpendicularly to the cortical surface (20). The
dynamic properties of alpha rhythms, such as their frequency,
depend on the parameters of the neuronal populations, such as
feedback gains, strength of connections, time properties of
synaptic potentials, and nonlinear properties (i.e., thresholds
and saturation).
Does this rhythmic activity propagate along the cortex or
form a standing field? It has been shown experimentally (20)
that, in the visual cortex of the dog, small cortical areas appear
to act as “epicenters” from which alpha rhythm activity
“spreads” in different directions. The existence of this type of
spread has been concluded on the basis of phase differences in
rhythmic components measured at different locations on the
cortical surface. It has been shown in a model study (21) that a
neuronal chain, where the model neurons are interconnected by
means of recurrent collaterals and interneurons such that the
connectivity functions are homogeneous and their strength
decays as an exponential function of distance, shows propaga-
tion of electrical activities characterized by frequencies in the
alpha band with a velocity c in the order of magnitude of
0.3 m/sec, which is in the range of experimental results. The
dynamics of the activity of neuronal populations are treated in
more detail in Chapter 4.
The propagation of electric activity in the cortex may result in
the interesting phenomenon of time dilation, that is, an EEG
transient recorded at the cortex will seem to be of shorter dura-
tion than the transient recorded simultaneously at the scalp. This
has been shown to occur in relation to epileptiform spikes (22).
The phenomenon of time dilation is due to the fact that a mov-
ing dipole seen from a distance can be observed over a longer
time than when seen from nearby. The physical basis for this phe-
nomenon is accounted for in the Appendix (expression 5.18).
INFLUENCE OF INHOMOGENEITIES
It has been assumed that the brain could be considered an infi-
nite homogeneous medium, but in reality this is not the case.
The field potentials are influenced not only by the geometry of
the neuronal populations and the electrical properties of indi-
vidual neurons but also by the existence of regions with differ-
ent conductivities in the head, that is, by the presence of
inhomogeneities. For an interpretation of field potentials meas-
ured at the scalp, therefore, it is important to take into consid-
eration the layers lying around the brain: the cerebrospinal fluid
(CSF), the skull, and the scalp. These layers account, at least in
part, for the attenuation of EEG signals measured at the scalp as
compared to those recorded at the underlying cortical surface.
This can be formulated in mathematical terms, as in the
Appendix (expression 5.19).
 Chapter 5 ■ Biophysical Aspects of EEG and Magnetoencephalogram Generation
To compute the potential distribution at the surface of the
scalp caused by a dipole placed within the brain, it is necessary
to compute the potential distribution at the surface of the differ-
ent shells (i.e., to solve the boundary value problem). This can be
done by solving the Poisson equation (see equation 5.7) in polar
coordinates as done in the Appendix. Already in the 1970s, the
influence on the EEG of the specific electrical properties of the
tissues surrounding the brain was recognized (23–27). The sys-
tematic study of Ary et al. (28) led to the proposal of a method
that would correct the errors introduced by variations in skull
and scalp thickness. Initially, the models of the head volume
conductor consisted of concentric spheres, usually three to
account for the brain, skull, and scalp; in some models a fourth
layer was included representing the CSF. In the classic models,
the three main concentric spheres typically had radii of 90, 83,
and 78 mm. The values of the conductivities of the three layers
commonly used were proposed by Geddes and Baker (29): for
the brain 0.33, for the skull 0.0042, and for the scalp also
0.33 S.m 1; in case the CSF was included, it had a conductivity
of 1.0 S.m 1. Recently more precise estimates of these conduc-
tivities were obtained by way of measurements in vivo, using an
electric impedance tomography (EIT) method combined with a
realistic model of the head (30–32). The results of the study of
Gonçalves et al. (32) showed that the ratio between the conduc-
tivities of the skull and the brain lies between 20 to 50 (for six
subjects) rather than the traditionally assumed value of 80. The
average values found in this investigation are the following:
brain: 0.33 S.m 1; skull: 0.0081 S.m 1. It is also important to
note that the intersubject variance of the estimates decreased by
half when a realistic model was used, in comparison with a
spherical model. However, a factor of 2.4 was found between the
subjects with the extreme values of the ratio of conductivities.
This implies that these values should be estimated for each indi-
vidual to increase the reliability of the estimated conductivities.
In addition, it is known that the conductivity of the various
tissues is not homogeneous. The conductivity of the skull varies
with its thickness and bone structure. Law et al. (33) showed
that there is an inverse relation between skull resistivity and
thickness. Cuffin (34) and Eshel et al. (35) investigated the
effects of varying the conductivity of the skull; the former
determined that the effect of local variations in skull and scalp
thickness was slightly larger on the EEG than on the MEG,
while the latter showed a correlation between interhemispheric
asymmetry in skull thickness and the amplitude of the scalp
EEG. The very thorough studies of Haueisen (36,37) on the
influence of various combinations of conductivities on both the
EEG and MEG showed that the scalp EEG is most influenced by
the conductivities of the skull and the scalp, while the MEG was
most influenced by the conductivities of the brain tissue and
the CSF. We should note that the conductivity of the tissues of
the head can also present anisotropy. Indeed, it is known that in
brain tissue, the conductivity measured in a direction parallel to
a fiber tract can be 10 times larger than in the perpendicular
direction (38). It is, however, difficult to integrate this finding in
global models of the whole head, since fiber tracts are organized
in a most complex way within the anatomical constraints of the
folds of the brain. Nevertheless, De Munck (39), using a model
95
consisting of five concentric shells, solved the forward problem
in an analytic form for the case in which the skull and the cor-
tex had anisotropic properties. The effects of the anisotropic
conductivity of the skull were determined by Bertrand et al.
(40) and by Van den Broek (41). The latter showed that the
anisotropy of the skull causes the smearing out of the distribu-
tion of the EEG over the scalp, whereas the normal component
of the MEG is not affected.
In addition, it is relevant to emphasize that the skull does not
have a homogeneous surface, due to both the existence of
regions with different thickness and the occipital and the eye
sockets openings (42). The existence of holes in the skull, for
example, due to surgical interventions, influences also the dis-
tribution of the EEG over the scalp as demonstrated by
Bertrand et al. (40). In general, the localization of the dipoles is
then shifted toward the hole.
REALISTIC MODELS OF THE HEAD
AND NUMERICAL METHODS
In the past, most solutions of the forward problem assumed
that the head could be reasonably approximated by a spherical
model with three or four shells. However, the geometry of this
volume conductor is clearly not spherical. Therefore, since the
late 1980s, realistically shaped models were gradually intro-
duced and applied (43–47). This was made possible by the
advances in magnetic resonance imaging (MRI) that allow
making realistic estimates of the three-dimensional geometry
of the brain and surrounding tissues (48). To solve the forward
problem in a realistic shaped volume of the head, it is necessary
to apply numerical methods. This involves decomposing the
volume conductor in a mesh consisting of a relatively large
number of discrete elements. The accuracy of the method
increases with the number of these elements, and, of course, is
inversely proportional to their size, but the computational
effort also increases accordingly. The most commonly used
methods to compute the EEG potential distributions are the
boundary-element method (49–53), which is appropriate to the
case where the volume conductor is assumed to be isotropic
and piecewise homogeneous, and the volume element methods,
which are needed in case the assumptions of isotropy and
piecewise homogeneity do not hold. Examples of the latter are
the finite-volume (35), the finite-difference, and the finite-
element methods (for details see reviews by van Uitert,
Weinstein, and Johnson (54)). These methods are powerful, but
they need reliable information about the structure of the brain
and surrounding tissues, obtained from MRI scans, and of the
corresponding conductivities, to be applied in a sensible way.
They are particularly suited to calculate the influence of specific
inhomogeneities such as the influence of the cerebral ventricles
filled with CSF (41) and anisotropic conductivities (36).
In a number of studies, the advantage of using realistic mod-
els in comparison with spherical models has been demon-
strated; differences in dipole reconstructions using EEG data
were reported to reach 20 mm (55,56). Relatively smaller differ-
ences were obtained with MEG data, in the order of 4 mm (57).
This implies that the necessity of using realistic models,
96 Part I ■ Basic Principles
particularly those accommodating inhomogeneous volumes, is
stronger for EEG than for MEG recordings. In the latter case,
most calculations are based on the boundary-element method.
THE INVERSE PROBLEM: APPROACHES
TO ESTIMATE SOLUTIONS
The introduction stated that the problem of calculating the
intracerebral sources of the potentials measured at the scalp, that
is, the so-called inverse problem of electroencephalography, has
no unique solution. This means that different combinations of
intracerebral sources can result in the same potential distribution
at the scalp. The only way to tackle this problem is to make spe-
cific assumptions about the intracerebral sources that are
assumed to cause a given EEG potential distribution and to make
a model of the conductive media lying between the sources and
the recording electrodes. This implies that first a particular type
of forward problem is solved; thereafter, the theoretical scalp
potentials obtained in this way are compared with those recorded
experimentally. It should be noted that, on the one hand, a satis-
factory comparison does not necessarily imply that the assumed
sources are those really responsible for the measured EEG poten-
tials. On the other hand, however, a bad comparison must lead to
the conclusion that the assumptions are not correct with respect
to the sources or to the model, or both.
The numerical procedures by means of which the measured
EEG/MEG distribution over the scalp is compared with that
computed using the forward approach are not trivial. Usually
one expresses the difference between the real and the simulated
distributions as a summed squared difference. The question is
to find a minimum of this function, which can be accomplished
by several algorithms. The mostly used of this is that of
Marquardt (58). In the last decades many solutions of the
inverse problem have been proposed, starting from the pioneer
work of Schneider and Gerin (59), Schneider (25,26), Smith et
al. (60), Henderson et al. (61), and many others as reviewed by
Scherg (62,63) and Stok (64).
According to this approach, the intracerebral source is
assumed to be an equivalent current dipole localized within the
brain. Here we consider, in general terms, those aspects of the
inverse problem that are common to EEG and MEG. The aim
of the inverse procedure is to obtain an equivalent source,
which in general is a dipole, determined by six parameters. This
solution is called the equivalent dipole (ED), which must be
considered a mathematical abstraction. It represents the theo-
retical dipole, which generates a potential (or magnetic field)
distribution at the surface of the outer shell, that is the closest
in the least square sense to the measured distribution at the
scalp. It should be emphasized that the estimation of an ED
located within the brain gives only a rough estimate of the
center of gravity of the cortical area active at a given moment.
For example, in the case of the dipole layer shown in Figure 5.3
the corresponding ED would be located much deeper than the
surface where the individual dipolar sources are placed.
Assuming that the activity of each cortical macrocolumn
would be described by an equivalent current dipole, one would
need several thousands to account for an EEG or MEG scalp
distribution. The number of sensors, however, is much smaller,
in the order of a maximum of 128 for the EEG and 300 for the
MEG. Thus, the problem of estimating brain sources of EEG or
MEG data is underdetermined. Regularization methods have
been proposed to help solve this major difficulty.
For solving the inverse problem, the method of the least-
squares source estimation has been applied both to the activity
measured at a single moment in time or within a time interval
(65,66). In this context a convenient approach, proposed by De
Munck (39,67), is to split the set of parameters of the ED into
linear and nonlinear parameters. This procedure consists
schematically of the following steps: first, the time functions
that describe the change of the source as function of time must
be estimated for a given position and orientation of the ED; sec-
ond, the orientations must be found given the dipole time func-
tions and position. These two steps must be performed
alternatively a number of times until the best dipole orientation
and time functions are found for given dipole positions.
Thereafter, the nonlinear position parameters must be updated
and the process repeated until a best fit is obtained. This proce-
dure starts from the assumption that the position and orienta-
tion of the initial dipole are known, that is, it is based on the
stationary dipole model proposed by Scherg and Von Cramon
(65), although it is not constrained to the initial choice.
If a reasonable assumption on the initial position and orien-
tation cannot be made, one may use the alternative approach of
estimating EDs for a number of successive time samples of a
given potential, or magnetic, distribution (so-called moving
dipole approach) as done by Stok (64) and Stok et al. (68) and
many others. In this way a series of EDs is obtained, as shown in
Figure 5.4. Those EDs that have about the same position and ori-
entation can be clustered and the clustered dipoles can be aver-
aged. If desired, these average ED clusters can be used as seed
points for the procedure of de Munck as described earlier (69).
We should note that, in practice, if the distribution of the
EEG potential or of the MEG field over the scalp is relatively
simple, a single ED may be an appropriate model. Of course,
one needs all available a priori information to speed up the
solution and to give it high reliability. For instance, if it is to be
expected that a given activity occurs simultaneously in both
hemispheres in areas well defined anatomically, it is advisable to
start by assuming two symmetrical dipoles at the appropriate
sites (63). An increase in the number of dipoles can easily lead
to rather complex and ambiguous interpretations. Validation of
ED models can be obtained in retrospect by relating the solu-
tions obtained to independent physiological or anatomical
information. An example of an anatomical–physiological vali-
dation is the case of the dipolar sources of alpha, mu rhythms,
and sleep spindles (70) that are distributed on brain areas as
expected on the basis of animal and human physiological
observations. An example of anatomical–pathological valida-
tion is given by the distribution of sources of epileptiform tran-
sients in patients with cortical lesions visible in MRIs (Fig. 5.5),
where the main epileptiform sources are located around the
lesions (71,72).
A number of alternative methods have been developed to
circumvent the obvious limitations of the ED approach. One of
 Chapter 5 ■ Biophysical Aspects of EEG and Magnetoencephalogram Generation 97

Figure 5.4 Projection of equivalent dipoles (EDs) based on visual evoked potentials (VEPs) (left) and visual evoked fields
(VEFs) (right) during the on-response of subject JM, to the appearance of a checkerboard presented to the left half-visual
field. Each ED is shown in three projection planes, namely occipital (top), horizontal (middle), and sagittal (bottom). An
arrow indicates the direction and length of the projection of the dipole and the starting point indicates the dipole location.
Each arrow is identified by a small number that runs from 1 to 11. A group of arrows represents an interval of 50 msec, in
steps of 5 msec, spanning the time interval of 55 to 160 msec after start of the stimulus. Scale: one division indicates 10 mm
for the position and 6.7 10–9 A m, for the components. Each dipole is plotted with a line width corresponding to the
measure of fit 2: thick for 0.1, medium for 0.1 0.2, thin for 0.2 0.4, very thin for 0.4. (Modified
from Stok CJ, Spekreijse HJ, Peters MJ, et al. A comparative EEG/ MEG equivalent dipole study of the pattern onset visual
response. New trends and advanced techniques in clinical neurophysiology. EEG Suppl. 1990;41:34–50.)

these consists in applying spatial filtering to the data so that sig-
nals from a given location may be privileged with respect to the
rest. This is the so-called linearly constrained minimum vari-
ance beam forming (73) that has been applied in practice with
some interesting results (74). Related methods are the synthetic
aperture magnetometry (SAM) method introduced by
Robinson and Vrba (75), and the parametric mapping method
applied by Dale et al. (76). In addition, methods have been pro-
posed to obtain estimates of multiple dipoles with little a priori
information, such as the multiple signal classification (MUSIC)
algorithm (77–79). A number of laboratories have pursued this
issue by creating new algorithms based on the assumption that
the sources are at fixed locations within the brain, namely at a
given node of the triangular tessellation of the cortical surface
as extracted from the subject’s MRI. In this way the inverse
solution is simplified to finding the linear parameters of
the sources. Nevertheless, this problem is still underdetermined.
The methods used to solve this problem use different forms of
regularization parameters (80). In this way particular forms of
inverse solutions can be obtained: minimum norm (MN)
(81–84), weighted resolution optimization (WROP) method of
Grave de Peralta Menendez et al. (85), and the low-resolution
98 Part I ■ Basic Principles

Figure 5.5 Top: Channel overlays and

topographic maps of the average magnetic
fields for the five spike clusters in patient
C. The field maps represent the magnetic
field distribution at the time of the marker,
indicated in the channel overlays by the
dashed vertical line. Above each map the
cluster number and, within brackets, the
number of spikes it contains. Bottom:
Position of equivalent dipoles. In the top
axial and sagittal magnetic resonance
imaging (MRI) slices the dipoles fitted to
the average magnetoencephalogram
(MEG) spike of clusters 1 and 2, com-
bined. In the bottom slices are the dipole
locations for cluster 5. For clusters 3 and 4
equivalent dipole sources with less than
10% residual error were not found. The
boundary of the structural lesion in the
right frontal lobe for this patient is marked
by dotted lines. (Adapted from Van’t Ent
D, Manshanden I, Ossenblok P, et al. Spike
cluster analysis in neocortical localization
related epilepsy yields clinically significant
equivalent source localization results in
magnetoencephalogram [MEG]. Clin
Neurophysiol. 2003;114(10):1948–1962.)

brain electromagnetic tomography (LORETA). The latter was
proposed by Pascual-Marqui et al. (86) and uses the discrete
spatial Laplacian operator for regularization. This method
yields a very smooth inverse solution (87).
In any case, dipole models provide less ambiguous solutions
than the more sophisticated methods described above, since
they are based on simpler assumptions. Nonetheless, they yield
images that are less attractive than the latter methods. In any
case, dipolar methods are only meaningful if the scalp field has
approximately focal character.
Alternatively, the scalp EEG may be described using the surface
Laplacian that estimates the local normal component of the cur-
rent through the skull (33) and can be computed by applying
either a local or a global approach. According to the local
approach (88–90), the surface Laplacian is computed locally using
a subset of neighboring sensors for differentiation. According to
the global approach (33,91,92), an interpolation function is first
applied to the measurements obtained over the whole head and
the differentiation is applied afterward. The relationship between
the surface Laplacian and the cortical potential distribution is only
simple in case of homogeneous skull and scalp layers. Zanow (93)
compared, in a computer simulation study, the performance of
the surface Laplacian and a linear estimation approach with
dipoles having free moment orientations, the so-called cortical
potential reconstruction based on linear estimation (Fig. 5.6).
Compared to the surface Laplacian method, the latter linear esti-
mation method provided a more accurate cortical image of the
sources for a comprehensive review of these methodologies, see
Ref. 94. Another alternative method consists in reconstructing the
potential distribution on the inner boundary of the skull by the
method of spatial deconvolution (89).
SPATIAL DISTRIBUTIONS OF SOURCES
OF NEUROPHYSIOLOGICAL SIGNALS:
THE CONTRIBUTION OF HEMODYNAMIC
INFORMATION
Currently a number of groups are attempting to solve the
problem of the inherent ambiguity of the EEG/MEG inverse
solutions by combining functional MRI (fMRI) data with
EEG/MEG data (76), but these approaches are still in
 Chapter 5 ■ Biophysical Aspects of EEG and Magnetoencephalogram Generation
development. A problem is that the time course of the fMRI
signal, that is, the blood-oxygen-level dependent (BOLD) sig-
nal differs from the corresponding local neurophysiological
signals, since it is delayed with respect to the latter.
Nonetheless, Logothetis and Wandell (95) examined the rela-
tion between LFPs (or local EEG), single- and multiunit activ-
ity (MUA), and high spatiotemporal fMRI responses recorded
simultaneously in monkey visual cortex, and were able to
demonstrate that while both LFP and MUA activity are corre-
lated with BOLD for short stimulation paradigms, this is not
the case for sustained oscillations, in the latter case only the
LFP signal is significantly correlated with the hemodynamic
response. The LFPs are commonly thought to be related to the
input of a neural ensemble and dendritic processing within the
network in contrast with the action potentials or “units” that
consist of the output of neurons (96). Thus, LFPs consist
99
Figure 5.6 The EEG (top row), the brain
potential reconstructed by means of linear
estimation with minimum norm constraint
(center row), and the surface Laplacian
maps (bottom row) of the deviant tones
responses to a mismatch negativity exper-
iment are shown at two latencies: at 140
msec (left column) and 180 msec (right
column) after stimulus onset. Results were
computed with Advanced Source Analysis
(ASA) software package. [Adapted from
Zanow F. Laplacian maps (bottom row) of
the deviant tones responses to a mismatch
neg-1997. Realistically shaped models of
the head and their application to EEG and
MEG, p. 144. Ph.D. thesis, University of
Twente, Enschedé.]
mainly of synchronized synaptic potentials, either excitatory or
inhibitory, with a contribution from membrane oscillations
and spike afterpotentials.
Devor et al. (97) reported the results of a systematic study in
rat somatosensory cortex, to characterize the relationship
between changes in blood-oxygen content and the neural spik-
ing and synaptic activity, showing that there is a nonlinear rela-
tionship between electrophysiological measures of neuronal
activity and the hemodynamic response.
The methodologies developed to interpret the results of
combining EEG recordings and fMRI, particularly in epileptic
disorders, have been critically reviewed (98). More recently, a
study by Shmuel and Leopold (99) combining fMRI-BOLD and
intracortical neurophysiological signals confirmed the existence
of correlations between neuronal signals consisting of either the
firing rate of a small group of neurons, or of relative power
100 Part I ■ Basic Principles
changes in the MUA band, but particularly of LFPs in the
gamma frequency band, with BOLD signals at a time lag (peaks
at about 6 sec lag) relative to neuronal activity.
THE MAGNETOENCEPHALOGRAM (MEG)
As discussed above, the ionic currents originating from bio-
chemical sources at the cellular level in the central nervous sys-
tem generate not only electric fields but also magnetic fields.
Until Cohen’s work (100), magnetic fields could not be
detected owing to the extremely low strength of the MEG and
the absence of sufficiently sensitive transducers. The magnetic
field of the brain has a strength in the order of magnitude of
10 8 gauss (G) or, in meter-kilogram-second (m kg s) units, of
10 12 Weber m 2 or tesla (W m 2 or T) (101), whereas the
earth magnetic field has a strength in the order of magnitude of
0.5 G (102). However, with the introduction of the supercon-
ducting quantum interference device (SQUID) magnetometer,
based on a superconducting effect at liquid helium tempera-
ture, sensitivities in the order of magnitude of 10 10 G have
been obtained (101).
The earth magnetic field and the urban magnetic noise fields
(10 4 to 10 3 G) can be practically removed while recording
the MEG by using the gradiometer technique. This is based on
the fact that the disturbing fields are almost constant over large
distances, whereas the MEG falls off rapidly with distance. This
is done by subtracting the induced potentials in two coils placed
close to each other; in this way, the earth and urban noise mag-
netic fields are suppressed. The magnetic field in stationary
conditions obeys Ampere’s laws and can be derived from the
microscopic sources at the cellular level, as shown in the
Appendix (from expression 5.22 onward). Although the electri-
cal potential reflects the strength and localization of the current
sources, the magnetic field reflects not only these properties but
also the direction of the current densities. This can be seen in
the Appendix by comparing expression 5.29 for the magnetic
quantities with expression 5.8 for the electric potential.
Whether the information about the direction of current densi-
ties in the brain may be derived from the MEG depends on the
practical possibility of recording the magnetic fields in three
directions. In general we may state that the EEG reflects mainly
the volume or extracellular currents, whereas the MEG is more
sensitive to the primary intracellular currents. There are very
few experimental studies where magnetic and electric fields
have been measured along with elementary electrophysiological
properties of simple brain preparations. The recent study of
Murakami et al. (103) validates the assumption indicated
above. This study examined whether evoked magnetic fields
and intra- and extracellular potentials from longitudinal hip-
pocampal CA3 slices of guinea pig can be interpreted within a
single theoretical framework that incorporates ligand- and volt-
age-sensitive conductances in the dendrites and soma of the
pyramidal cells. The intracellular potentials in this validated
model reveal that the spikes and slow waves of the magnetic
fields are generated in or near the soma and apical dendrites,
respectively.
In the section on the influence of inhomogeneities we have
seen that, to explain the EEG, we must assume the existence of
shells with different conductivities. The existence of these
shells affects the electric potential in two ways: it causes an
attenuation of the potential and it introduces ED layers at the
boundaries of the inhomogeneities as shown in expression
5.19. However, the brain and surrounding tissues behave as a
medium with constant magnetic permeability. Therefore, the
magnetic field, in contrast to the electric field, is not influ-
enced by those layers; it is nevertheless also affected by the
induced dipole layers existing at the boundaries of the inho-
mogeneities in conductivity, as shown in the Appendix
(expression 5.30).
Other properties of the MEG that should be mentioned are
the following. Neither electrodes nor a reference point are nec-
essary for recording the MEG; the transducers for the MEG
need not touch the scalp, because the magnetic field does not
disappear where conductivity is zero (free space). From expres-
sion 5.29 it can be seen that the vector potential has the same
direction as the intracerebral current density. Because the mag-
netic field is perpendicular to the vector potential, we can con-
clude that current dipoles directed perpendicularly to the
surface of the skull result in magnetic fields tangential to the
surface of the head and, vice versa, tangential current dipoles
result in magnetic fields perpendicular to the skull.
Research on MEG has advanced considerably in recent years
(see Chapter 42). Both practical and theoretical aspects of magne-
toencephalography have been elucidated, as summarized by
Cohen and Cuffin (104), Okada (105), Cuffin (106), Modena et al.
(107), Hari and Ilmoniemi (108), Hämäläinen (109), Lewine and
Orrison (110), and de Jongh et al. (111). Some theoretical aspects
of general interest will be mentioned here. Although EEG and
MEG reflect in essence the same elementary phenomena, in prac-
tice the two types of measurement differ in a number of aspects.
First, the EEG is a relative measurement; it always needs a
reference electrode. The magnetic field measurement does not
need a reference point.
Second, the MEG is a measure of the magnetic fields perpen-
dicular to the skull, which are caused by tangential current
dipoles. The radial component of the current dipole does not
generate a magnetic field outside a sphere-shaped volume con-
ductor, and thus does not contribute to the MEG. By contrast, the
EEG is a measure of both components. This means that the MEG
reveals tangential current dipoles in a clear way, which in the EEG
may be obscured by radial sources. The main intracortical dipo-
lar current sources are perpendicular to the cortical surface
(Fig. 5.1). Therefore, Cohen and Cuffin (104) have pointed out
that the MEG mainly measures the cortical current dipoles lying
in the sulci and not on the convexity of the gyri. Thus, the former
causes mainly tangential dipoles to the skull surface, whereas the
latter produces mainly radial dipoles (Fig. 5.4).
Third, the map representing the distribution of the mag-
netic field at the surface of the head caused by a tangential
current dipole is rotated by 90 degrees to the corresponding
EEG potential map as shown in Figure 5.7. This means that to
localize a dipole-like source in the x and y directions, both
MEG and EEG should be used (104), but it cannot be stated
 Chapter 5 ■ Biophysical Aspects of EEG and Magnetoencephalogram Generation 101

Figure 5.7 Examples of measured and computed maps corresponding to two time samples of a visual evoked potential
(VEP) and magnetic field (VEF). The stimulus was the appearance of a checkerboard pattern as described by Stok (64) and
Stok et al. (68). Top row: Measured VEF and VEP contour plots for two instants of time during the on-response: 100 and
130 msec. Middle row: Simulated VEF and VEP contour plots that correspond to the EDs that were estimated from the
measured VEP distributions. Bottom row: The simulated contour plots that correspond to the VEF-based EDs. The dashed
rectangles indicate the simulation area of 12 14 cm 2. Measures are in centimeters and with respect to the inion. In the
contour plots of measured data, negativity is indicated by dashed lines and in the contour plots of simulated data by thin
lines. The thick lines indicate the zero level. Note that these plots present theoretical and experimental examples of the
three main differences that can be expected between EEG and MEG contour plots. (Modified from Stok CJ, Spekreijse HJ,
Peters MJ, et al. A comparative EEG/ MEG equivalent dipole study of the pattern onset visual response. New trends and
advanced techniques in clinical neurophysiology. EEG Suppl. 1990;41:34–50.)

that one type of measurement is better than the other.
Differences in localization of sources with both methods may
be determined mainly by the direction along which the meas-
urements were made.
Fourth, the EEG represents a sum of the potentials caused by
primary and secondary sources or volume currents. In the
Appendix, it is shown that the media of the head (brain, CSF,
skull, and scalp) have different conductivities; thus, secondary
sources of electric (expression 5.19) and magnetic fields
(expression 5.30) are introduced. It is often stated, however,
that the MEG is not affected by these volume currents
(104,105). It is still debatable how much these secondary
sources contribute to the EEG and MEG under the conditions
in which these signals are experimentally measured. Plonsey
(112) pointed out that both the electric and magnetic fields are
affected by secondary sources. Van Rotterdam (113), using a
model of a volume conductor represented by a lattice including
inhomogeneities, showed that the magnetic field changes
significantly in respect to the case of the homogeneous volume
conductor under the influence of inhomogeneities in
conductivity. In certain cases of spherical symmetry, however,
the magnetic field distribution due to a current dipole is
independent of spherically symmetrical variations in resistivity
of the volume conductor (114). Whether this will hold true for
a real head is questionable.
Nevertheless, Cohen and Cuffin (104) assume that the MEG
would not be influenced by the volume currents and that this
would explain why the MEG pattern is more focal, or tighter,
than the corresponding EEG potential distribution. It appears
that still tighter MEG patterns may be obtained by using trans-
ducers composed of a number of magnetic field–sensing coils
in an array.
Fifth, there are differences between MEG and EEG measure-
ments regarding the representation of some types of nondipo-
lar sources, as discussed in detail by Cuffin (106). Different
types of sources give different results. For side-by-side line
102 Part I ■ Basic Principles
sources forming a dipole layer, the EEG gives a more accurate
representation of the actual sources than does the MEG; the
reverse is true for in-line sources, that is, a series of dipoles
aligned along the same direction. However, it should be added
that both EEG and MEG solutions of the inverse problems for
the side-by-side and in-line sources lie deeper in the head and
have a larger amplitude than the actual sources. In particular,
the sources that are in the form of lines and are more than 2 cm
long give different results when measured using EEG or MEG;
the MEG solutions are deeper for the side-by-side sources, and
EEG solutions are deeper for the in-line sources. These differ-
ences may be helpful in identifying the type of such sources in
the brain (106).
Sixth, MEG and EEG appear to differ in their capability of
localizing a current dipole source, given a set of observations
corrupted by noise. A single current dipole can be represented
by six parameters; three determine its position inside the head
and three define its components. It has been shown by Stok et
al. (115) that the EEG gives better estimates of the three com-
ponents of the current dipole than the MEG, based on obser-
vations to which noise was added. The latter, however, gives
better estimates of the three position parameters. Moreover,
the inverse procedure, for cases with bad signal-to-noise ratios,
fails more often with EEG than with MEG data. The EDs esti-
mated by the inverse procedure using EEG and MEG data sep-
arately may be combined in order to obtain an average
estimate. Alternatively, the ED may be calculated directly by
determining the dipole that provides the least-squares fit to
EEG and MEG simultaneously. It can be demonstrated that the
latter procedure provides slightly better results than the com-
bination of the two estimates into an average estimate. This
reinforces the statement that the two types of measurement,
MEG and EEG, indeed provide complementary information
and should be combined whenever precise source localization
is desired.
Seventh, spherical models of the head perform much bet-
ter with MEG than with EEG data, because the former is less
sensitive to the influence of volume currents, while these cur-
rents are affected appreciably by deviations from the ideal
head. This means that inverse solutions based on MEG data
may be simpler to obtain since spherical models may be more
readily used in this case than with EEG measurements (116).
Nevertheless, volume currents may still affect the MEG as
shown by Van Uitert et al. (54). Certainly one of the benefits
that has been obtained by the introduction of the MEG tech-
nique is that a considerable amount of research on brain
activity by competent multidisciplinary groups has led to a
renewed interest in the biophysical study of the electric and
magnetic activity recorded from the scalp. This led to the
development of more sophisticated models of the neuronal
sources and of the volume conductor. These developments
have been of interest for obtaining a better understanding not
only of the MEG but also of the EEG. Initially the experimen-
tal results obtained by the new technique have led to the
thought that the MEG is superior to the EEG in locating brain
sources (108).
However, the question of whether the MEG offers real
practical advantages over the EEG has been raised (64). Stok
et al. (68) made a systematic theoretical and experimental
comparison of EEG and MEG and concluded that the MEG is
not, in general, superior over the EEG, although the MEG
localizes the equivalent sources slightly better than the EEG,
but it estimates the dipole components slightly worse. Two
general conclusions of these studies should be put in evidence:
(i) when model assumptions are not violated, the MEG and
EEG lead essentially to the same inferences about the source
(excluding magnetically silent sources); and (ii) when model
assumptions are violated, the MEG and EEG respond, in gen-
eral, in different ways such that a comparison of the results
obtained with both methods may give clues regarding the
kind of model violation present. Therefore, it is important to
use both methods to obtain a higher reliability in functional
localization studies (117).
This question has been addressed directly by comparing
MEG and EEG maps obtained in patients where dipoles were
implanted, using indwelling intracranial electrodes that were
placed for the diagnosis of an epileptogenic focus (55,118).
These authors concluded, from a limited number of experi-
mental measurements, that the average errors of localization
were 10 mm for the EEG and 8 mm for the MEG, and thus that
the MEG was not superior to the EEG. Nevertheless, the main
advantage of the MEG may be the fact that the MEG is not
affected by radial sources and, in this way, the use of both MEG
and EEG may complement each other and help to establish a
better model of the source, when it cannot be described as
being a simple dipole.
In conclusion, more research is needed where the two meth-
ods are directly compared using the same data. In the tutorial
book on “MEG Methods” edited by Hansen et al (119), some
frequently asked questions about these issues are discussed in
more detail (120).
APPENDIX: THE MATHEMATICAL BASIS
AB Van Rotterdam
The ionic currents in the brain cause electrical and magnetic
fields obeying Maxwell’s and Ohm’s laws. These fields have a
direction and magnitude; they must, therefore, be represented
as vector functions. The electric field is given by E (r ,t) and the
magnetic field by H (r ,t) , because these vectors depend on
location r and on time t. (Vector quantities are indicated by an
arrow above the symbol.)
In brain and surrounding tissues, the material constants
are approximately the following: conductivity, ≈ 10–1 –1 m –1;
dielectric constant, ≈ 10–9 F m –1; and magnetic permeability,
≈ 10–6 H m –1. Do potential changes taking place in the brain
at a frequency of, for example, 1 k Hz cause propagated elec-
tromagnetic waves in the medium that may be detected after
a certain time at a distant point such as the scalp? It can be
proved (121) that, in a medium such as the brain with the
material constants indicated above, the propagation velocity
of the electromagnetic waves is of the order of 105 m s–1,
 Chapter 5 ■ Biophysical Aspects of EEG and Magnetoencephalogram Generation 103

change of current in the volume of the cube. When no net cur-

rent flows into the cube (i.e., the current that goes into the vol-
ume comes out again), this total change must be zero, as
indicated in expression 5.1. Thus, we may state that the current
is stationary. In the case of a stationary current, the electrical
field E(r ) , or simply E, is related to the electrical potential V(r ) ,
or simply V, by the following expression:

E gradV (5.3)

In this expression the operator gradient (grad) indicates dif-

ferentiation of the scalar function V, of space and time (time is
considered here to be fixed) in the following way:

V V V T
grad V a , , b (5.4)
x y z
where gradV is a vector quantity called the gradient of V.
Figure 5.8 Cubic volume with sides dx, dy, and dz. Current density In a conductive medium, Ohm’s law is valid; that is:
flowing in: Jz (x — dx, y — dy, z — dz); Jy (,y — dy) and Jz (,,z — dz)
or J(r dr ) . Current densities flowing out: Jx (x,y,z); Jy (,y,) and Jz J Ji sE (5.5)
(,,z) or J(r ) . When divJ 0, no charge is accumulated in volume
where J is the current density injected in the medium (e.g., a
dxdydz. Inset: Decomposition of current density vector J in components transmembrane current density caused by synaptic activity in a
in x, y, and z direction as Jz , Jy, and Jx. neuron).
Taking the divergence of both sides of expression 5.5 and
using expression 5.1, one may write:
which implies that one need not be concerned with the prop-
agation of electromagnetic waves caused by potential changes divJi
within the brain. This means that, in practical terms, the divE (5.6)
s
effects of those potential changes may be detected simultane-
ously at any point in the brain or surrounding tissues. This
Substituting E by – gradV according to expression 5.3, one
important statement leads to the conclusion that the currents
can write an expression that is called the Poisson equation for
caused by sources in the brain at any moment in time behave
the potential field owing to an injected current:
in a stationary way. This means that no charge is accumulated
at any time in the brain (122). 2 2 2
V V V divJi
Therefore, it can be stated that for the current density J(r ) divgrad V (5.7)
or, for simplicity, J, at any moment of time: x2 y2 z2 s

div J O (5.1) In case the medium where r o is situated is assumed to be

where the operation div J is called divergence of the vector J; infinite, isotropic, and homogeneous, it can be proved that the
J is the current density in A m –2. The operator div must be solution of the Poisson equation is the following:
understood as indicating differentiation of a vector; expression
5.1 can also be written in another way that may help in under- 1 divJi 3
V(r o ) dr (5.8)
standing its meaning. Let us consider the expression for J for 4 s R
vol
the case of an infinitesimally small cube with sides dx, dy, and
dz (Fig. 5.8). It holds that: which gives the values of the potential at a point r o in the
volume conductor from injected current densities Ji, at points
Jx Jy Jz
divJ dxdydz a b dx a b dy a dxdyb dz r , lying at a distance R from r o ( R 0r r o 0); the integral
x dydz y dydz z represents the summation over all current sources within the
(5.2) volume; note that divergence of the vector J gives a scalar
quantity.
In this expression the factors between brackets, such as Summarizing, it has been shown that the space-varying and
( Jx/ x dydz), represent the rate of change of current through the time-varying parts of the electrical variables can be sepa-
the surfaces dydz, dxdz, and dxdy. To account for the total rated; the time-varying part is the same everywhere in the
change of current, it is necessary that those factors are multi- medium owing to the large propagation velocity, and the space-
plied by the sides of the cube dx, dy, and dz, then these products varying part can be computed for every instant in time accord-
must be summed. Thus expression 5.2 represents the total ing to expression 5.8.
104 Part I ■ Basic Principles

This general expression is not the only way to define V(r o ) . To understand the effect of the solid angle in a more practi-
Two other expressions (5.10 and 5.12) are given below. These cal way, it is useful to derive an explicit expression for d :
are more convenient because they relate the extracellular
potential field directly to membrane processes of the individ- 1 # R cos
ual neurons in the brain. One of these expressions can be d (r r o) grada b dO (r ) #
3 dO (r ) dO
R R R2
derived assuming that the neuronal membrane can be consid-
(5.11)
ered as equivalent to a double layer with an inner (intracellu-
lar) membrane potential Vm and an outer (extracellular) where is the angle between R and dO; it should be noted that
potential V. In these terms, V( r o ) is given by the following R is the vector connecting the point r o where V is measured and
expressions: the point r where the dipole is located (see Fig. 5.8).
It should be noted in expression 5.9 that a change in mem-
1 1 # brane potential such as a depolarization in a small piece of
V(r o ) (s iVm (r ) s e(r ))grada b dO (r ) (5.9)
4 s R membrane potential results in an intracellular potential Vm and
s
an extracellular potential V. In this case, the membrane behaves
where i is the intracellular and e is the extracellular conduc- like a current dipole consisting of a sink at the outside and a
tivity, and dO is pointing to the extracellular medium; the inner source at the inside. Thus, there is a jump in potential across the
product grad (1>R) #dO is called the solid angle d (r r o )
membrane. A current dipole, or simply a dipole, is defined by its
strength and direction; it is represented by the vector p (r ) . In the
subtended by the surface element dO (r ) on the membrane sur- example described above, the dipole is oriented perpendicularly
face S and seen from the extracellular point r o. It must be to the membrane surface dO; thus, the dipole is represented as
emphasized that expression 5.8 takes the form of expression 5.9 p(r )dO. According to the above, another expression for the
only when the injected current densities Ji originate at the cell potential V(r o ) caused by one dipole can be derived (123):
membrane as derived by Plonsey (121). Expression 5.9 can be
simplified to the following approximate expression, because p(r ) 1 #
V(r o ) grad a b dO (r )
0s eV(r ) 0V 0s iVm (r ) 0: 4 s R

or
si
V(r o ) Vm (r )d (r ro) (5.10)
4 se p(r ) cos
s
V(r o ) dO (5.12)
4 s R2
The solid angle d subtended by a membrane surface dO
seen from a point r o can be interpreted in a geometrical way, as From expression 5.12 it can be concluded that in the case of
shown in Figure 5.9. a single dipole at the membrane, V decreases with the square of
the distant R to the source and it is also proportional to cos ;
thus, it is maximal when R is perpendicular to the surface dO.
In many cases, the membrane depolarization (or hyperpo-
larization) does not remain limited to a small piece of mem-
brane; rather, it spreads over a more or less extended membrane
surface. In this way the equivalent current source cannot be
accounted for in terms of single dipole. One must, instead,
assume that there exists a dipole layer at the membrane.
For a dipole layer located at the surface of the membrane,
one must sum the effects of all individual dipoles; therefore,
expression 5.12 becomes:

1 1 #
V(r o ) p(r )grad a b dO (r )
4 s s R

or

1
V(r o ) p(r )d (r ro) (5.13)
4 s s

Figure 5.9 Solid angle subtended by infinitesimal surface dO with

From expression 5.13, it can be noted that, if p is constant
direction dO on membrane cylinder S seen from the extracellular point over the closed membrane surface S, then V(r o ) 0, because
r o lying at a distance R 0ro r 0 from dO; the angle between the the integral of the solid angle d over an external closed surface
direction of the surface dO and r o r is . is zero. Thus, to obtain a value of V(r o ) different from zero,
 Chapter 5 ■ Biophysical Aspects of EEG and Magnetoencephalogram Generation 105

Figure 5.10 A: Dipole layer of length 1 and height h in the y-z plane seen from extracellular point r o on the x axis.
B: Attenuation (relative potential, rel. pot.) of extracellular potential as a function of distance (r o) in a log–log plot for the
case of a single dipole and for the case of a dipole layer. Note that, in the latter case, two slopes can be distinguished, the
decay of the potential with distance being more accentuated in the far field (ro 1). The arbitrary values of h = 0.0625
and I = 1 used in this model are indicated on the horizontal axis.

p must vary over the membrane surface; in other words, the where the time behavior of the field is the same as that of the
membrane potential must vary over the surface owing, for sources.
example, to local synaptic activity. In general terms, to allow for time-varying sources, expres-
In this case, the attenuation of the potential in the volume sion 5.10 must be changed as follows:
conductor depends not only on distance R and angle , but also
on the geometry of the dipole layer. For instance, for a dipole 1
layer of constant strength with length 1 and height h located V(r o, t) Vm (r , t)d (r ro ) (5.16)
4 se s
at x = 0 in the y-z plane (Fig. 5.10A) the potential field at a point
r o along the x axis will be given by the following expression: This effect can be illustrated by the following example (Fig.
5.11A). Let us assume a cell that can be simplified to an equiva-
lent cylinder (124) oriented along the z axis where a membrane
1>2 ¢ h>2 ro depolarization front propagates along the cylinder with an arbi-
V(r o ) p 2 2 2 3>2
dyzy (5.14)
y 1>2 z ¢ h>2 (r 0 y z) trary velocity c. In this case, we may write that the resulting
potential V(r o ) is proportional to a time-varying solid angle.
For ro W h and h V 1, this can be estimated as follows: From expressions 5.11 and 5.16, the potential V(r o,t) caused
by a depolarization front along the cylinder is given by:
1
V(r o ) p¢ h (5.15)
ro5 1 2 1>2
(r o>1) 6 cos ( (t) )
V(r o, t) 2 dO (5.17)
0ctn z r o 0
It can be seen that in the far field, r o W 1, the potential tends
to decay as in the case of the single dipole with r 2o where for where n z is the unit vector in the z direction. Note that the
short distances h ro 1, the potential decays proportion- denominator in expression 5.17 represents the square of the
ally to distance ro. In Figure 5.10B, it can be seen that in the case distance R(t) between r o and the time-varying location of the
of the dipole layer the attenuation has two slopes depending on front.
the relation between r o and 1, whereas in the case of a simple At a point further away in the same direction, r 1 a r o, V
dipole there is only one slope. becomes:
Until now, the discussion has considered only the behavior
of standing sources of EEG signals, that is, sources that do not 1
change in space. However, there is experimental evidence that V(r 1, t) 2 V(r o, t>a ) (5.18)
a
leads one to assume that some macroscopic sources in the
brain do propagate in space (14,20,92). This propagation of Therefore, the potential at a far location r 1 is equal to the
sources has a particular effect on the behavior of the potential potential at location r o attenuated in amplitude and in a more
field in time. This effect does not appear for standing sources dilated time scale (Fig. 5.11B).
106 Part I ■ Basic Principles

a (s e s k )V(r )d (5.19)
k 1 sk

s e V(r )d
s

in which S1 represents the membrane surface that encloses the

Figure 5.11 A: Membrane cylinder oriented along the z axis in which
a depolarization front propagates (with velocity c) along the z direc-
tion, r o and r 1 are two extracellular points at which extracellular
potentials are measured r 1 a r o. Note that the potentials V at r o and
at r 1 are different, because and R differ when seen from r 1 instead
of r o. The changes in and R as a function of time depend on
the propagation velocity c. B: Extracellular potentials V measured
simultaneously at r o and r 1 (plot with arbitrary scales) showing time
dilation.
(i)
first neuron with membrane potential Vm (r ) and intracellular
conductivity i; the extracellular conductivity is e except in the
regions bounded by the surfaces Sk, within these regions the
conductivity is k; s (r o ) is the conductivity for the region
where point r o lies: s (r o ) s k if r o is bounded by Sk and it is
e elsewhere, d are the solid angles of surfaces pointing to the
region with conductivity e.
The second term on the right describes the influence that
neurons have as inhomogeneities on the extracellular field; the
third term describes the distorting influences of the M inhomo-
geneities, whereas the fourth term expresses the influence of the
border of the medium.
Using expression 5.19, it is in principle possible to compute
the EEG measured on the scalp and generated by N neurons.
In the following model, the brain and surrounding tissues
are represented as a sphere with a number of shells within
which a single dipole is situated. The potential field in and on
the sphere is a solution of the Poisson equation obeying the
following boundary conditions: the current densities normal
to the surfaces of the shells are continuous and the tangential
components of the electric field at the boundaries are also
continuous.
Assuming radial symmetry, the solution for a radial dipole
can be written as a Legendre series, in polar coordinates as
follows (123):
q
V(r, f , ) (n 1)
bn r n 6P0n (cos ) (5.20)
a 5 an r
n 0
where P0n ( cos ) is a Legendre polynomial of the nth order. For
a tangential dipole, the solution is:

q
(n 1)
V(r, , ) a 5 cn r dnr n 6 P1n ( cos ) cos (5.21)
n 0

INFLUENCE OF INHOMOGENEITIES
Let us assume that the field potential V(r o ) generated by the where P1n (cos ) is an associated Legendre polynomial.
activity of N neurons situated in a finite homogeneous medium
The coefficients an, bn , cn , and dn for all n in both cases should
containing M structures with different conductivities and
bounded by surface S (outside the medium the conductivity is be evaluated on basis of the boundary conditions at the different
zero) can be expressed as follows (125): boundaries and the fact that the source is a dipole (23).

N
4 s (r 0 )V(r 0 ) sia (i)
Vm (r )d MAGNETIC FIELDS
i i si
N
In the stationary approximation of Maxwell’s equations for the
sea V(r )d electromagnetic fields generated by microscopic sources at
i i si the cellular level, the behavior of the magnetic field H or the
 Chapter 5 ■ Biophysical Aspects of EEG and Magnetoencephalogram Generation 107

induction field B mH is governed by two experimental In this way, it can be seen that the magnetic vector potential
physical laws. There is Ampere’s law in the stationary A reflects the direction of the current densities Ji. If there are
approximation: inhomogeneities in conductivity, the expression becomes (126):

m Ji (r ) 3 V(r )
curlH J (5.22) A( r 0 ) c dr a (s e s k) dO (r )d (5.30)
4p vol R k Sk R
and
where Sk are surfaces enclosing regions with conductivity k;
divB 0 (5.23) elsewhere is the conductivity e. The surfaces dO point to the
region with conductivity e; is assumed to be constant.
which expresses the experimental fact that magnetic monopoles
do not exist. The term curl H is a differential operation called REFERENCES
curl or rotation on a vector function; it is defined as follows:
1. Helmholtz H. Ueber einige Gesetze der Verteilung elektrischer
0H y 0H z 0H z 0H x 0H x 0H y T Ströme in Körperlichen Leitern mit Anwendung auf die
curlH ca b #a b #a bd thierischelektrischen Versuche. Prog Ann Physik Chemie.
0z 0y 0x 0x 0y 0H y 1855;89:211–233, 353–377.
(5.24) 2. Lorente de No R. Action potential of the motoneurons of the
hypoglossus nucleus. J Cell Comp Physiol. 1947;29:207–287.
For conservative or rotation-free fields, that is, fields deriv- 3. Hubbard JI, Llinas R, Quastel DMJ. Electrophysiological Analysis of
able from vector potentials (e.g., E gradV), this differenti- Synaptic Transmission. London: Edward Arnold Ltd; 1969.
ation results in a zero vector. However, for vector fields directed 4. Klee M, Rall W. Computed potentials of cortically arranged popu-
lations of neurons. J Neurophysiol. 1977;40:647–666.
along a closed path (e.g., closed current density loops), the 5. Humphrey DR. Reanalysis of the antidromic cortical response. II.
operation results in a vector field perpendicular to the original On the contribution of cell discharge and PSPs to the evoked
vector field. It can be proved that expression 5.23 is equivalent potentials. Electroencephalogr Clin Neurophysiol. 1968;25:421–442.
6. Pettersen KH, Einevoll GT. Amplitude variability and extracellular
to the statement that B can be derived from a vector potential A
low-pass filtering of neuronal spikes. Biophys J. 2008;94(3):
as follows: 784–802.
7. Swinney KR, Wikswo JP Jr. A calculation of the magnetic field of a
B curlA (5.25) nerve action potential. Biophys J. 1980;32(2):719–731.
8. Murakami S, Okada YC. Contributions of principal neocortical
By substituting expression 5.25 in 5.22 and using the defini- neurons to magnetoencephalography and electroencephalography
tion of J as a function of impressed current densities (expres- signals. J Physiol. 2006;575(3):925–936.
sion 5.5), we obtain: 9. Mainen ZF, Sejnowski TJ. Influence of dendritic structure on fir-
ing pattern in model neocortical neurons. Nature.
1 1996;382:363–366.
curlcurlA Ji sE (5.26) 10. Hämäläinen MS, Hari R, Ilmoniemi R, et al. Magneto -
m encephalography: theory, instrumentation and applications to the
noninvasive study of human brain function. Rev Med Physics.
It can be shown that curlcurlA divgrad A grad divA, 1993;65:413–497.
so that 5.26 becomes 11. Kyuhou S, Okada YC. Detection of magnetic signals from isolated
transverse CA1 hippocampal slice of the guinea pig.
divgrad A mJi grad divA ms E (5.27) J Neurophysiol. 1993;70:2665–2668.
12. Okada YC, Wu J, Kyuhou S. Genesis of MEG signals in a mam-
One can simplify expression 5.27 by making use of the ambi- malian CNS structure. Electroencephalogr Clin Neurophysiol.
1997;103(4):474–485.
guity in the definition of A (123) and choose A such that 13. Mountcastle VB. The columnar organization of the neocortex.
divA ms V the two last terms of expression 5.27 then can- Brain. 1997;120(pt 4):701–722.
cel out; expression 5.27 becomes the Poisson equation that for 14. Freeman WJ. Mass Action in the Nervous System. New York, NY:
Academic Press; 1975.
each separate component of A can be expressed as follows:
15. Katchalsky AK, Rowland V, Blumenthal R. Dynamic patterns of
brain cell assemblies. Neurosci Res Prog Bull. 1974;12(1):90–94.
divgrad A mJi (5.28) 16. Lopes da Silva FH, Hoeks A, Smits A, et al. Model of brain rhyth-
mic activity. Kybernetik. 1974;15:27–37.
17. Lopes da Silva FH, Van Rotterdam A, Barts P, et al. Models of neu-
In an infinite medium with constant and , the solution of ronal populations. The basic mechanisms of rhythmicity. Prog
equation 5.28 can be proved to be (121): Brain Res. 1976;45:281–308.
18. Wilson HR, Cowan JD. A mathematical theory of the functional
m Ji (r ) 3 dynamics of cortical and thalamic nervous tissue. Kybernetik.
A(r 0 ) dr (5.29)
4 vol R 1973;13:55–80.
108 Part I ■ Basic Principles
19. Schmitt FO, Dev P, Smith BH. Electronic processing of informa-
tion by brain cells. Science. 1976;193:114–120.
20. Lopes da Silva FH, Storm van Leeuwen W. The cortical alpha
rhythm in dog: the depth and surface profile of phase. In: Brazier
MA, Petsche H, eds. Architectonics of Cerebral Cortex. New York,
NY: Raven Press; 1978.
21. Van Rotterdam A, Lopes da Silva FH, van den Ende J, et al. A
model of the spatial-temporal characteristics of the alpha rhythm.
Bull Math Biol. 1982;44:283–305.
22. Lopes da Silva FH, van Hulten K. Analyse quantitative de l’activite
intercritique en E.E.G. et S.E.E.G. dans l’epilepsie. Rev EEG
Neurophysiol. 1978;8:198–204.
23. Hosek RS, Sances A, Jodat RW, et al. The contributions of intrac-
erebral currents to the EEG and evoked potentials. IEEE Trans
Biomed Eng. 1978;25:405–415.
24. Kavanagh RN, Darcey TM, Lehmann D, et al. Evaluation of meth-
ods for three-dimensional localization of electrical sources in the
human brain. IEEE Trans Biomed Eng. 1978;25:421–429.
25. Schneider MR. A multistage process for computing virtual dipolar
sources of EEG discharges from surface information. IEEE Trans
Biomed Eng. 1972;19:1–12.
26. Schneider MR. Effect of inhomogeneities on surface signals com-
ing from a cerebral current dipole source. IEEE Trans Biomed Eng.
1974;21:52–54.
27. Witwer JG, Tresek GJ, Jewett DL. The effect of media inhomo-
geneities upon intercranial electric fields. IEEE Trans Biomed Eng.
1972;19:352–362.
28. Ary JP, Klein SA, Fender DH. Locations of sources of evoked scalp
potentials; corrections for skull and scalp thicknesses. IEEE Trans
Biomed Eng. 1981;28:447–452.
29. Geddes LA, Baker LE. The specific resistance of biological; materi-
als—a compendium of data for the biomedical engineer and phys-
iologist. Med Biol Eng. 1967;5:271–293.
30. Ferree TC, Eriksen KJ, Tucker DM. Regional head tissue conduc-
tivity estimation for improved EEG analysis. IEEE Trans Biomed
Eng. 2000;47:1584–1592.
31. Gonçalves SI, de Munck JC, Heethaar RM, et al. The application of
electrical impedance tomography to reduce systematic errors in
the inverse EEG problem—a simulation study. Physiol Meas.
2000;21:379–393.
32. Gonçalves SI, de Munck JC, Verbunt JPA, et al. In vivo measure-
ment of the brain and skull resistivities using an EIT-based
method and realistic models of the head. IEEE Trans Biomed Eng.
2003;50:754–767.
33. Law SK, Nunez PL, Wijesinghe RS. High-resolution EEG using
spline generated surface Laplacians on spherical and ellipsoidal
surfaces. IEEE Trans Biomed Eng. 1993;40:145–153.
34. Cuffin BN. Effects of local variations in skull and scalp EEG’s and
MEG’s. IEEE Trans Biomed Eng. 1993;40:42–48.
35. Eshel Y, Witman S, Rosenfeld M, et al. Correlation between skull
reflects the direction of the current densities. IEEE Trans Biomed
Eng. 1995;42:232–249.
36. Haueisen J. Methods of numerical field calculation for neuromag-
netic source localization. Ph.D. thesis, Technische Universität
Ilmenau, Shaker Verlag, Aachen, 1996.
37. Haueisen J, Ramon C, Czapski P, et al. On the influence of volume
currents and extended sources on neuromagnetic fields: a simula-
tion study. Ann Biomed Eng. 1995;23:728–739.
38. Nicholson PW. Specific impedance of cerebral white matter. Exp
Neurol. 1965;13:386–401.
39. De Munck JC. A mathematical and physical interpretation of the
electromagnetic field of the brain. Ph.D. thesis, University of
Amsterdam, the Netherlands, 1989.
40. Bertrand O, Thevenet M, Perrin F, et al. Effects of skull holes on.
In: Proceedings of the 14th International Conference of the IEEE
Engineering in Medicine and Biology Society, Satellite Symposium
on Neuroscience and Technology. Lyon, France; November, 1992.
pp. 42–45.
41. Van den Broek SP. Volume conduction effects in EEG and MEG.
Ph.D. thesis, University of Twente, Enschedé, 1997.
42. Yan Y, Nunez PL, Hart RT. Finite element model of the human
head: scalp potentials due to dipole sources. Med Biol Eng Comput.
1991;29:475–481.
43. Hämäläinen MS, Sarvas J. Realistic conductivity geometry model
of the human head for interpretation of neuromagnetic data. IEEE
Trans Biomed Eng. 1989;36(2):165–171.
44. He B, Musha T, Okamoto Y, et al. IEEE Trans Biomed Eng.
1987;34:406–414.
45. Meijs JWH, Voorde BJ. ten, Peters MJ, et al. The influence of vari-
ous head models on EEGs and MEGs. In: Pfurtscheller G, Lopes da
Silva FH, eds. Functional Brain Imaging. Toronto: Hans Huber;
1988:31–45.
46. Meijs JWH, Peters MJ, Boom HBK, et al. Relative influence of
model assumptions and measurement procedures in the analysis
of the MEG. Med Biol Eng Comput. 1988;26:136–142.
47. Yvert B, Bertrand O, Echallier JF, et al. Improved forward EEG cal-
culations using local mesh refinement of realistic head geometries.
Electroencephalogr Clin Neurophysiol. 1995;95:381–392.
48. Wieringa HJ, Peters MJ. Processing MRI data for electromagnetic
source imaging. Med Biol Eng Comput. 1993;31:600–606.
49. Meijs JWH, Weier OW, Peters MJ, et al. On the numerical accuracy
of the boundary element method. IEEE Trans Biomed Eng.
1989;36:1038–1049.
50. De Munck JC, Peters MJ. Mathematical aspects of biomagnetic
modeling. In: Proceedings of the 8th International Conference on
Biomagnetism. In: Hoke M, ed. Munster: Excerpta Medica; 1991:
11–12.
51. Ferguson AS, Zhang X, Stroink G. A complete linear discretization
for calculating the magnetic field using the boundary element
method. IEEE Trans Biomed Eng. 1994;42:455–459.
52. Kybic J, Clerc M, Abboud T, et al. A common formalism for the
integral formulations of the forward EEG problem. IEEE Trans
Med Im. 2005;24:12–28.
53. Kybic J, Clerc M, Faugeras O, et al. Generalized head models for
MEG/EEG: boundary element method beyond nested volumes.
Phys Med Biol. 2006;51:1333–1346.
54. Van Uitert R, Weinstein D, Johnson C. Volume currents in forward
and inverse magnetoencephalographic simulations using realistic
head models. Ann Biomed Eng. 2003;31(1):21–31.
55. Cuffin BN. EEG localization accuracy improvements using realis-
tically shaped head models. IEEE Trans Biomed Eng.
1996;43:299–303.
56. Roth BJ, Balish M, Gorbach A, et al. How well does a threesphere
model predict positions of dipoles in a realistically shaped head?
Electroencephalogr Clin Neurophysiol. 1993;87:175–184.
57. Menninghaus E, Lütkenhöner B, Gonzalez SL. Localization of a
bipolar source in a skull phantom: realistic versus spherical model.
IEEE Trans Biomed Eng. 1994;41:986–989.
58. Marquardt DW. An algorithm for least squares estimation of non-
linear parameters. J Soc Indust Appl Math. 1963;11:431–441.
59. Schneider MR, Gerin P. Une methode de localisation des dipoles
cerebraux. Electroencephalogr Clin Neurophysiol. 1970;28:68–78.
60. Smith DB, Lell ME, Sideman RD, et al. Nasopharyngeal phase
reversal of cerebral evoked potentials and theoretical dipole impli-
cations. Electroencephalogr Clin Neurophysiol. 1973;34:654–658.
 Chapter 5 ■ Biophysical Aspects of EEG and Magnetoencephalogram Generation
61. Henderson CJ, Buler SR, Glass A. The localization of equivalent
dipoles of EEG sources by the application of electric field theory.
Electroencephalogr Clin Neurophysiol. 1975;39:117–130.
62. Scherg M. Fundamentals of dipole source potential analysis. In:
Grandori F, Hoke M, Romani GL, eds. Auditory Evoked Magnetic
Fields and Potentials. Advances in Audiology. Basal: S Karger; 1990;6:
40–69.
63. Scherg M. Functional imaging and localization of electromagnetic
activity. Brain Topogr. 1992;5:103–111.
64. Stok CJ. The inverse problem in EEG and MEG with application to
visual evoked responses. Ph.D. thesis, University of Twente, The
Netherlands, 1986.
65. Scherg M, Von Cramon D. Two bilateral sources of the late AEP as
identified by a spatiotemporal dipole model. Electroencephalogr
Clin Neurophysiol. 1985;62:32–44.
66. Wood CC. Application of dipole localization methods to source
identification in human evoked potentials. Ann N Y Acad Sci.
1982;388:139–155.
67. De Munck JC. The estimation of time varying dipoles on the basis
of evoked potentials. Electroencephalogr Clin Neurophysiol.
1990;77:156–160.
68. Stok CJ, Spekreijse HJ, Peters MJ, et al. A comparative EEG/MEG
equivalent dipole study of the pattern onset visual response. New
trends and advanced techniques in clinical neurophysiology. EEG
Suppl. 1990;41:34–50.
69. Lopes da Silva FH, Spekreijse H. Localization of brain sources of
visually evoked responses: using single and multiple dipoles. An
overview of different approaches. In: Brumia CHM, Mulder G,
Verbaten MN, eds. Event-Related Brain Research. Amsterdam:
Elsevier; 1991; EEG Suppl. 42:38–46.
70. Manshanden I, De Munck JC, Simon NR, et al. Source localization
of MEG sleep spindles and the relation to sources of alpha band
rhythms. Clin Neurophysiol. 2002;113(12):1937–1947.
71. Schwartz DP, Badier JM, Vignal JP, et al. Non-supervised spatio-
temporal analysis of interictal magnetic spikes: comparison with
intracerebral recordings. Clin Neurophysiol. 114(3):438–449.
72. Van’t Ent D, Manshanden I, Ossenblok P, et al. Spike cluster
analysis in neocortical localization related epilepsy yields clinically
significant equivalent source localization results in magnetoen-
cephalogram (MEG). Clin Neurophysiol. 2003;114(10):
1948–1962.
73. Baillet S, Mosher JC, Leahy RM. Electromagnetic brain mapping.
IEEE Signal Process Mag. 2001;18(6):14–30.
74. Gross J, Kujala J, Hämäläinen M, et al. Dynamic imaging of coher-
ent sources: studying neural interactions in the human brain. Proc
Natl Acad Sci U S A. 2001;98:694–699.
75. Robinson SE, Vrba J. Functional brain imaging by synthetic aper-
ture magnetometry (SAM). In: Yoshimoto T, Kotani M, Kuriki S,
et al., eds. Recent Advances in Biomagnetism. Sendai, Japan: Tohoku
University Press; 302–305.
76. Dale AM, Liu AK, Fischl BR, et al. Dynamic statistical parameter
mapping: combining fMRI and MEG for high resolution imaging
of cortical activity. Neuron. 2000;26:55–67.
77. Mosher JC, Leahy RM. Source localization using recursively and
projected (RAP) MUSIC. IEEE Trans Signal Process. 1999;47:
332–340.
78. Mosher JC, Lewis PS, Leahy RM. Multiple dipole modeling and
localization from spatiotemporal MEG data. IEEE Trans Biomed
Eng. 1992;39:541–557.
79. Mosher JC, Baillet S, Leahy RM. EEG source localization using
multiple signal classification approaches. J Clin Neurophysiol.
1999;16:225–238.
109
80. Baillet S, Garnero L. A Baysenian approach to introducing
anatamo-functional priors in the EEG/MEG inverse problem.
IEEE Trans Biomed Eng. 1997;44:374–385.
81. Hämäläinen MS, Ilmoniemi RJ. Interpreting measured magnetic
fields of the brain: estimates of current distributions. Technical
Report TKKFA559. Helsinki: University of Technology; 1984.
82. Hämäläinen MS, Ilmoniemi RJ. Interpreting magnetic fields of the
brain: minimum norm estimates. Med Biol Eng Comp. 1994;
32:35–42.
83. Wang JZ, Williamson SJ, Kaufman L. Magnetic source images deter-
mined by a lead-field analysis: the unique minimum-norm least-
squares estimation. IEEE Trans Biomed Eng. 1992;39(7):665–675.
84. Wang JZ. Minimum-norm least-squares estimation: magnetic
source for a spherical model head. IEEE Trans Biomed Eng.
1993;40(4):387–396.
85. Grave de Peralta Menendez R, Hauk O, Gonzalez Andino S, et al.
Linear inverse solutions with optimal resolution kernels applied
to electromagnetic tomography. Hum Brain Map. 1997;5:
454–467.
86. Pascual-Marqui RD, Michel CM, Lehmann D. Low resolution elec-
tromagnetic tomography: a new method for localizing electrical
activity in the brain. Int J Psychophysiol. 1994;18:49–65.
87. Pascual-Marqui RD, Esslen M, Kochi K, et al. Functional imaging
with low-resolution brain electromagnetic tomography
(LORETA): a review. Methods Find Exp Clin Pharmacol. 2002;
24(suppl C):91–95.
88. Hjorth B. An online transformation of EEG scalp potentials into
orthogonal source derivations. Electroencephalogr Clin
Neurophysiol. 1975;39:526–530.
89. Le J, Gevins AS. Method to reduce blur distortion from EEGs using
a realistic head model. IEEE Trans Biomed Eng. 1993;40:517–528.
90. Le J, Menon V, Gevins AS. Local estimate of surface Laplacian der-
ivation on a realistically shaped scalp surface and its performance
on noisy data. Electroencephalogr Clin Neurophysiol. 1994;26:
193–199.
91. Babiloni F, Babiloni C, Carducci F, et al. A high resolution EEG
method based on the correction of the surface Laplacian estimate
for the subject’s variable scalp thickness. Electroencephalogr Clin
Neurophysiol. 1997;103:486–492.
92. Nunez PL. Electrical Fields of the Brain. New York, NY: Oxford
University Press; 1981.
93. Zanow F. Realistically shaped models of the head and their appli-
cation to EEG and MEG. Ph.D. thesis, University of Twente,
Enschedé, 1997.
94. Knösche Th. Solutions of neuroelectromagnetic inverse problems
an evaluation study. Ph.D. thesis, University of Twente, Enschedé,
1977.
95. Logothetis NK, Wandell BA. Interpreting the BOLD signal. Annu
Rev Physiol. 2004;66:735–769.
96. Berens P, Keliris GA, Ecker AS, et al. Feature selectivity of the
gamma-band of the local field potential in primate primary visual
cortex. Front Neurosci. 2008;2(2):199–207.
97. Devor A, Dunn AK, Andermann ML, et al. Coupling of total
hemoglobin concentration, oxygenation, and neural activity in rat
somatosensory cortex. Neuron. 2003;39(2):353–359.
98. Gotman J, Kobayashi E, Bagshaw AP, et al. Combining EEG and
fMRI: a multimodal tool for epilepsy research. J Magn Reson
Imaging. 2006;23(6):906–920.
99. Shmuel A, Leopold DA. Neuronal correlates of spontaneous fluc-
tuations in fMRI signals in monkey visual cortex: implications for
functional connectivity at rest. Hum Brain Map. 2008;29(7):
751–761.
110 Part I ■ Basic Principles
100. Cohen, D. Magnetoencephalography. Evidence of magnetic fields
produced by alpha rhythm current. Science. 1968;161:784–786.
101. Reite H, Zimmerman JE, Edrich J, et al. The human magnetoen-
cephalogram: some EEG and related correlations. Electro-
encephalogr Clin Neurophysiol. 1976;40:59–66.
102. Geselowitz DB. Magnetocardiography: an overview. IEEE Trans
Biomed Eng. 1979;26:497–504.
103. Murakami S, Zhang T, Hirose A, et al. Physiological origins of
evoked magnetic fields and extracellular field potentials produced
by guinea-pig CA3 hippocampal slices. J Physiol. 2002;544(pt 1):
237–251.
104. Cohen D, Cuffin BN. Demonstration of useful differences
between magnetoencephalogram and electroencephalogram.
Electroencephalogr Clin Neurophysiol. 1983;56:38–51.
105. Okada YC. Inferences concerning anatomy and physiology of the
human brain based on its magnetic field. Il Nuovo Cimento D.
1983;2:379–409.
106. Cuffin BN. A comparison of moving dipole inverse solutions using
EEG’s and MEG’s. IEEE Trans Biomed Eng. 1985;32: 905–910.
107.Modena I, Ricci GB, Barbanera S, et al. Biomagnetic measurements
of spontaneous activity in epileptic patients. Electroencephalogr
Clin Neurophysiol. 1982;54:622–628.
108. Hari R, Ilmoniemi R. Cerebral magnetic fields. CRC Crit Rev
Biomed Eng. 1986;14:93–126.
109. Hämäläinen MS. Magnetoencephalography: a tool for functional
brain imaging. Brain Topogr. 1992;5:95–103.
110. Lewine JD, Orrison W Jr. Magnetoencephalography and mag-
netic source imaging. Functional Brain Imaging. St. Louis, MO:
Mosby Year Book; 1995:369–417.
111. de Jongh A, de Munck JC, Gonçalves SI, et al. Differences in
MEG/EEG epileptic spike yields explained by regional differences
in signal to noise ratios. J Clin Neurophysiol. 2005;22(2):153–158.
112. Plonsey R. The nature of sources of bioelectric and biomagnetic
fields. Biophys J. 39:309–315.
113. Van Rotterdam A. Electric and magnetic fields of the brain: a sys-
tems analysis approach. Ph.D. thesis, University of Amsterdam,
The Netherlands, 1986.
114. Nunez PL. The brain’s magnetic field: some effects of multiple
sources on localization methods. Electroencephalogr Clin
Neurophysiol. 1986;63:75–85.
115. Stok CJ, Kouijzer WJJ, Peters MJ. Source localization based on
EEG’s and MEG’s. In: Proceedings of the 5th Conference on
Biomagnetics. Vancouver, Canada; 1984.
116. Leahy RM, Mosher JC, Spencer ME, et al. A study of dipole local-
ization accuracy for MEG and EEG using a human skull phan-
tom. Electroencephalogr Clin Neurophysiol. 1998;107:159–173.
117. Dale AM, Sereno MI. Improving localization of cortical activity
by combining EEG and MEG with MRI cortical surface recon-
struction: a linear approach. J Cogn Neurosci. 1993;5:162–176.
118. Cohen D, Cuffin BN, Yunokuchi K, et al. MEG versus EEG local-
ization test using implanted sources in the human brain. Ann
Neurol. 1990;28:811–817.
119. Hansen P, Kringelbach M, Salmelin R, eds.MEG: An Introduction
to Methods. New York, NY: Oxford University Press; 2009.
120. Lopes da Silva FH. Electrophysiological basis of MEG signals. In:
Hansen P, Kringelbach M, Salmelin R, eds. MEG: An Introduction
to Methods. New York, NY: Oxford University Press; 2009.
121. Plonsey R. Bioelectric Phenomena. New York, NY: McGraw-Hill;
1969.
122. Rosenfalck R. Intra and extracellular potential fields of active
nerve and muscle fibres. A physicomathematical analysis of dif-
ferent models. Acta Physiol Scand Suppl. 1969;321:1–168.
123. Jackson JD. Classical Electrodynamics. New York, NY: Wiley;
1962.
124. Rall W. Electrophysiology of a dendritic neuron model. Biophys J.
1962;2:145–167.
125. Van Rotterdam A. A one-dimensional formalism for the compu-
tation of extracellular potentials: linear systems analysis applied
to volume conduction. In: van Eijnsbergen B, Lopes da Silva FH,
eds. Progress Report No. PR6. Utrecht: Institute of Medical
Physics TNO; 1978:115–122.
126. Plonsey R. Capability and limitations of electrocardiography and
magnetocardiography. IEEE Trans Biomed Eng. 1972;19:239–244.

Analog Signal Recording Principles
ANALOG AND DIGITAL RECORDING
Digital analysis and storage have transformed modern neuro-
physiology in ways that could only be imagined and envied by
the founders of EEG. But the “front end” of all recording sys-
tems, the portion that interfaces with the patient, remains ana-
log; and understanding the operation of analog components
can also give insight into aspects of signal processing that are
now performed mostly by digital techniques. Analog refers to
methods of handling data that are physically similar to the orig-
inal signal. For example, the moving pens and ink traces of the
original electroencephalographs produced waveforms that mir-
rored the electrical potential differences on the scalp, and the
wavy grooves of old-fashioned vinyl records follow the frequen-
cies and amplitudes of the sound waves produced by stereo sys-
tems. Simple amplification of the microvolt potentials in EEG is
an analog process. Analog data “looks like” the original signal,
which the zeroes and ones of digital representations do not.
Ideally, each type of data processing would produce a perfect
reflection of the original waveforms. In practice, each is subject
to its own limitations and potential for inaccuracy, which will
be discussed in the following chapters. More detailed discussion
of the material here can be found online in the Guidelines of
the American Clinical Neurophysiology Society (1) and in those
of the International Federation of Clinical Neurophysiology (2).
THE HOSTILE WORLD OF CLINICAL
NEUROPHYSIOLOGY
Figure 6.1 shows the approximate amplitudes of electromyo-
graphic activity (EMG), EEG, and evoked potential signals in
relation to the many sources of noise that constantly threaten to
overwhelm them. This vast range of voltages requires the use of
a logarithmic scale. So, for example, ECG recorded from the
head can be a thousand times larger than a brainstem evoked
potential. Within electrified buildings, the 50- or 60-Hz electric
field on human bodies can be 10,000 times larger than the EEG.
Some physiologic signals, such as EMG and the standing poten-
tial on the retina, can also represent noise in relation to smaller
signals that we need to see behind them.
At first glance, it may seem impossible that any interpretable
recording could be performed through noise that is orders of
magnitude larger than the signal of interest. This chapter dis-
cusses some of the ubiquitous artifacts in Figure 6.1, the fea-
tures of the analog front end that are designed to defeat them
(while allowing faithful recording of the desired signal) and the
ways in which they may nonetheless make an unwanted
CHAPTER
CHARLES M. EPSTEIN 6
appearance in EEG and other neurophysiologic recordings. The
major entry points and barriers to noise consist of the electrode
interface, the amplifier inputs, and analog filters. The following
sections describe from front to back the different components
of the system and the features that contribute to accurate
recording of neurophysiologic signals.
ELECTRODES
EEG and other recording electrodes provide an interface
between lead wires, which conduct current as the flow of elec-
trons and human tissues, which transmit it through the move-
ment of ions. Because cutaneous oils and keratin are good
insulators, placing dry metal electrodes directly on dry skin is
usually ineffective. Keratin is often removed by rubbing the
skin gently with mild abrasives, and sometimes oils are
removed separately with alcohol wipes. (Despite such prepara-
tions, the epidermis remains the single greatest barrier to cur-
rent flow between the body and the recording system.) In
addition an ionic solution, or electrolyte, is placed between the
skin surface and the electrode. As with human extracellular
fluids, the most abundant ions in the electrolyte are sodium
and chloride. The consistency of the electrolyte ranges from
free-flowing solutions to thick sticky pastes. The latter are used
to help hold cup electrodes mechanically against the scalp. The
former are chosen when the cup is attached by other means
and must be filled with electrolyte through a small hole in the
middle.
Contact between the electrolyte and the electrode causes
opposite charges to line up at the metal–ionic boundary. The
simplest model for this arrangement is the “electrical double
layer,” first proposed in the 1850s by Helmholtz. The double
layer contributes to the stability of the interface and also rep-
resents a capacitance that augments the apparent current
flow. At times, especially with new electrodes, the electrical
double layer undergoes sudden, poorly understood transi-
tions that produce electrode “pops” and a variety of bizarre
oscillations.
The junction between the metal electrode and the ionic elec-
trolyte also develops a half-cell potential, which reflects the
electrical force of incipient chemical reactions between them. It
also represents half of an electrical battery. The half-cell poten-
tial can be much smaller than that in a flashlight battery and
still be enormously larger than the EEG. Ideally this would not
be problem, because the half-cell potentials on any pair of
matching electrodes would be identical, and thus would simply
cancel out when those electrodes are connected to an amplifier.
111
112 Part I ■ Basic Principles

Intracranial strip and depth electrodes are generally made

from stainless steel or platinum. The more favorable electrical
properties of platinum tend to be offset by its greater cost.
With modern amplifiers the high impedance and polarization
of stainless steel represent less of a problem than they did in
the past.

LEAD WIRES, JACKBOX, AND CABLES

The lead wires, jackbox, and cables are basically passive compo-
nents whose function is transparent. Excessive separation of the
lead wires should be avoided. The capacitance of extremely long
cables can attenuate very high frequencies. These effects can be
ameliorated by placing the amplifiers inside the jackbox or even
directly on the electrodes and by using fiberoptic connections
to replace electrical cables. In exceptional circumstances, such
Figure 6.1 Potential sources of noise in neurophysiologic recording. as inside the bore of a high-field MRI system, resistors should
Sources may represent noise in some contexts and signals in others. be placed in series with every electrode lead to prevent induc-
Note the logarithmic scale. tion of excessive current. When troubleshooting, it is worth
remembering that if things mysteriously go wrong, the com-
monest culprit is a bad lead wire, cable, or connector.
However, half-cell potentials can be altered in a number of
ways, including drying out of the electrolyte, its dilution by ELEMENTARY ELECTRONICS
sweat or other fluids, relative differences in temperature, or
exposure of different metallic surfaces by wear and tear. These Classically the building blocks of the EEG front end are ampli-
potentials are also affected by movement of the electrode fiers, filters, switches, and potentiometers. The top portions of
against the electrolyte and the skin, which disrupts the electri- Figures 6.2 through 6.6 sketch the simplified “transfer functions”
cal double layer. The effect of movement is reduced by using of an amplifier, a voltage divider, a high filter (which, confusingly,
cup-shaped electrodes, which hold a larger reservoir of moist
electrolyte.
Chemical interactions at the electrode surface consist of
redox reactions, in which an electron is donated or removed
from an ion. Since the product of such interactions is com-
monly a toxic species, such as hydroxyl ions or chlorine gas, care
must be taken to avoid injury of the underlying tissues. The
safest operating zone for recording and stimulating electrodes is
one where they carry only a limited alternating or balanced
pulse current, small enough to be transmitted entirely through
the double-layer capacitance rather than through redox reac-
tions. Passage of direct current through the double layer will
more readily produce chemical changes at the metal surface,
including electrical polarization, which impedes the free flow of
current in one direction more than another, increases imped-
ance at low frequencies, and may be associated with a substan-
tial change in the half-cell potential.
Scalp and other cutaneous electrodes are usually made
from gold—which is inert and therefore produces smaller
half-cell potentials—or silver coated with a layer of silver Figure 6.2 Amplifier. The top sketch shows the amplifier as just a
chloride. Disposable stick-on electrodes almost always use “black box” that magnifies the size of the signal. At bottom left is the
a silver–silver chloride interface. Combined with a chloride standard circuit symbol for an amplifier, with two inputs on the left and
electrolyte, the latter represents the most stable electrode an output on the right. At bottom right is a graph of the output. It shows
available for neurophysiologic applications, resistant to polar- a straight line representing a gain of 2.0 at the top of the y-axis, because
ization and with the best characteristics for low-frequency and the gain is the same at all frequencies. This simple “single-ended”
direct-current recording. Occasionally the exposure of metallic amplifier has asymmetric inputs: one terminal is “hot” and the other is
silver to the light pulse from photic stimulators produces usually connected to system ground (the bent trident symbol). The
synchronous photoelectric discharges, which may be puzzling inherent asymmetry prevents accurate rejection of noise signals, even
to novice interpreters. when they are equal at both inputs.

Figure 6.3 Resistance bridge. The circuit symbol for a resistor [R] is a
cylinder. The two resistors here are labeled R1 and R2, and both have
the same value. Current from the AC input signal at the top flows
through the resistors to system ground (the bent trident symbol) at the
bottom. The output voltage graph is a straight line with a gain of 0.5,
because a pure resistive circuit has the same gain at all frequencies.
an engineer would call a low pass filter), and a low filter (LF)
(which an engineer would call a high pass filter). The amplifier
(Fig. 6.2) makes the desired signal larger. The purpose of the LF
is to block frequencies so low that they carry more noise than
signal, and thus are better removed. The high filter blocks fre-
quencies so high that they carry more noise than signal. These
simple functions remain the basic components of analog
recording. For this discussion the amplifier will remain mostly
a “black box,” which is the way electrical engineers prefer to use
modern integrated circuits anyway. The voltage divider and fil-
ters are represented in the form of their simplest analog com-
ponents: resistors and capacitors.
Figure 6.4 Low filter (LF). The low-frequency waveform coming out is
smaller than the one going in. The circuit symbol for the capacitor (C)
is a pair of parallel lines. Current from the AC input signal at the top
flows through the capacitor and the resistor [R] to system ground at the
bottom. The log–log output voltage graph shows a horizontal line indi-
cating a gain of one on the right side, where frequencies are above the
cutoff. Below the cutoff frequency, gain drops steadily as frequency
increases.
Chapter 6 ■ Analog Signal Recording Principles 113
Figure 6.5 Sketch of how changing frequencies produce the LF behav-
ior graphed in Figure 6.4. At low frequencies (A) the capacitor imped-
ance becomes very high, so the capacitor and signal source practically
drop out of the circuit. At high frequencies (B) the capacitor impedance
becomes very low, and the resistor practically drops out of the circuit.
Amplifiers (Fig. 6.2) have the primary purpose of magnify-
ing the signal from a range measured in microvolts up to sev-
eral volts, which can do something useful such as moving
mechanical pens or undergoing analog-to-digital conversion to
a stream of numbers. The ratio of the signal coming out of the
amplifier to its size going in is called gain. For both analog and
digital systems gain should be in the range of 100,000 or more.
In neither case, however, is gain easily apparent from the system
output. This is because the EEG signal is universally shown in
terms of sensitivity, the ratio of microvolts to millimeters on a
paper or electronic display.
Figure 6.3 shows a simple voltage divider or potentiometer
in the form of a two-resistor network. A resistor is a circuit ele-
ment with two characteristics: it partially blocks the flow of cur-
rent according to Ohm’s law (voltage = current resistance),
and this relationship is independent of signal frequency.
Potentiometers are useful when for any reason voltage must be
reduced rather than increased (i.e., the gain should be less than
one). The signal enters the two-resistor network as the AC volt-
age source at the top and passes through the resistors to system
ground. According to Ohm’s law, if the resistors share the same
current and have equal values, the voltage output at the middle
of the network must be half the voltage at the top. Since the
function of the resistors is the same at all frequencies, the graph
shows a steady gain of 0.5 all the way across.
The middle and bottom of Figure 6.4 show an LF composed
of one resistor and one capacitor. A capacitor consists basically of
two metal plates that are close together but separated by an insu-
lator. Current cannot cross the insulator directly, but positive
and negative charges attract and repel electrostatically on the
plates, producing AC current flow in and out of the capacitor.
The behavior of the capacitor changes predictably with fre-
quency, according to the equation Z = 1/(2 FC). Here Z repre-
sents the impedance of the capacitor to current flow. Impedance,
which is a generic term for all obstacles to electric current, goes
down as either frequency F or capacitance C goes up.
114 Part I ■ Basic Principles
Figure 6.6 High filter (HF). The circuit is simply an inverted LF. The
high-frequency waveform coming out is smaller than the one going in.
The log–log output voltage graph is a horizontal line for a gain of one
on the left side, where frequencies are below the cutoff. Above the cut-
off frequency, gain drops steadily as frequency decreases.
LF gain changes strikingly with frequency. The right side of
the graph in Figure 6.4 shows a value of 1, meaning that the sig-
nal passing through is unchanged. On the left side of the graph
the signal comes through at lower and lower amplitude as the
frequency drops—essentially, low frequencies are filtered out.
In the middle of the graph is a break point where the imped-
ance of the capacitor matches the resistance of the resistor. This
break point is the nominal filter frequency, cutoff frequency, or
3-dB point.1
The behavior of analog filters far from the cutoff frequency
is sketched for an LF in Figure 6.5. It may be understood most
easily by remembering the equivalence of the capacitor imped-
ance and the resistor resistance at the cutoff point in Figure 6.4.
At lower frequencies the impedance of the capacitor is so high
that the AC signal has less and less influence on the output ter-
minal, and the output approaches a short circuit to system
ground. At higher frequencies, however, the impedance of the
capacitor becomes much lower than the resistance, so that the
system ground has less and less influence on the output
terminal—which is now practically short-circuited to the signal
source. The flat passband with a gain of 1.0 reflects this short
circuit.
The high filter (HF) circuit (Fig. 6.6) and its transfer func-
tion are simply the reverse of the LF. The elegant symmetry of
the filter curves is seen only in log–log displays, which is one
reason for preferring this type of graph over linear representa-
tions. Merging the HF and LF curves would form a complete
passband of 1.0 to 100 Hz.
1Readers familiar with decibel notation may notice that the output of the RC fil-
ter is reduced by only the square root of 2 when the resistive impedance equals
the capacitive impedance, whereas it is attenuated by a factor of 2 in the two-
resistor network. Explaining the difference requires the use of imaginary num-
bers, which will not be attempted here.
AMPLIFIERS FOR NEUROPHYSIOLOGY
In addition to gain, neurophysiologic amplifiers require three
other essential properties: linearity and flatness across the entire
frequency range of possible signals, high common-mode signal
rejection, and high input impedance. Any reasonable commer-
cial device should have excellent linearity. The other require-
ments are explained below.
Removing the enormous fields caused by AC wiring—and
by other electrical noise (Fig. 6.1)—is the purpose of the
Faraday cages used with the earliest electroencephalograms and
of shielded rooms still required for modern magnetoen-
cephalograms. As an alternative, patients have occasionally
been connected directly to earth ground. Faraday cages are
unacceptably clumsy and grounding patients is prohibitively
dangerous in medical environments, so more practical alterna-
tives have been found. The key to effective recording of small
neurophysiologic signals is the fact that, large as the noise may
be, most sources of electrical noise are of similar amplitude in
nearby regions of the body. Thus the ECG potentials at adjacent
EEG electrodes are almost identical—at least, in most normal
individuals. The large AC field produced by capacitive coupling
between protoplasm and AC power systems is similar all over
the body. Potentials that are similar at different electrode sites
are called common-mode signals. Common-mode signals are
removed by very precisely subtracting the whole combination
of potentials at one electrode site from the combination of
potentials at another, leaving behind only those potentials
that are different at the two “hot” electrodes—the so-called
differential-mode signal. This process is illustrated in Figure 6.7.
The simpler amplifiers used in radios, stereos, and other famil-
iar electronics (sometimes referred to as single-ended ampli-
fiers) are incapable of the very precise subtraction needed, so
that all neurophysiologic amplifiers have some version of the
symmetrical inputs in Figure 6.7. The latter are called differen-
tial, balanced, or instrumentation amplifiers to distinguish
them from the single-ended variety. The two “hot” inputs in
Figure 6.7 Differential amplifier with balanced “hot” inputs and sys-
tem ground. The vertical symmetry of this more complicated circuit
allows precise rejection of large common-mode signals that are equal at
both “hot” inputs. The relatively small differential-mode signal—the
small square at the far left—is the only signal that passes through the
amplifier.

Figure 6.7 are symmetrical and equally active, despite the uni-
versal practice of labeling one of them as “active” and the other
as “reference” in many EEG recording montages and in
EMG/NCV studies. Their electrical behavior is indistinguish-
able; the only difference is in terminology. The “ground” or
neutral connection in Figure 6.7 is normally present during all
neurophysiologic recording (although modern amplifiers may
appear to function passably without it).
In EEG, a single “ground” or system reference is shared by all
recording channels and is almost universally placed on the fore-
head. The ground must be distinguished from the recording ref-
erence, used in constructing referential EEG montages. The
“reference” in referential montages is the second hot terminal of
the three-terminal differential amplifier, not ordinarily the
“ground.” For digital EEG, the initial recording is most often
performed using a recording reference in the vicinity of the
central-parietal region. With modern electronics, it is possible
to use a single digital recording reference (or one of the regular
scalp electrodes) rather than a separate recording reference and
ground. But noise rejection will be better using both, provided
there is provision for them in the jackbox.
The classic differential amplifier takes the common-mode
noise signal as it comes—whatever its size—and then attempts
to subtract it out. But smaller common-mode voltages would
make the task easier. A more recent technique takes the common-
mode signals seen by all of the amplifiers on the head, averages
and inverts them, and injects the result back into the body
through the former “ground” electrode. Since the necessary cur-
rents are measured in microamperes, no risk is involved. This
method, still referred to anachronistically from its original ECG
usage as “driven right leg,” can cancel and reduce excess noise in
some applications. It also removes the “ground” label and with it
the possibility for a misunderstanding to harm the patient (see
section “Electrical Safety and Ground”).
The ability of an amplifier to pass differential-mode signals
and block common-mode signals is described by its common-
mode rejection ratio (CMRR) and is usually expressed in deci-
bels (dB). Thus an amplifier that reduces the common-mode
signal by a factor of 100,000 (105) relative to the differential-
mode signal has a CMRR of 100 dB. Modern integrated circuit
amplifiers can come off the assembly line with CMRR specifi-
cations of 120 dB (1,000,000:1) or higher. However, many
extraneous factors can defeat them and allow noise to get
through despite the nominal superb performance of the differ-
ential amplifier. One mundane consideration is simply that
common-mode noise is not absolutely identical at different
electrode sites, and the difference grows larger as the electrodes
are moved farther apart. Another example is the induction of
AC potentials in the electrode lead wires if they become widely
separated and form an open loop. A third example is the stim-
ulus artifact from the electrical stimulators used in nerve con-
duction studies and evoked potentials. At hundreds of volts, the
latter can produce an enormous difference at the two hot
amplifier inputs. And very commonly, as discussed below, poor
electrode application can defeat the differential amplifier.
An ideal amplifier senses electrical potentials without any
current actually passing through the inputs. This perfect device
Chapter 6 ■ Analog Signal Recording Principles 115
would have infinite input impedance. Although perfect ampli-
fiers do not exist, the highest practical impedance is preferred
for neurophysiologic recording. Figure 6.8 shows why. Imagine
an amplifier in which the two inputs have an impedance of
10,000,000 , connected to two “hot” electrodes with imped-
ances Z E1 of 1,000 and ZE2 of 10,000 , respectively. Both elec-
trodes are picking up an equal 60 Hz common-mode noise
signal of 10 mV. Assuming that the input impedances of the
amplifier are still much larger than the electrode impedances,
the simplified equation in Figure 6.8 allows calculation of the
result. The unequal electrode impedances will produce a differ-
ential-mode noise signal at the amplifier inputs of 0.01
(10,000 – 1,000)/10,000,000 = 0.000009 V or 9 V. This noise
will be faithfully magnified by the differential amplifier to pro-
duce visible widening of the EEG tracings—just as seen in
Figure 6.9B.
Since it would be impossibly cumbersome to make all the
electrode impedances exactly equal, we make sure all electrode
impedances are low—in which case the difference between
them cannot get too large—and make the amplifier input
impedances as high as practical. Recent guidelines specify that
the differential input impedance of the amplifier must be at
least 100 M (1). The CMRR should be at least 80 dB
(10,000:1) at the highest sensitivity of the amplifier when the
common-mode signal is applied between both inputs and neu-
tral (1). Amplifier input impedances higher than this can be
obtained without much difficulty, but their performance is
liable to be degraded by the capacitance between conductors in
the lead wires and cables.
If the only problem caused by unbalanced (i.e., excessively
large) electrode impedance were 50 or 60 Hz noise, it would be
easy to deal with using notch filters. But Figure 6.9 illustrates a
Figure 6.8 Model of the differential amplifier showing the impedances
of the electrodes and the two amplifier inputs. Making some reasonable
approximations, the equation at the upper right represents the way in
which unequal electrode impedances ZE1 and ZE2 convert a portion of
the common-mode noise voltage VCM into a differential-mode voltage
VDM, which passes through the amplifier.
116 Part I ■ Basic Principles

Figure 6.9 ICU EEG study with ground recording artifacts in the F4 channels, including apparent slowing and a spike at
the tip of the arrow (A). 60-Hz noise with the notch filter off (B) is an important clue to the existence of a serious prob-
lem. See text for further discussion.

far more insidious consequence. Panel A of this ICU recording
appears to show focal delta slowing involving electrode F4, cul-
minating in an apparent spike-wave discharge. Panel B, with the
60 Hz notch filter turned off, indicates that in fact there is a
poor, high-impedance connection at F4 and that 60-Hz com-
mon-mode AC noise has entered the involved channels. The
crucial message of Figure 6.9 is that the apparent focal abnor-
malities also represent common-mode noise. The F4 slowing is
actually time-locked to the ECG and reflects an obscure compo-
nent of the t-wave. The apparent spike is due to the small pha-
sic EMG artifact seen in other leads, picked up by the ground
lead on the forehead and magnified across the now-unbalanced
amplifier inputs. Recognizing the presence of ground recording
artifacts, and the importance of AC power line artifacts in
assisting that recognition, is vital exactly because these artifacts
may not be obvious. In several decades of teaching EEG, none
of our residents or fellows has ever recognized or deciphered
ground recording independently. Recording routinely with the
AC notch filter on badly skews the odds that many electroen-
cephalographers ever would.
One corollary of the preceding discussion is noteworthy
for being even more counterintuitive. Poor, high-impedance
 Chapter 6 ■ Analog Signal Recording Principles 117

electrode connections do not generally reduce the size of the
EEG signal. Indeed, as in Figure 6.9, channels 9 to 10, they
commonly make it look bigger.
RANDOM BACKGROUND NOISE
In addition to the physiologic and man-made artifacts
described above, several sources of random noise can interfere
with recording of very small electrical signals. Of these the
most irreducible is thermal or “Johnson” noise, which repre-
sents Brownian movement of electrons in resistive conduc-
tors. The magnitude of thermal noise increases with
temperature, circuit resistance, and recording bandwidth. At
body temperature, with a high filter of 70 Hz and total elec-
trode resistances of 1000 , thermal noise is about 0.035 V
and unlikely to make any difference in recording. But in a
brainstem auditory evoked potential (BAEP) or concentric
needle EMG study, high electrode resistance can raise the
minimum noise to the approximate magnitude of the desired
signal. Some examples of possible recording conditions are
given in Table 6-1. Note that other noise sources in the ampli-
fier may increase actual noise figures several times above those
listed. Recent guidelines specify that the noise level of an EEG
amplifier must not exceed 2 V rms with the inputs con-
nected to neutral and with a bandpass of 0.1 to 5000 Hz (1).
ELECTRICAL SAFETY AND GROUND
earth ground at the power station.2 In modern neurophysiology,
this feature of the power system is the most important concept
for electrical safety. To operate electrical devices plugged into
the AC wall jack, current flows from the “hot” wire of the power
line to the “neutral” wire at earth ground. If patients are con-
nected to earth ground, any stray currents (especially those from
the hot power line) will happily flow through the patient on the
way there. A third wire, which represents a second and more
direct earth ground connection, is required in hospital equip-
ment as well. This additional ground, commonly referred to as
“the” ground, increases safety in some situations—but also con-
stitutes a risk if the patient is connected to it. Modern safety
standards require that the recording ground on the patient cannot
form a direct connection to the power line ground(s) or to other
sources of earth ground, including “the” ground wire.
Traditionally discussions of electrical safety have focused on
“ground loops,” which may be formed when the patient is con-
nected to more than one ground electrode through different
pieces of medical equipment. Decades ago, currents flowing
through the power line ground system could place the different
components of that system at somewhat different voltages—in
which case the voltages would attempt to equalize through the
body of the patient. But the modern requirement that record-
ing ground on the patient must never be connected to earth
ground means that ground loops can seldom present a safety
risk using commercial systems. However, ground loops may
contribute substantially to electrical noise.

In electronic circuits “ground” is a reference point that remains

FULL ANALOG RECORDING
at a virtual zero voltage. Without such a reference point, circuit
operation easily becomes unstable. Many years ago the circuit Prior to the digital era, the vast amount of information con-
ground, and with it the ground electrode on the head, was actu- tained in even a routine EEG could be stored only on paper. The
ally connected to real earth ground through metal plumbing or original signal was amplified up to 1,000,000 times, enough to
other low-resistance conduits. Earth ground is a wonderful power an electromagnetic coil that moved a long mechanical
zero-reference point—a near-infinite reservoir of electrons— pen which carried ink onto fanfold paper. Mechanical pens
which will never be changed by local circuit operation. It is also could move only so fast, setting the upper frequency limit
an unacceptable connection for human bodies, especially in around 70 Hz. They could move only so far, causing the wave-
medical environments. This is because one of the two wires in forms to “square-off” at the limits of the pen range (Fig. 6.10).
alternating current power lines is almost always connected to Special circuitry was used to prevent pens from colliding and
tangling together. The fanfold paper advanced from one tray to
Tabl e 6 . 1 another via a mechanical chart drive with rollers, which could
move at speeds of 15, 30, and 60 mm/sec. The paper was prone
Minimum System Noise Under Representative to ripping or jamming. Ordinarily, EEGs were recorded at
Recording Conditions 30 mm/sec on pages 300 mm wide. The resulting scale of
10 sec/page has been carried over to digital systems. The obso-
Bandwidth Electrode Thermal lete designation of paper speed in mm/sec is so engrained in
Modality (Hz) Resistance Noise older technologists and electroencephalographers that even
( ) ( V) now computer displays are labeled anachronistically in the
same manner.
EEG 35 1,000 0.024
In fully analog EEG systems, constructing different mon-
EEG 70 10,000 0.13 tages requires the use of elaborate mechanical switches. Rows of
BAEP 1,000 1,000 0.13 individual switches are also needed for every channel to set
gain, HF, and LF, and AC notch filters. The available range of
BAEP 3,000 10,000 0.72
EMGa 10,000 100,000 4.1 2Grounding of the power line is necessary because of the astonishingly large
voltage gradients that can develop at different elevations in the atmosphere and
aConcentric needle electrode. over long distances across the earth’s surface.
118 Part I ■ Basic Principles
Figure 6.10 Detail of two channels from an analog paper writeout.
During a burst of high-voltage frontal theta, the pen in the second chan-
nel reaches the limits of its mechanical range, causing the waveform to
“square-off” in the vicinity of the shaded area. Within the shaded area
is a very narrow spike-like artifact, apparently caused by the collision-
avoidance system meant to protect the EEG pens. Note the curvature of
the high-voltage waveforms and of the artifact, due to the radial nature
of the pen movement.
settings is consequently limited, most notably to the legendary
HF choices of 15, 35, and 70 Hz. A few generations of “hybrid”
systems are likely to still be in use. In hybrid systems switching
and the associated notations are done electronically, but the
final writeout remains pen and ink.
In order to produce a comprehensive overview of cerebral
activity, several montages must be used during the course of
each analog recording. Because the physical settings are not
noted automatically on the paper record, analog interpreters
rely on the technologist to write by hand every setting and every
change in parameters at the time of recording. Traditionally,
every combination of parameters is then replicated during a
terminal calibration. Producing a true calibration, which faith-
fully displays the performance of the entire pathway from
inputs to writeout, is a virtue of analog recording that not every
digital system properly duplicates. (On the other hand, the
mechanical writeout, the weakest link in the analog recording
chain, has been abolished from digital systems.)
Perhaps the greatest limitation of analog recording is the
dependence on the technologist to make notations accurately,
to recognize abnormalities “online,” and to optimize their dis-
play for the reader by making appropriate changes in montages
and recording parameters. This feature does, however, have a
silver lining. The technologist by necessity is far more engaged
in the recording and its eventual interpretation.
Other virtues of paper should not be entirely dismissed. The
spatial resolution of paper EEG is in the range of 100 dots/cm,
as compared to about 30 pixels/cm for conventional computer
monitors. To give equivalent resolution, while leaving room for
data labeling, digital monitors would have to be about a meter
wide and incorporate pixel counts that only now are being
approached by high-definition television. Furthermore, paper
can easily be unfolded to provide an overview of several min-
utes’ data while maintaining the same exquisite detail. The
inability of computer displays to show more than 10 seconds
worth of data without severe compromise in frequency resolu-
tion is an underappreciated handicap. It can be overcome only
by knowledgeable adjustment of the display on the part of
interpreters who understand its limitations for slowly evolving
activity. Finally, there is no recorded instance of an entire paper
EEG tracing vanishing forever on its way to the server, when
somebody accidentally pushed the “delete” button or when the
operator forgot to hit “record.” And the recent history of digital
recording media suggests that ancient paper records will still be
quite readable when many digital studies can only be viewed in
a museum.
ARTIFACTS UNIQUE TO
ANALOG RECORDING
Artifacts related to the electrode interface, poor common-
mode rejection, and ground recording have been discussed pre-
viously. Mechanical pen writers are subject to radial
displacement due to the pens moving in an arc rather than in a
straight line (Fig. 6.10)—at times causing inaccurate alignment
of high-voltage, sharp-contoured waveforms. The ink may skip
or puddle on the page. The moving paper can transiently slip or
jam on the roller, stretching or compressing the EEG wave-
forms. Figure 6.10 also shows an unusual artifact apparently
caused by the collision-avoidance system in a hybrid electroen-
cephalograph. What appeared at first glance to be spike wave
consists only of high-voltage theta slowing plus a spike-like
artifact. When the pen reaches the limit of movement, it para-
doxically moves across its full range in the opposite direction
and quickly back. We have observed this behavior in several dif-
ferent brands of paper electroencephalograph, where it was
misinterpreted as evidence of epilepsy.
REFERENCES
1. American Clinical Neurophysiology Society Guidelines.
https://www.acns.org.
2. Ebner A, Sciarretta G, Epstein CM, et al. EEG instrumentation.
The International Federation of Clinical Neurophysiology.
[Practice Guideline]. Electroencephalogr Clin Neurophysiol Suppl.
1999;52:7–10.
 CHAPTER

DOUGLAS MAUS, CHARLES M. EPSTEIN, AND SUSAN T. HERMAN
INTRODUCTION AND HISTORY
OF DIGITAL EEG
Digital electroencephalography (EEG) refers to the recording,
storage, and review of the EEG on digital equipment (i.e.,
computers). The process requires conversion of the continuous
analog EEG signal, as described in Chapter 6, into discrete
numerical values. Over the past 15 years, digital EEG has grad-
ually replaced analog systems and is now the standard for clin-
ical EEG. Digital acquisition and processing of EEG allows
many advantages, but at the same time imposes limitations that
should be understood by the clinician to avoid inaccurate inter-
pretation. The practical advantages of digital EEG over analog
EEG are summarized in Table 7.1. In this chapter, we will out-
line the principles of digital signal acquisition and processing
that are important in the design and proper use of clinical EEG.
Clinical digital EEG represents the convergence of two scien-
tific fields that evolved in parallel for the last 70 years. In 1929,
Hans Berger first reported his studies on the electroencephalo-
gram (1). The utility of EEG for clinical diagnosis was quickly
established, and EEG has flourished as a tool of neurologists,
Digital EEG
7
and epileptologists in particular. Practical digital computation
began with the development of ENIAC around 1943 to 1944
(2). As early as the 1950s, the potential advantages of computer
analysis of EEG were foreseen (3). However, the routine use of
computers for EEG acquisition and analysis required several
decades for commercial digital electronics to achieve bench-
marks that equal or surpass the capabilities of analog EEG
recording apparatus. Reliable, power-efficient computation
began with the semiconductor transistor, invented at Bell Labs
by William Shockley in 1947. The development of integrated
circuits in the 1950s permitted mass production and commer-
cialization. A notable computer algorithm for spectral analysis,
the Tukey–Cooley version of the discrete fast Fourier transform
(FFT), was published in 1965 (4). First-generation digital EEG
systems, with limited digitization technology and screen resolu-
tion, could not match the display properties of paper-based
EEG (5), but digital electronics gradually achieved fast sam-
pling rates and larger data throughput. Finally, in the late 1990s,
digitization rates and computer storage available for reasonably
priced digital commercial products were sufficient to match
routine analog EEG machines.

Tabl e 7 . 1

Advantages of Digital EEG

Efficiency Computer-based display generally permits quick review of many studies

Nondestructive processing
Precision
Archiving
Transmission/comparison
Increased frequency range
Reliability
Portability
Digital signal processing
Associated software
Raw data are stored digitally, allowing reconstructed display with different montages, different
filters, and different sensitivity
Can precisely measure frequencies and amplitudes for longitudinal or intergroup comparisons
Easily make exact copies, retain for easy access on disk drive, or save on optical media
(CD, DVD)
Digital copies can be easily and quickly transmitted for second opinion or conferences
Current and near-future digital sampling rates easily exceed the frequency range accessible
by traditional analog EEG
Stable and reliable electronics, controlled by microprocessors, decreases likelihood of technical
artifacts compared to analog equipment requiring manual calibration for each channel
Amplifiers can be small enough for patients to carry during ambulatory and long-term studies
Allows extensive on-line or postacquisition analysis, such as automated detection of spikes and
seizures, topographic maps, and graphical displays of quantitative EEG
May include patient information databases, report writing programs, and integration with
hospital electronic medical records

119
120 Part I ■ Basic Principles
DIGITAL EEG COMPONENTS
Figure 7.1 shows the basic components of the digital EEG sig-
nal path from the electrodes to the display. Several components
(jackbox, differential amplifier, display) parallel those encoun-
tered in analog EEG machines (Chapter 6). The features of
these components that differ from analog EEG will be high-
lighted in the sections below. Components unique to digital
EEG systems include (i) the analog-to-digital converter (ADC),
which samples the analog EEG signal and converts it into dis-
crete digital data; (ii) the computer or microprocessor, which
stores and manipulates the digital data for optimal review;
(iii) monitors for display of EEG data; and (iv) digital storage
devices and networking equipment, which allow archiving of
EEG data and transmission to remote sites. Some systems
contain video processors and sound cards for simultaneous
acquisition and synchronization of video and audio data. The
hardware and software capabilities of current digital EEG sys-
tems vary significantly between vendors. This review cannot
describe all possible digital system specifications but aims to
provide a summary of the most important features.
Electrodes
Along the signal acquisition path, we turn to the first compo-
nent—the electrodes. These are discussed in detail in Chapter
6. One significant difference from analog EEG is the require-
ment for one additional electrode, the machine reference,
which we shall discuss below in the section on (differential)
amplification. As with analog EEG, a good recording depends
on low-impedance electrical contact with the scalp. Most
current commercial EEG acquisition machines have function-
ality for checking electrode impedances. In general, these are
software automated. For each electrode, the impedance is esti-
mated by forming a complete circuit, starting with that elec-
trode, then creating a return path with all the other electrodes
joined (so that section of the circuit has an impedance equiva-
lent to all other electrodes in parallel, and thus small). A small
(microvolt) voltage (typically oscillatory, on the order of 30
Hz) is applied, and the current is measured. The impedance of
that single electrode contact is thereby estimated by the voltage
(known) divided by the current (measured). For standard
scalp electrodes applied with electrolytic paste, impedances
should be less than 5 k . This technique relies, of course, on
the assumption that at least some of the electrodes are well
connected. If all electrodes have poor contact, spurious results
can be obtained.
Electrode
Figure 7.1 The schematic of a digital EEG
machine shows components of the signal
"Jackbox"
path from the electrode to the display.
preamp
Reference
Routine digital EEG has no requirements for specialized elec-
trodes. Current protocols for experimental digital EEG acquisi-
tion at high sampling rate with intracranial implantation may
derive benefit from specialized microelectrodes that have a
smaller contact area and may be able to detect spatially local-
ized multiunit or even single-unit firing trains.
Electrode Input Board or Jackbox
Jackboxes for digital EEG typically have a greater number of
electrode inputs than analog machines. The standard number
of inputs for routine EEG is 32, but jackboxes accepting 64, 128,
or even 256 inputs are common for epilepsy monitoring and
intracranial EEG recordings. Additional inputs can accommo-
date additional EEG electrodes, such as 10-10 modified
International System placements, as well electro-oculogram,
electromyogram, or other polygraphic inputs (6). Jackbox
inputs are labeled with the location of the electrode on the head
according to the International 10-20 System (6) or numbered.
In addition, there are inputs for ground and for a machine ref-
erence (sometimes called recording reference) electrode, which is
usually placed in a location not susceptible to large artifacts,
such as the midline central-parietal region.
In contrast to analog systems using electrode selectors, most
digital amplifiers are “hard-wired,” with the signal from each
jackbox position going to Input 1 of a specific amplifier. Input
2 for all amplifiers is the machine reference electrode. Some
digital systems have electronically switched jackboxes to allow
amplifier input channels to be switched among different elec-
trodes. Because all EEG activity is acquired in relation to the
machine reference electrode, the data in all channels may be
distorted if this reference electrode is of high impedance. With
a high impedance reference, external noise may overwhelm the
EEG signal, producing a very low amplitude recording with lit-
tle signal. (This is different from effects of high impedance in
any other electrode, which can increase the apparent EEG
amplitude through ground reference recording, in addition to
showing the noise.) To prevent this problem, the EEG should be
reviewed during the recording in a machine reference montage
to check for excessive noise in the reference electrode.
Digital EEG jackboxes may be housed in the same enclosure
as other digital components, such as preamplifiers (see the sec-
tion “Amplifiers”). The short distance between the scalp elec-
trode and this first amplification step reduces electrical
interference and artifact in the EEG signal. ADCs may also be
present in the jackbox itself, as in systems for ambulatory
recording (7). The jackbox also contains or is connected to an
Software:
Differential Anti-alias Analog montage
digital filter Display
amplifier filter
converter sensitivity
time-scale
Digital
storage

Paper Pens Low-cut High-cut Montage
(fixed) (fixed) (fixed)
Fp1−F7
Fp1−F7 0.5 Hz 70 Hz
F7−T3
F7−T3 0.5 Hz 70 Hz
T3−T5
T3−T5 0.5 Hz 70 Hz
T5−O1
T5−O1 0.5 Hz 70 Hz
impedance meter to allow measurement of electrode imped-
ance (see the section “Electrodes”) without unplugging the
electrode wires from the electrode input panel.
Amplifiers
The first active components in the signal path are the
(pre)amplifiers. Prior to digitization, the signals—which are in
the microvolt range—must be amplified. This may be done via
one or more cascaded amplifiers. Modern amplifiers are small,
low-power, single-chip multichannel devices with solid-state
integrated circuits. Noise (generated by random movement of
electrons within amplifier circuits such as resistors and semi-
conductors) should be less than 2 V peak to peak. Amplifiers
are electrically or optically isolated to prevent conduction of
current from the EEG machine to the patient through the EEG
electrodes.
Properly designed, the first amplifier (often called a preampli-
fier) has high input impedance (on the order of mega- to tera-
ohms), which allows the source signals to be precisely captured,
without drawing significant current which would create distor-
tions in the voltage to be measured (8). They also raise the signal
into a range where stray device noise (powerline, television,
radio) is negligible and so cannot disturb the signal. Finally, com-
mercial ADCs have an expected input range, such as 5 V. To
take full advantage of the amplitude resolution of the ADC, input
signals must be brought up to this voltage range by amplification
(though not over it, since this would “clip” the signal).
For EEG, one of the most important aspects of the amplifi-
cation stage is the use of differential amplifiers. Both analog and
digital EEG employ differential amplifiers. Differential ampli-
fiers have three input connections: (i) inverting input; (ii) non-
inverting input; and (iii) ground, typically placed on the
forehead, and a single output. The output is the difference
between the two inputs multiplied by some constant—the
amplification factor, or gain. Gain for clinical EEG machines is
usually in the range of 2000 to 20,000, but can be up to
1,000,000. The gain equals output voltage divided by input volt-
age. Gain is typically expressed as a logarithmic ratio in decibels
(dB), with gain (dB) = 20(log(Vout/Vin)). For instance, with
inputs of 106 V (noninverting) and 104 V (inverting), and
amplification factor (gain) of 20, the output will be 40 V.
Chapter 7 ■ Digital EEG 121
Figure 7.2 Schematic of amplifier inputs for analog EEG for
a longitudinal bipolar montage. One additional electrode
+ Fp1
input—the ground—is omitted for simplicity.
−
F7
+
−
T3
+
−
T5
+
− O1
(In our figures, we show the electrical symbol for an op-amp,
short for operational amplifier. Op-amps do amplify the differ-
ence between their inputs, but this is a simplification. Actual
hardware differential amplifiers are more complex circuits, using
a combination of op-amps with other circuit elements, specifi-
cally designed and tuned to amplify the difference in voltage
even when both voltages may be quite large in absolute value.
See more about common-mode rejection below.)
In analog EEG, the montage is fixed at acquisition time by
the fact that the electrodes are mechanically connected to a set
of differential amplifiers for that montage. Figure 7.2 shows a
simple diagram of the arrangement of differential amplifiers for
analog EEG for a longitudinal bipolar montage. Digital EEG
also employs differential amplifiers, but with a significant dis-
tinction. Instead of determining the difference between elec-
trode inputs at acquisition time, each electrode is acquired
relative to one fixed common electrode, the machine reference.
These individual electrode channels are then digitized and
stored. Later review with tracings displayed in different mon-
tages can be performed by digitally reconstructing the mon-
tages. Figure 7.3 shows a simple diagram highlighting this
difference from analog EEG. Note that the vertical dashed line
separates the fixed, unalterable acquired data on the left, from
computations on the right showing the data in one of many
possible different montages, which can be easily reconstructed.
Also note that the input to the inverting contact for each differ-
ential amplifier is the machine reference for all the electrode
channels recorded.
There are several other aspects of amplifiers that could
impact the data acquisition. The common-mode-rejection ratio
(CMRR) is a measure of how well the differential amplifier
reflects the difference between the inputs, especially when both
inputs are large in magnitude. Real differential amplifiers can
be best modeled as generating an output voltage V0 as
1
V0 Ad (V V ) 2As(V V )
The coefficient Ad is the differential gain and As is called the
common-mode gain. The common-mode rejection ratio is
Ad
CMRR 20log10 a b
As
122 Part I ■ Basic Principles
Figure 7.3 Differential amplifiers
for digital EEG, with each electrode
input amplified compared to the +
Fp1
machine reference. Computer soft-
−
ware then converts this referential
montage to a longitudinal bipolar
F7 +
montage. See text for details. The
− Computer Display
ground electrode is omitted for
simplicity.
T3 +
−
T5 + T5−O1
−
O1 +
−
Ref
So, a CMRR of 80 dB, as recommended by American Clinical
Neurophysiology Society (ACNS) (9), indicates that the ampli-
fier will begin to lose accuracy when the absolute magnitude is
10,000 (104) times the difference in magnitudes (e.g., 10,001
V – 10,000 V). Most of the common mode signal is 60 Hz
noise, and the magnitude of that noise is determined more by
the electrode impedances than circuit specifications.
Analog Filters
Prior to digitization, analog electrical filtering is applied. The
fundamentals of analog filter design are discussed in Chapter 6.
In contrast to analog machines, digital EEGs are acquired using
broad bandpass filter settings, typically 0.1 to 70 or 100 Hz.
However, for specialized applications, the bandpass may
include very low frequencies (0.001 Hz) or even direct current
(DC), as well as high frequencies up to 3000 to 10,000 Hz (10).
The most important filter in this stage is the anti-aliasing fil-
ter as described below. This requires a low-pass filter (passes
low-frequency signals but attenuates high-frequency signals).
The simplest low-pass analog filter is a resistor (R) and capaci-
tor (C) in series, with the voltage across the capacitor transmit-
ted to the next stage, as diagrammed in Figure 7.4. This
first-order low-pass filter will attenuate high frequencies and
has a cutoff frequency (in Hz) of 1/(2 RC). Commercial anti-
alias filters may be this simple but are often engineered with
R understanding of digitizer capabilities and limitations is crucial.
Vin C Vout
Figure 7.4 A low-pass analog filter is a resistor (R) and capacitor (C) in
series, with the voltage across the capacitor transmitted to the next stage.
Digital
Transmission
Montage
/Storage
ADC …00101010… ‘Fp1’
‘Fp1−F7’
10111000
ADC …01110010… ‘F7’
‘F7−T3’
00001001 Fp1−F7
ADC …01101001… ‘T3’ F7−T3
‘T3−T5’ T3−T5
01010110
ADC …00010011… ‘T5’
‘T5−O1’
10111111
ADC …01010100… ‘O1’
more components to achieve a faster roll-off in the transition
band, thus a “sharp” cutoff. Such analog filters may be designed
according to Butterworth or Chebyshev topologies (11).
Other filters can be inserted prior to digitization. DC is not
commonly used in clinical EEG but may be used for research
purposes. In most commercial systems, filters are inserted to
remove the DC (constant) and very low-frequency components
( 0.5 Hz), which make the record harder to read and require a
larger magnitude resolution for the ADC. These have the addi-
tional benefit of maximally exploiting the ADC resolution, since
the average signal is then zero, and the greatest positive and neg-
ative excursions can be designed to fit within the digitizer input
range to take full advantage of the magnitude resolution of the
ADC. Another filter that may appear before digitization is a
power line “notch” filter, which is specifically tuned to eliminate
the frequencies matching that of the power line: 60 Hz in the
United States, 50 Hz in Europe and other countries.
Analog-to-Digital Conversion
The stage of conversion of the analog signal to digital represen-
tation is the most crucial step in the acquisition process. Design
of an acquisition system requires that the components be
matched and appropriate as a whole, so that the final data are
not hindered by poor design of any one element in the path.
Because parameters of the digitization impact acquisition com-
ponents both earlier and later in the signal path, thorough
The act of digitization collapses an EEG signal in two independ-
ent axes: time and magnitude.
Time
Typically, a time-varying signal is digitized at fixed (constant)
discrete intervals. The time intervals are described by sampling
rates. Typical sampling rates are on the order of several hundred
samples per second (Hz). For 200 Hz, this implies that samples
are taken every 5 msec. The time between subsequent samples

A Figure 7.5 Aliasing. A: A pure sinusoid at 11 Hz
+1
0 are “folded back” across the Nyquist, as if the paper
0.5
−1
Seconds
0 Nyquist Sample
(8 Hz) rate
is called the dwell time. The choice of sampling rate impacts
how well the obtained data reflect the analog source, with faster
sampling roughly giving more fine detail. A reasonable question
is: How fast is fast enough? A well-known theorem from signal
processing literature provides a quantitative answer to this
question. The Nyquist theorem states that the highest frequency
distinguishable is one half the sampling frequency. The down-
side of more rapid sampling is the creation of larger EEG files.
Armed with the Nyquist theorem, one knows the limitations
of what signal components in the source are capturable, but the
question is naturally raised as to the consequence of signals that
are above this Nyquist limit. If these were simply eliminated, the
issue would be moot, but sadly high-frequency components
above the Nyquist can actually create nefarious artifacts in the
digitized output. Figure 7.5A shows an example of a pure sinu-
soid at 11 Hz (the dashed curve) being sampled at 16 samples
per second (Nyquist frequency of 8 Hz). The sampled data
points can be seen to equivalently match a sinusoid at 5 Hz
Chapter 7 ■ Digital EEG 123
(the solid curve) sampled at 16 samples per second
(Nyquist frequency is then 8 Hz). The sampled data
points can be seen to equivalently match a sinusoid
at 5 Hz (dashed curve). When displayed on a screen
(connected by straight lines), these points would
more closely match the 5-Hz sinusoid than the origi-
nal 11-Hz sinusoid. B: Frequencies above the Nyquist
1 were folded with the crease at the Nyquist fre-
quency—best exemplified by the solid curves of the
frequencies between the Nyquist and the sampling
rate. The heavily darkened curve represents the alias-
ing of 11 to 5 Hz from A.
(solid curve). Indeed, in practice, when displayed on a screen
(connected by straight lines), these points would more closely
match the 5-Hz sinusoid than the original 11-Hz sinusoid. This
phenomenon is known as “aliasing.” Figure 7.5B shows the pat-
tern by which frequencies above the Nyquist are “folded back”
into the range between 0 and the Nyquist frequency, with the
example of 11 Hz being folded back to 5 Hz at sampling rate of
16 samples per second emphasized with the bold arc.
Figure 7.6 shows the deleterious effect of noise above the
Nyquist frequency. Fortunately, methods to eliminate this artifact
are quite simple. Electrical filters—specially designed circuits
composed of elements such as resistors and capacitors—can be
used to eliminate undesired frequencies in the analog signal prior
to digitization. In particular, ADC systems incorporate an analog
low-pass filter prior to digitization, with this filter designed to
pass frequencies below the Nyquist frequency and eliminate
those above it. Such a low-pass filter in this context is termed an
“anti-aliasing” filter.
124 Part I ■ Basic Principles

Spectrum

Original
0 Hz 256
Low−pass filtered
Filter

0 0

0 2.0 0 2.0
Noise folded
Digitize Digitize
128 Hz 128 Hz

0 0

0 Seconds 2.0 0 2.0

Figure 7.6 Deleterious effect of noise above the Nyquist frequency. Top left panel shows 2 seconds of “original” (simu-
lated) data composed of a signal at 5 Hz, plus noise with significant components up to 256 Hz. The inset (just above and
to right) shows the power spectrum of this data, showing the white noise and the large signal at 5 Hz. Bottom left panel
shows what would be the result of naïve direct digitization at 128 samples per second, rendering the signal at 5 Hz nearly
unrecognizable within the abundant noise. The signal-to-noise ratio (SNR) has actually been worsened because the noise
above the Nyquist limit (64 Hz = half of 128 samples per second) has been folded back, adding to the noise in the fre-
quencies below the Nyquist, as illustrated in its power spectrum. The right side panels show how anti-aliasing filtering
prior to digitization can prevent this problem. Top-right panel shows the tracing after a low-pass filter (modeled as
Butterworth third order) with cutoff at 50 Hz has been applied. Bottom-right panel shows the result of digitization (again
at 128 samples per second) of this filtered tracing. Frequencies above the Nyquist (64 Hz) have been attenuated by the
low-pass filter and are therefore not folded back, so digitization in this case yields a faithful reproduction of the signal.

The anti-alias filter and ADC in Figure 7.1 are deliberately
joined. The analog hardware anti-alias filter must obviously
precede the ADC. The anti-alias filter cutoff must be appropri-
ate for the sampling rate of the joined ADC. The filter must
adequately attenuate frequencies above the Nyquist, or else
high-frequency noise will be aliased into the digitized signal. If
the cutoff is too low, then signals with frequencies above this
cutoff and below the Nyquist will be unnecessarily attenuated
and lost, and expense will have been wasted in using an ADC
with higher sampling rate than necessary.
Another point regarding the time axis regards the “frequency
resolution,” a factor important for quantitative analysis of EEG.
The sampling rate does not impact whether one can distinguish a
signal as being one of two closely spaced frequencies, for example,
8.3 Hz versus 8.4 Hz. Frequency resolution is determined by the
duration of the acquired signal. In general, the duration required
is approximately the inverse of the difference in frequency. This
reflects the fact that two sinusoids with some difference in fre-
quency will become out of phase with each other over that time
frame. For instance, to reliably distinguish a 0.1 Hz difference in
frequency requires a total acquired signal on the order of 1/(0.1
Hz) = 10 seconds. This applies irrespective of the absolute fre-
quency—that is, distinguishing 8.3 from 8.4 Hz or 124.7 from
124.8 Hz both require approximately 10 seconds of acquisition.
Magnitude or Amplitude
Digitization occurs along the magnitude axis as well. In this
respect, the effect of digitization is typically discussed in terms of
digitization “depth.” Digitizers presume that each successive
point is at a known time interval, so this time interval does not
need to be stored. Instead, only the value or magnitude at each
successive point is stored. Digital equipment is optimized when
storing each successive magnitude in a predefined length. To be
concrete, the output is a stream of 0’s and 1’s. The number of

4-bit 6-bit 8-bit
−7 to +7 −31 to +31 −127 to +127
binary digits (bits) set aside for each measurement is the same.
This storage length can differ between equipment manufacturers.
Early digitizers used 8, 10, or 12 bits to store each measurement.
Digitizers now routinely use 16 to 24 bits. The number of bits
determines explicitly the number of distinguishable magnitudes
of the measurements. For example, 8 bits per measurement can
store up to 28 = 256 different values (0 to 255 unsigned, or –127
to 127 signed two’s complement system). Figure 7.7 illustrates
the practical effect of 4-, 6-, and 8-bit digitization depth.
The three parameters—maximum value (or dynamic range),
minimum distinguishable difference, and digitization depth—
are interconnected, and not independent. Choosing any two
specifies the third. With a 12-bit linear digitizer (212 = 4096 dif-
ferent levels), in order to be able to represent magnitudes up to
1 mV, then the “resolution”—the smallest detectable differ-
ence in magnitude—will be (2000 V/4096) ~0.5 V. The more
bits available, the more accurate each sample. For very low-
amplitude signals (e.g., electrocerebral inactivity recordings),
ADC resolution less than 0.5 V is necessary (12).
ADC Effects on File Size
Both sampling rate and digitization depth have an impact on
acquired file lengths. For example, for 32 channels with digiti-
zation of 16 bits (2 bytes) per sample and a sampling rate of 400
samples per second, 1 hour of EEG (3600 seconds) will gener-
ate (32 2 400 3600) = 92,160,000 bytes = 90,000 kB =
87.89 MB. A full day of EEG with these settings would generate
~2 GB. Increasing the sampling rate to 1000 Hz would increase
the file sizes by 2.5, generating ~5 GB per day.
Guidelines
There are specific guidelines for minimum requirements of dig-
ital EEG acquisition, specifically touching on ADC. The ACNS
guideline (9) suggests that 11-bit depth (including sign bit) is
Chapter 7 ■ Digital EEG 125
Figure 7.7 The practical effect of 4-, 6-, and 8-bit
digitization depth. Note carefully that for the 6-bit
depth, the digital step is most apparent at the peak
and trough, where for each time step the voltage
change is small, but these changes are smaller than
the smallest discrete step and so cannot be repre-
sented.
the minimum, with 12 bits or more being preferred. The mini-
mum amplitude resolution is recommended to be 0.5 V.
Regarding the sampling rate, the guideline has no absolute min-
imum sampling rate but states that “Higher rates are prefer-
able.” The guideline does suggest that “minimum sampling rate
(be) three times the high-frequency (anti-aliasing) filter,” which
is equivalent to saying that for a given sampling rate, the anti-
aliasing filter cutoff be at 2/3 the Nyquist frequency. Since the
ADC rate and anti-aliasing filter are hardware determined,
these are under the control of the manufacturer but should be
investigated and documented clearly when purchasing deci-
sions are considered.
ADC Hardware
Some of the basic principles of the hardware involved in AD
conversion are worth mentioning. First, no AD converter oper-
ates instantaneously. The input voltage is typically measured
over some finite time (obviously less than the time between
samples). The time over which the signal is actually measured is
called the conversion time. Sampling skew, or loss of time axis
integrity, can occur if the ADC samples each channel sequen-
tially, since it takes some time to sample one channel and con-
vert the amplitude to a numerical value before it moves on to
the next channel’s sample. Sampling skew is most problematic
with high-frequency activity and may cause misalignment of
high-frequency events such as spikes between the first and the
last channels sampled. Most ADCs use an input circuit called a
sample-and-hold, employing a capacitor to “hold” the voltage
during this conversion time. Then, comparator circuits are
applied to bracket the “held” voltage to the desired precision. A
second method is to have a series of ADCs, one for each chan-
nel, all triggered by a single computer command. Some digital
instruments have combinations of these two methods, with
ADCs for blocks of four to eight channels using the “sample
126 Part I ■ Basic Principles
and hold” method. Since large ADCs with high throughput are
expensive, multiple ADCs may be more cost-efficient.
There are various ADC hardware structures available. These
structures include direct conversion or flash, successive approx-
imation, ramp-compare, delta-encoded, pipeline, and sigma-
delta. Some of these structures actually involve complex
combinations of simpler ADC designs. For example, the
sigma–delta converter first oversamples at a very high rate, dig-
itally filters and may use a Flash ADC, then finally downsamples
to the desired sampling rate.
The most obvious trade-off of AD converters is between
sampling rate and number of channels. The product (rate
channels) is basically a fixed constant, with this constant being
proportional to the expense of the digitizer. With current hard-
ware digitizers, it is often possible to sample very fast (2 kHz)
on a handful (~16) of channels, or at slow rates (256 Hz) on
several dozen channels (64 to 256). Most commercial EEG
machines aim for a compromise with the ability to monitor 32
to 64 channels at sampling rates of 256 to 1024 Hz, and these
rates can be user programmed according to the task at hand.
Calibration
Digital calibration differs significantly from analog calibration.
The ADC must be calibrated so that it “knows” how large to
represent each voltage point. A 50- or 100- V signal (usually a
30-Hz sine wave) is passed through each amplifier. The ADC
takes the amplifier outputs and assigns a 50 V value to the
peak of each channel. All other amplitude values are deter-
mined based on this initial calibration. Calibration information
is saved with each EEG record. The digital EEG software should
contain a voltage cursor that allows the user to determine the
voltage of individual waveforms. This cursor is then applied to
the known signal from the signal generator to determine
whether the system is properly calibrated.
Square wave calibration pulses can sometimes be generated
using a digital EEG machine to verify the amplifier and analog
filter functioning. The calibration signal should be time-locked
to the ADC clock (13). The square wave calibration should be
examined in the machine reference montage with all digital fil-
ters off. Biologic calibration may not be possible with digital
EEG machines.
Ambulatory and Long-Term EEG Portable Recorders
Systems for acquisition of long-term EEG (hours to days) can
have a variety of configurations. The number of hours of EEG
that can be stored depends on the number of channels recorded,
recorder battery life, and memory storage capacity. Fully portable
ambulatory recorders, designed for home EEG recording, are
small enclosures containing an amplifier, ADC, batteries, and
data storage (usually flash memory card) (7,14). Modern systems
can store up to 96 hours of continuous 16–32-channel EEG data.
When recording is complete, data are downloaded to a computer
via a universal serial bus (USB) or proprietary connection for
review.
In systems for long-term monitoring (LTM) of epilepsy,
amplifiers/ADCs can be connected directly to a computer via
USB or other cable, or have TCP/IP connectivity that allows the
amplifier to be plugged into a network jack near the patient and
to stream data to an acquisition machine anywhere on the net-
work. Advanced LTM recorders may have bluetooth connectivity
and can wirelessly stream data to nearby storage (15) and often
have battery backup and local storage to allow patients increased
mobility while ensuring uninterrupted data acquisition.
Video and Audio Acquisition
Simultaneous video and audio recording is most commonly
done in LTM for epilepsy and in the intensive care unit, to allow
correlation of EEG features with behavioral events (16). Most
digital EEG systems today utilize digital cameras and store EEG
and video data to computer hard drives, although some older
systems with analog video and audio recorded to videotape are
still in operation (17). Early proprietary video capture systems
have largely been replaced by standard digital video recording
equipment, markedly decreasing the cost and increasing the
efficiency of video acquisition and review. Digital video
recorded directly to hard drives can be accessed randomly,
allowing the reviewer to quickly jump to any portion of the
record, rather than having to fast-forward or rewind a series of
videotapes. Video equipment can be customized to optimize
recording for specific types of monitoring. Camera options
include analog versus digital, color versus black-and-white, low
light versus infrared, fixed versus pan-tilt, and wide-angle ver-
sus remote zoom. Many systems have controls for remote pan-
tilt and zoom functions. While early systems required custom
cabling for video, modern systems can transmit and control
digital video over standard TCP/IP cables.
Video and audio are usually encoded into MPEG, MPEG2, or
MPEG4 formats. These are standard formats for digital video,
differing in resolution and compression algorithms. All data
should include time markers so that EEG, video, and audio can
be synchronized precisely (18). Video should be synchronized to
EEG in the millisecond range, or frame-by-frame. Video files
contain large amounts of data; 24 hours of video recording can
be 8 to 30 GB, depending on resolution (usually 320 240 pix-
els or 640 480 pixels), color depth, frame rate, and data com-
pression algorithm. EEG machines for video-EEG recording
should be able to store at least 24 hours of continuous video and
EEG data. Currently available systems, with hard drive capacities
of 200 GB or more, may allow continuous recording for more
than a week. Because of the size of video files, the entire video is
not usually archived; rather, video segments of interest (seizures
and behavioral events) are clipped and archived.
Computer and Software
The EEG acquisition and digitization devices described above are
connected to a central computer, which can be either a standard
desktop personal computer or a laptop computer. The computer
contains a central processing unit (CPU) that performs calcula-
tions, random access memory (RAM) which holds data tem-
porarily while they are being manipulated, and a hard disk drive
for permanent data storage, as well as a monitor, keyboard,
mouse, network connectivity devices, and various connections for
removable media such as CDs, DVDs, and external hard drives.
Most EEG vendors currently use standard commercially available
 Chapter 7 ■ Digital EEG 127

computers, running on commonly used software platforms such

as Microsoft Windows or Macintosh operating systems.
Most of the variability in modern digital EEG systems comes
from proprietary software applications that allow customization
of EEG acquisition as well as manipulation and display of the EEG
and video data through a graphical user interface (GUI). Features
of review software are discussed in more detail below. In addition,
vendor-provided software applications may include databases for
patient and study information, interfaces to hospital patient infor-
mation systems, report-generation software, security packages,
and a variety of postacquisition signal processing software.
Because the central machine is a standard computer, review sta-
tions can also run other software that may enhance laboratory
productivity such as word processors, voice recognition software,
and hospital electronic medical record systems.

Display
Digital EEG can be displayed in a variety of methods. The
recorded signal can be output to an oscilloscope, printed out on
paper, reconverted to analog signal via digital-to-analog con-
version (DAC) and written by traditional pen-writing systems,
or viewed on a computer monitor. The most practical and ver-
satile method is display on a computer monitor.
Visual display on screen is also a digital process. Whether dis-
played on older CRT (cathode ray tube) or modern digital LCD,
images are drawn as discrete pixels. In fact, current display pixel
resolution may be the limiting factor impeding precise represen-
tation of the signal characteristics to the electroencephalogra-
pher (19). The EEG signal consists of points discretized in time
and voltage but obviously not necessarily in adjacent pixels, so
intermediate pixels are filled to construct lines connecting the
points. The top diagram in Figure 7.8 is an idealized example of
a set of such points and where they might lie on a pixel display. Figure 7.8 Effects of horizontal pixel resolution on EEG display.
Modern software draws lines between these points using line- Supposing two time steps per pixel-width (e.g., 200 Hz digital sampling
drawing routines based on Bresenham’s algorithm (20). rate and 1000 pixel-widths for 10 seconds = 100 pixel-widths per sec-
Screen resolution is described in terms of the number of hor- ond). Top panel shows idealized signals at 12.5, 25, 50, and 100 Hz (the
izontal and vertical points or pixels. In 2010, typical mid-range digitization Nyquist frequency). Bottom panel shows how these fre-
display monitors may have 1280 pixels across with 1024 pixels quencies would be actually drawn with pixels. As can be seen, 50 Hz is
vertically. We examine horizontal resolution first. Display of 10 the highest frequency that can be visualized with these pixel settings
seconds of EEG on a screen implies each second being allotted (and not terribly well) with frequencies above 50 Hz appearing
128 pixels, in the best case. If the sampling rate was 256 samples “smudged.” In analogy with the digitization Nyquist frequency, the
per second, then horizontally there would be two EEG points highest frequency represented by pixels can be said to be half the num-
per pixel. Suppose there were (for simplicity) 100 pixels per sec- ber of pixels per second (100 pixel-widths per second can represent
ond, and an EEG channel sampled at 200 Hz, with signals up to signals up to 50 Hz).
100 Hz (which is at the Nyquist frequency, and representable in
the digitized data stream). Figure 7.8 shows a set of points with
sine waves up to 100 Hz. Unfortunately, by the Bresenham line occur when viewing fast frequency activity, such as 60-Hz arti-
drawing algorithm, these signals will not be able to be traced fact or EMG activity, and can occur even when following rec-
out, instead resulting in blurred uninterpretable lines, as shown ommended guidelines for video resolution. When such video
in Figure 7.8 (bottom). In general, the maximum frequency aliasing artifacts are suspected, decreasing the number of sec-
faithfully representable on a pixel display is one half the number onds displayed per page (e.g., from 10 to 5 seconds) or increas-
of pixels per second. In this case, that would be (100/2) = 50 Hz. ing screen resolution will allow display of all samples in each
Display of EEG on monitors with inadequate resolution (i.e., second and resolve the artifact (Fig. 7.9).
lower than the sampling rate) can result in aliasing, the intro- We now turn to the vertical display resolution. The dimen-
duction of spurious waveforms caused by “undersampling” the sion of pixels on current LCD screens is on the order of 0.25
EEG signal (21). This is analogous to violating the Nyquist the- mm. If we try to replicate the vertical scale of analog EEG,
orem when choosing a sampling rate. Aliasing artifacts typically then 7 V/mm translates into 7 V over 4 pixels. Thus, each
128 Part I ■ Basic Principles

Figure 7.9 Aliasing by inadequate moni- A

tor resolution. A: EEG was viewed on a
hospital computer with monitor resolution
set at 800 600 pixels. An unusual 9.5-Hz
artifact is seen in channel T6 – reference.
B: When the EEG was viewed on a moni-
tor with screen resolution 1600 1200
pixels, this signal is seen to be a sinusoidal
60-Hz artifact.

50 µV
1 sec
B

50 µV
1 sec

pixel spans about 2 V in vertical height. With current digi-
tizers (12- or 16-bit), the amplitude resolution is typically
better than 0.5 V, perhaps as low as 0.03 V. Figure 7.10
shows an idealized example with a low-amplitude signal that
ranges from 1 V to –1 V. This variation is lost with pix-
els of 0.25 mm at a sensitivity of 7 V/mm, resulting in a flat
line. Conversely, to achieve 0.5 V per pixel requires setting
the sensitivity to 2 V/mm. With our 1024 pixels in the ver-
tical dimension, and viewing 20 channels, we see that we have
approximately 50 pixels per channel. Using our standard 7
V/mm, EEG signals of 50 V can be displayed. Signals
larger than 50 V will begin to spill over into adjacent
channel display, causing overlap of tracings in adjacent chan-
nels and making the EEG more difficult to interpret.

Figure 7.10 Effects of vertical display resolution on EEG. Idealized
example with a low-amplitude signal that ranges from 1 to –1 V.
This variation is lost with pixels of 0.25 mm at a sensitivity of 7
V/ mm, resulting in a flat line.
Several methods can be used to improve signal display.
Screen size for review stations should measure 20 diagonally or
more. Dot pitch (the diagonal distance between display pixels)
should be as small as possible. Display is best when the moni-
tor’s “native” screen resolution is used. Large (20 or 24 ) mon-
itors with high resolution (1600 1200 pixels, or widescreen
1920 1200 pixels) are currently reasonably priced, and higher
resolution larger monitors are available for those with larger
budgets (Table 7.2). A minimum horizontal resolution of 100
pixels per second is recommended (9). For lower resolution
monitors, decreasing the number of seconds per screen can
allow display of every data sample point. Similarly, for vertical
resolution, one can reduce the number of channels displayed or
use a “zoom” feature to hone in on waveforms of interest. Dual
monitor displays may be a particularly attractive option. Two
side-by-side 24 widescreen WUXGA monitors, rotated 90
degrees, give an effective resolution of 2400 1920 pixels. This
allows viewing of nearly every data point in 10 seconds of EEG
sampled at 250 samples per second.
Digital EEG signals can also be printed on paper. Printers
using fan-fold paper allow printing of a continuous EEG record
similar to that acquired with analog systems. Laser jet printers
Chapter 7 ■ Digital EEG 129
should be able to print at resolutions of 600 dots per inch
(DPI), or approximately 350 samples per second. Optimally, the
system should be able to print a subset of channels on each page
to maximize vertical resolution.
COMPUTER ACQUISITION AND REVIEW
Digital EEG has several advantages over analog recordings (22). In
routine clinical practice, the ability to reformat montages, change
sensitivities and “paper speed,” and manipulate filter settings dur-
ing EEG review are some of the major advantages. For analog
EEG, what the technologist records is what the electroencephalo-
grapher reviews, with no ability to change or reformat data to bet-
ter localize abnormalities or exclude artifacts. Digital EEG allows
review of the data “as acquired” (as the technologist viewed the
data as they were being recorded) or with a nearly limitless array
of post hoc montage, sensitivity, and filter settings.
Digital reformatting improves EEG interpretation. Levy et al.
(23) demonstrated that interrater agreement for “as-acquired”
digital EEG was the same as for paper EEG, with weighted kappa
scores of approximately 0.65 (good agreement). When elec-
troencephalographers were given the ability to reformat the
digital EEG, however, interpreter agreement improved to kappa
scores of 0.8 (very good agreement). Montage reformatting was
most useful for distinguishing normal variants from abnormal
patterns, identifying artifacts, and classifying abnormalities as
focal or generalized. Correlation of digital EEG with simultane-
ously recorded video also enhances the ability to detect and cor-
rectly identify environmental and biologic artifacts.
Digital EEG can be reviewed page by page (usually 10 seconds
per page) or in continuous “scroll” mode. Scroll mode is typically
used during acquisition or when reviewing simultaneous video
and EEG, and page mode is used during review. Page mode is
much faster; up to 5 to 10 pages can be presented each second.
Only superficial features of the EEG can be determined with this
rapid paging speed, but it may be useful for screening for seizures
or determination of sleep states. In page mode, EEG abnormali-
ties or comments at the edge of the page may be missed. This can
be avoided by overlapping the pages by several seconds.
Patient Information
At a minimum, the patient’s name, date of birth, medical record
number, laboratory EEG number, and the date the study was
performed should be recorded and stored with the EEG (24).
This information can be stored in the same file as the EEG data
itself, as a separate text file, or in a database linked to the EEG
data file. Most systems allow recording of more detailed infor-
mation including patient history, medications, and recording
conditions. More advanced systems may include report-gener-
ation software, allowing reports to be stored with the digital
EEG record.
Data Labeling
During recording, the EEG data should be automatically
labeled with time of day, all montage, sensitivity, and filter set-
tings used by the technologist during recording, and start and
130 Part I ■ Basic Principles

Tabl e 7 . 2

Commonly Used Computer Monitor Resolutions

Code Name Aspect Ratio Resolution

VGA Video Graphics Array 4:3 640 480
SVGA Super Video Graphics Array 4:3 800 600
XGA eXtended Graphics Array 4:3 1024 768
SXGA Super eXtended Graphics Array 5:4 1280 1024
WXGA Widescreen eXtended Graphics Array 16:9 1366 768
UXGA Ultra eXtended Graphics Array 4:3 1600 1200
WSXGA Widescreen Super eXtended Graphics Array Plus 16:10 1680 1050
WUXGA Wide Ultra eXtended Graphics Array 16:10 1920 1200
WQXGA Wide Quad eXtended Graphics Array 16:10 2560 1600
WQUXGA Wide Quad Ultra eXtended Graphics Array 16:10 3840 2400

end of procedures such as photic stimulation and hyperventila- Montage reformatting allows the electroencephalographer
tion. The system should support entry of comments by the to mentally reconstruct the three-dimensional EEG by combin-
technologist, both as stored “codes” for common events (e.g., ing views of three or more montages, such as longitudinal bipo-
“eyes open,” “awake,” or “seizure”) and as free text. Comments lar, transverse bipolar, and referential montages.
can also be generated by digital analysis programs, such as spike Additional reference montages can be easily constructed by
and seizure detection algorithms. taking the average of data from two or more electrodes (25).
During review, the electroencephalographer should be able to Laplacian montages (source current density montages) are also
view all notations made during acquisition, as well as add com- easy to implement using “nearest neighbor” methods to
ments. A scrollable list of comments can allow the reviewer to approximate the laplacian (26). Montages can include elec-
immediately find an EEG segment of interest. A particularly use- trodes of the modified 10-10 International System, as well as
ful feature is the ability to create a “context-sensitive” comment, other physiologic monitors such as electro-oculogram, electro-
which stores both a text annotation and the montage, sensitivity, cardiogram, oxygen saturation, and electromyogram.
and filter settings at which the EEG was originally viewed. In addition to changing montages, digital EEG software usu-
ally allows the reviewer to optimize vertical resolution by select-
Montage Reformatting ing only certain channels or groups of channels for display or
Since digital EEG data are acquired using a hard-wired referen- “hiding” channels contaminated by artifact. Individual or
tial montage (input 2 for all channels is the machine reference, groups of channels can be moved or reordered to improve com-
or REF), the computer can perform calculations on the stored parisons between homologous brain regions. Extra space or
data to create any desired montage, nearly instantaneously. For “gaps” between channel groups can facilitate review, albeit at the
example, transforming the recorded montage to a bipolar mon- expense of vertical resolution. Using different colors (red vs.
tage requires the following calculations: blue) for left versus right hemisphere channels can help the elec-
troencephalographer quickly localize waveforms but may intro-
Channel Recording Calculations Display duce subtle visual illusions that can mislead the interpreter.
Montage Montage
Sensitivity Changes
1 Fp1-REF Fp1-REF – Fp1-F7
(F7-REF) In analog EEG, sensitivity is the input voltage ( V) divided by
the output pen deflection (mm). For example, a 50- V signal at
2 F7-REF F7-REF – F7-T3 a sensitivity of 7 V/mm would be 7.1 mm high. In digital EEG,
(T3-REF) the same waveform on the display may not actually measure 7.1
3 T3-REF T3-REF – T3-T5 mm. Rather, the amplitude scale is displayed as a vertical icon
(T5-REF) (calibration line) on the monitor, and the amplitude of EEG
4 T5-REF T5-REF – T5-O1 signals is determined relative to this reference amplitude. The
(O1-REF) scale can be changed to provide a variety of sensitivity settings.
Sensitivity changes can be performed in two ways with a
5 O1-REF digital EEG machine. Rarely, the technician can use a software

sensitivity switch during EEG acquisition to change the actual
amplifier sensitivity (e.g., for intracranial vs. surface EEG
recordings). More commonly, the display sensitivity is changed
during acquisition or review. To decrease the amplitude of the
waveforms on the page, the software program plots each data
point on the monitor as if it were half as large. Sensitivity
changes can be made for a single channel, groups of channels, or
the entire page.
Time Scale Changes
Most EEG software programs display EEG at 10 seconds per
page, or for larger or widescreen monitors, 25 to 35 mm per sec-
ond. The number of seconds displayed per page can be
adjusted, however, to optimize interpretation. For example, dis-
playing 20 seconds or more per page will enhance slow activity
and allow analysis of slow periodic complexes or prolonged
events such as seizures. Spreading out the EEG to show only 2
to 5 seconds per page will spread out faster frequencies, similar
to increasing the paper speed on an analog machine. This
allows more precise analysis of time relationships of signals in
adjacent channels, for example, to analyze the propagation of
epileptiform spikes. In addition, time cursors can usually be
placed on the screen to measure time relationships exactly.
Many software programs allow the number of seconds per page
to be varied from 1 to 100 seconds.
Unfortunately, even the largest monitors cannot adequately
display more than 20 to 30 seconds of EEG, limiting simultane-
ous analysis of long segments of EEG (e.g., evolution of seizures
over several minutes). Printing the EEG to fan-fold paper may
enhance the interpretation of long EEG epochs. Alternatively,
most EEG software programs allow split-screen views to allow
side-by-side comparison of noncontiguous segments of EEG.
Waveform Measurements
Waveforms can be measured in several ways. Some software dis-
plays the maximal “peak-to-peak” amplitudes available for each
Tabl e 7 . 3
Time Domain Digital Filters
Filter Type Pros
FIR Absolutely stable
Linear phase, constant group delay
(peaks and other features of wave
shapes will have the same delay
in two channels processed by the
same filter)
IIR Computationally cheaper
Can obtain good performance with
low order
Can simply and exactly reproduce
analog filters
Chapter 7 ■ Digital EEG 131
channel. Others contain a calibration icon that shows the size of a
50- or 100- V signal. The calibration icon can be moved around
the screen and compared to various waveforms. Most programs
also contain a feature that allows a user to draw a line from peak
to trough of any given waveform to get a numerical value for the
amplitude in microvolts, or the duration in milliseconds. Finally,
more advanced waveform measurements may include average
amplitude or frequency for a selected segment of a channel of
EEG, or even a spectrogram of the component frequencies.
Digital Filtering
One of the major advantages of digital EEG is the ability to
apply, remove, and reapply a variety of digital filters to enhance
or minimize activity in certain frequency bands, without alter-
ing the original EEG data. This is analogous to inserting electri-
cal filters in the acquisition path (see Chapter 6). The field of
digital time domain filters is vast and complex, but some basic
principles are important to illustrate (Table 7.3).
The three common methods of digital filtering include two
time domain methods (finite impulse response [FIR] and
infinite impulse response [IIR]) and one frequency domain
method using the fast Fourier transform. While most commer-
cial systems do not incorporate all the filter choices discussed
below, it is important to understand how different digital filter
types can be designed to mimic (or improve upon) the analog
filters discussed in Chapter 6. Signal processing programs and
toolboxes used for quantitative EEG analysis (e.g., MatLab) will
incorporate a larger number of filters and the ability to design
custom digital filters.
Time Domain Filters
Processing discrete time sequences to accentuate or diminish
frequency bands is made possible by a technique called digital
filtering. We begin with an extremely simple example, the
two-point moving average. If one takes a sequence and
replaces each value by the average of it and the previous value,
Cons
Computationally more expensive
Require higher order for equivalent
performance
Difficult to reproduce effect of analog
filters
Potentially unstable (stable in typical
practice)
Nonlinear phase, group delay is
frequency dependent
132 Part I ■ Basic Principles
one generates a new sequence. If xk is the input at time t k and
yk is the output at t k, then
1 This simple algorithm (average of 2 points) acts as a filter that
yk 2 (xk xk 1)
Figure 7.11 shows this concept applied to pure sinusoids of
several different frequencies, up to the Nyquist limit. The squares
(connected by solid curve) are the original data points. The cir-
cles (connected by dashed curve) are the points that result from
the two-point moving average. We note several features: (i) For
an input of a pure sinusoid, the amplitude of the output may be
altered, but the frequency of the output is the same as the input.
This is a general feature. (ii) The amplitude of the output dimin-
ishes with higher frequency. This is specific to our example.
Other filters might attenuate low frequencies and preserve high
frequencies. (iii) The output is shifted to the right. In particular,
the peaks in the output are all shifted ½ point to the right relative
to the input sinusoid. However, the apparent shift in phase
increases with higher frequency. (iv) At the Nyquist frequency,
the output is identically zero, for this example.
Figure 7.11B shows graphs of these effects. The top shows the
ratio of the output amplitude over the input amplitude, as a func-
tion of frequency (in this case, the absolute frequency divided by
the sampling frequency, as this effect does not depend on absolute
sampling rate, so 0.5 is the Nyquist frequency). The bottom shows
A
100
0 1
−100
0 Seconds
100
−100
0 Seconds
100
0 0
−100
0 Seconds
100
−100
0 Seconds
Figure 7.11 Filtering by FIR filter, two-point moving average. See text for explanation.
the phase delay (in radians and degrees) as a function of (reduced)
frequency. Note also that the phase shift is a straight line (linear).
attenuates higher frequency inputs and preserves low frequencies
and is thus termed a low-pass filter. A four-point moving average
would act similarly but would more dramatically attenuate high
frequencies. A simple filter that would attenuate low frequencies
and preserve high frequencies would be a two-point difference fil-
ter (computing the difference between the current input point
and the previous input point).
The graph of the magnitude of the output as a function of the
input frequency is known as the transfer function. Several differ-
ent regions or bands in this transfer function have specific names.
The frequency band in which the output is preserved (ratio of
output to input near 1) is called the passband. The frequency
band in which the output is severely attenuated (ratio near 0) is
termed the stopband. The frequency band inbetween is known as
the transition band. The slope of the transfer function in the tran-
sition band is called the roll-off. The graph of the phase shift ver-
sus frequency is known as the phase plot. There is a delay in the
output, which is named the group delay and is generally expressed
in number of sample points. The group delay may vary as a func-
tion of frequency. In general, the group delay at a given frequency
is the (negative) slope of the phase plot at that frequency. The
B
Magnitude
1
0.5
0
1 0 Frequency 0.5
0°
Phase
1
−π/4 −45°
−π/2 −90°
1 0 Frequency 0.5
 Chapter 7 ■ Digital EEG 133

order of a digital filter is the number of prior input values that are choices yield a low-pass filter. Note that these coefficients yield
used in the computation of the output. Thus, a two-point aver- a small passband and a cutoff near 0.05 (i.e., low frequencies
age is first order and a four-point moving average is third order. below 0.05 sampling rate are passed, but frequencies above
These are the principal features that are compared and contrasted 0.1 sampling rate are mostly attenuated).
between different digital filters. In this manner, given a set of filter coefficients, one can com-
pute the transfer function and phase plot. There are algorithms for
Finite Impulse Response. The examples described so far are part
generating the coefficients corresponding to traditional analog
of a class known as FIR filters. When the input consists of an
circuits (Butterworth, Chebyshev type 1 and 2, Bessel, etc.) with a
“impulse”—a single nonzero input value in a sequence of zero
given cutoff frequency. The inverse problem—having a desired
inputs—the output sequence (response) will be nonzero for
transfer function, and computing the required order and coeffi-
only a finite number of elements before it necessarily returns to
cients—is beyond the scope of this chapter. For details, see Ref. 27.
zero. This is a direct consequence of the fact that the output yk
In general, for IIR filters, phase shift is nonlinear, so group
depends only on current and prior inputs xk, xk – 1, … This fea-
delay varies with frequency. In fact, particular coefficients for
ture defines the FIR class of filters. FIR filters can be designed to
IIR filters can create a situation in which two finite sinusoidal
be low-pass, high-pass, band-pass, and many other forms. A
signals of different frequencies can appear reversed in order
particular feature of the FIR class of filters is that the phase shift
after being filtered, compared to the input sequence.
is a linear function of frequency, and thus the slope is constant,
Techniques to mitigate group delay for both FIR and IIR filters
so the group delay is the same at all frequencies.
are discussed below.
Infinite Impulse Response. The other class of digital filters is
known as IIR filters. They are also known by other names Undesired Effects
including autoregressive (AR) or recursive. A simple example is Group delay can produce unexpected distortions in digital sig-
1 4 nals, particularly for multichannel comparisons, as with EEG.
yk xk yk 1 Figure 7.13 shows the distorting effect of different filter settings
5 5 when visually comparing channels. Group delay can be prob-
That is, the output (yk) is computed as 1/5 the current input
lematic, and it affects both FIR and IIR filters. As mentioned
(xk) plus 4/5 the prior output.
previously, the group delay generally increases with increasing
Using prior output values is what gives this filter class its
order of the filter. Since FIR filters typically require significantly
important characteristics. For an input sequence that contains an
higher orders to obtain the same transfer function attenuation
impulse (a single nonzero value), this filter will yield an output
as IIR filters, group delay can quickly become very problematic
that continues to have nonzero outputs. Figure 7.12A shows such
for FIR filters. Techniques to eliminate it have been developed.
an IIR (though the values trend toward and approach zero).
Filters that are acausal can eliminate group delay. Acausal filters
This filter applied to another input sequence known as a step
are ones in which future time points as well as past time points
function is shown in Figure 7.12B. This plot will probably
are included in the computation, as opposed to the filters dis-
remind many readers of the graph of voltage charging a capac-
cussed previously that include only past time points, which are
itor. In fact, this filter is the digital analogy of a simple single
called causal. While causal filers can be used to filter live on a
resistor and capacitor circuit arranged as a low-pass filter.
point-by-point basis (but introducing the group-delay artifact),
Digital filters are specified by the coefficients that multiply
acausal filters cannot be implemented live. They can, however,
prior inputs and output to compute the current output. Given
be implemented with a short delay.
a set of coefficients, it is straightforward (though possibly com-
If the entire time series of points is known and immutable,
plicated) to determine the transfer function and phase plot. In
having been acquired and stored, then acausal filters are simple
general, the equation for a filter of order n follows:
to implement and can completely eliminate the group delay. In
yk 5 b0xk 0 b1xk 1 b2xk 2 p bnxk n6 fact, a simple scheme is to filter the entire time series in the for-
5a1yk 1 a2yk 2
p anyk n 6 ward direction, then reverse and use the same filter coefficients
in the backward direction. This eliminates the group delay and
The transfer function (H) and phase are generated by calcu- can even eliminate the phase distortion, achieving zero-phase
lating (or having a graphing program display) distortion. This forward–backward zero-phase filtering works
5b0 b1e i2 x
b2e 2i2 x p 6 for both FIR and IIR filters. A further trick can be employed to
H ` i2 x 2i2 x p 6` minimize transients at the end of the time series, producing the
51 a1e a2e
same result whether filtering forward first or backward first (28).
and
Frequency Domain Filters
1
Imag(H)
phase tan e f Frequency domain filters use FFT to remove unwanted fre-
Real(H) quencies from the signal (4,29). According to Fourier’s theo-
For our simple example, b0 = 0.2 and a1 = 0.8. With these rem, any time series waveform can be modeled as the sum of a
coefficients, the transfer function (solid) and phase (dashed, in set of sinusoidal waveforms, each with different frequency,
radians) are graphed in Figure 7.12C. It demonstrates that these amplitude, and phase. In frequency filtering, the input signal is
134 Part I ■ Basic Principles

A C
+1 1

Magnitude

0.5

0 0
0 10 0 Frequency 0.5

B
+1 0 0°

Phase

−π/4 −45°

0 −π/2 −90°
0 10 0 Frequency 0.5

Figure 7.12 Infinite impulse response (IIR) filter. A: Filter response to an impulse input. The open squares connected by
wide-dash line shows the input, which is zero everywhere but time point 0 where it is 1, and the filled squares connected
by narrow-dash line plots the output, which approaches but never reaches zero, having a persisting, infinite response to
an impulse input, thus the term “infinite impulse response.” B: IIR filter applied to another input sequence known as a
step function. Again, the open squares connected by wide-dash line shows the input, and the filled squares connected by
narrow-dash line plots the output. C: Transfer function and phase plot of IIR low-pass filter. See text for details.

broken into short segments and then transformed into its com-
ponent sine wave signals using FFT. To filter the signal, the coef-
ficients of the unwanted frequencies above or below a specified
cutoff frequency are set to zero, and the inverse Fourier trans-
form is then computed to obtain the original EEG signal minus
the unwanted frequencies (Fig. 7.14).
Frequency domain filters have several limitations. Frequency
filtering cannot be performed until the full epoch to be filtered
has been acquired, precluding true real-time filtering. To improve
computational speed, certain limitations may be placed on the
time series, such as a requirement to contain exactly 2n members
(where n is a positive integer). Finally, if the EEG signal is not sta-
tionary (i.e., has variable frequency components) over the epoch,
the FFT may introduce artifactual high-frequency noise (30). For
this reason, frequency domain filtering is typically performed
over short epochs (e.g., 1 second) to improve the likelihood that
the signal will remain stationary during the entire epoch.
Breaking the EEG into short epochs, however, may introduce dis-
continuities between subsequent epochs. “Windowing” proce-
dures can minimize such discontinuities. Windowing is the
application of a weighting function (e.g., Hamming window) to
taper the ends of each epoch toward zero. Because windowing
necessarily loses some information at the ends of the epochs,
analysis is commonly performed on multiple epochs overlapping
each other by 25% to 75%. The length of the analyzed epoch, the
windowing function utilized, and the degree of overlap all influ-
ence the computational speed of frequency domain filtering.
Digital Analysis and Data Reduction
Digital EEGs can be very rapidly reviewed by displaying pages
at rates of 1 to 10 pages (10 seconds of EEG) per second. The
maximal speed of review is determined by the speed of the

Original signals
Delay
Filtered
Not filtered
0 Seconds 1.0
Figure 7.13 Distorting effect of different filter settings when visually
comparing channels. The top panel shows (synthesized) traces for two
sequential channels from a bipolar derivation with an idealized sharp-
and-slow wave complex, but with significant noise on the top channel.
It is tempting to try to eliminate the noise in that one channel by selec-
tively applying a low-pass filter. The bottom panel shows the result of
an FIR low-pass (cutoff 25 Hz) filter, with order of 50 (a not unreason-
able order for an FIR filter). The high-frequency noise is indeed
reduced, but a delay is induced by the filter. The (downgoing) peak of
the top tracing is delayed by 1/ 8th of a second. All traces are 1-second
duration at 200 samples per second.
network, computer processor, and graphics card. Even at very
rapid review speeds, however, review of an entire 24-hour EEG
(8640 10-second pages) will take 15 to 60 minutes or more. A
variety of data reduction methods can be used to identify EEG
segments of interest for more rapid interpretation.
Digital EEG can be analyzed quantitatively to assist interpre-
tation, automatically detect events, and produce graphical dis-
plays (31–33). A full discussion of these techniques is beyond
the scope of this chapter, but they are included in later chapters.
Signal processing can be performed either on-line, as EEG is
acquired, or off-line. On-line analyses are limited to those
Chapter 7 ■ Digital EEG 135
which can be performed in real time by the computer proces-
sor, while off-line analyses can be more complex. Digital pro-
cessing can highlight specific features of the EEG using
graphical displays, remove or correct artifacts, or detect com-
mon patterns such as seizures or sleep states.
Spike and Seizure Detection
Algorithms to detect spikes and seizures are commonly used in
long-term video-EEG monitoring, intensive care unit monitor-
ing, and ambulatory EEG. The algorithms are designed to detect
specific features in the digital data, such as the sharpness and
field of an epileptiform discharge or the change in frequencies or
amplitude at the beginning of a seizure (34,35). Other methods
use neural networks that can be trained and updated with a
database of seizures to improve detection accuracy (36,37). The
detection algorithms mark the sections of data which contain
events of interest. The electroencephalographer can then review
this prescreened data rather than the entire EEG recording. The
detection algorithms are not perfect, and may either miss
seizures or spikes, or have high numbers of false detections.
Topographic Display (Mapping)
Topographic maps display the EEG voltage at various locations on
the head (38). In order to construct smoothed maps, interpolation
techniques such as spline interpretation or nearest neighbor rules
are used (39,40). The maps are most accurate when large numbers
of electrodes are used. Topographic maps can be used to display
the voltage topography of epileptiform activity (41).
Dipole Source Modeling
This technique aids in localizing the generators of EEG events,
such as epileptiform discharges (42,43). Multichannel digital
EEG is modeled as being generated by a single (or a small num-
ber) dipole source. Each dipole source produces particular scalp
voltage waveforms when placed in a mathematical model of the
head. The locations and orientations of the dipoles are adjusted
to produce the best fit between the mathematical model and the
actual EEG. Source localizations can be coregistered with data
from MRI.
Spectral Analysis/Graphical Displays of Quantitative EEG
Spectral analysis involves analyzing the various frequencies that
make up the digital EEG signal (44). Typically, the fast Fourier
transform is used to calculate the frequencies present in the EEG.
The spectral analysis of a segment of EEG can then be plotted on
a topographic map or displayed as a graph of power in a partic-
ular frequency versus time. This application is increasing in use
for long-term EEG monitoring in the intensive care unit (45).
DATA SHARING, STORAGE, AND
NETWORKING
EEG File Formats and Data Sharing
EEG files acquired with current commercial EEG systems are
typically stored in proprietary formats that require vendor-
specific software to open and view the files. This limitation can
136 Part I ■ Basic Principles
Figure 7.14 Filtering in the fre- Original
quency domain. The top left shows
the original signal (this example has
512 samples per second). The fast
(discrete) Fourier transformation 0
(FFT) converts this signal into a fre-
quency domain representation (top
right). The spectrum is multiplied by
a desired filter function (middle
right), in this case preserving all fre- 0 Seconds
quencies between 2 and 32 Hz—a
bandpass—and zeroing all other fre-
quency components. The resulting Filtered
spectrum (bottom right) is then
inverse Fourier transformed back into
the time domain, yielding the filtered
signal (bottom left). The Fourier 0
domain representation also has
phase information—this has been
omitted for clarity of illustration.
0 Seconds
cause significant problems in transferring data between differ-
ent EEG laboratories for review, and even for maintaining read-
able archives as file formats are changed over time. There are
several approaches to overcome this Tower of Babel problem
(46). Most commercial systems can export EEGs to a compact
disk (CD) or digital video disk (DVD) with a limited-function
version of the EEG reader software, so the disk becomes a self-
contained reader station usable on any computer. Others pro-
vide a “light” version of review software that can be installed on
multiple computers at low cost. Some allow EEGs to be con-
verted to open-source formats (e.g., European Data Format
[EDF], European Data Format plus [EDF ] (47), Extensible
Biosignal [EBS] (48), ASCII (49), or the American Clinical
Neurophysiology Society’s Technical Standard 1 (50)) that can
be read by a variety of freely available reader programs and
some commercial systems. These open formats may not repli-
cate all of the features available in the original EEG software,
and most have been adopted by only a few vendors. Finally,
there are several commercial programs that can read EEG files
collected by different commercial EEG systems, but these may
not read all EEG file types and can be quite expensive.
Data Storage
Once EEG is reviewed, video and EEG data are typically clipped
for events of interest and archived to long-term storage.
Archiving schemes vary; most labs save all routine EEGs, but
only clips of significant events for long-term video-EEG moni-
toring. Some save all continuous 24-hour EEG data, but contin-
uous video files are too large to store in a cost-effective manner.
A laboratory performing 2000 routine EEGs, 1000 video-EEG
studies, and 500 ambulatory EEGs will produce more than a
terabyte of digital data each year.
Spectrum − power
FFT
0 Hz 256
× (multiply by)
1
0 Hz 256
Inverse
FFT
1
0 Hz 256
EEGs can be archived on a variety of media, including CD,
DVD, tape, hard drives (internal or external), servers, or network-
attached storage devices. These media vary greatly in size
(amount of data which can be stored), cost, convenience (time
required to archive and ease of use), and durability. Use of digital
media has significantly decreased space requirements and cost of
EEG storage. CDs (storing up to 700 MB) and DVDs (4.7 GB,
dual layer 9 GB) are most commonly used when EEGs need to be
transferred to another electroencephalographer for interpreta-
tion. Archiving to removable media is low cost but inefficient, as
each disk holds only a small amount of data, must be “burned”
individually, must be placed back in the computer for the data to
be reviewed, and may become unreadable over years. External
tape drive storage devices have large capacity and rapid backup,
but data retrieval can be slow because the tape must be trans-
ferred back to the review station hard drive before review. Tapes
may also deteriorate over time. External hard drives are inexpen-
sive and provide more rapid access to data than tape drives. Hard
drive costs have decreased significantly over recent years, making
storage of large amounts of data on servers or network-attached
storage devices economically feasible. EEG data stored on a server
can be rapidly retrieved from any computer on the network.
Maintaining digital EEGs archives (e.g., updating to new media
or to new file formats) can be costly and time-consuming.
The size of an EEG file depends on the number of channels
recorded, sampling rate of the ADC, and the duration of the
recording. One hour of 32-channel EEG sampled at 200 sam-
ples per second produces an approximately 50-MB file, while 1
hour of video data can be nearly 1 GB, depending on resolution
and compression. Compression techniques can be used to
reduce the number of bits necessary to represent information,
by removing repeated bits. For example, instead of encoding 10
 Chapter 7 ■ Digital EEG 137

Tabl e 7 . 4

Transmission Speed of Digital EEG and Video

EEG Video
Type Data Rate Transfer Time Real Time Transfer Time Real Time
Wired
Gigabit Ethernet 1 Gbps 0.5 sec Yes 5.3 sec Yes
T4 (fiberoptic 274.76 Mbps 1.7 sec Yes 19.2 sec Yes
trunk line)
Fast Ethernet 100 Mbps 4.8 sec Yes 52.8 sec Yes
Ethernet 10 Mbps 48 sec Yes 8 min 48 sec Yes
Cable modem 2.5–6 Mbps 1 min 20 sec– Yes 14 min 40 sec– Yes
3 min 12 sec 35 min 20 sec
T1 (trunk line) 1.54 Mbps 5 min 20 sec Yes 57 min 8 sec Yes
DSL 1–3 Mbps 2 min 40 sec–8 min Yes 29 min 20 sec– Maybe
88 min
Modem 56 kbps 2 hr 38 min No 26 hr 20 min No
Wireless
802.11n 100 Mbps 4.8 sec Yes 52.8 sec Yes
802.11a/g 54 Mbps 8.9 sec Yes 1 min 38 sec Yes
801.11b 11 Mbps 43.6 sec Yes 8 min Yes
4G cellular 6–100 Mbps 4.8 sec–1 min 20 sec Yes 52.8 sec– Yes
14 min 40 sec
3G cellular 144 kbps–2.4 Mbps 3 min 20 sec– Yes 36 min 20 sec– Maybe
55 min 33 sec 10 hrs 10 min

Digital EEG, 32 channels, 16-bit resolution, sampled at 256 Hz, 1 hour = 60 MB = 480 Mbit. MPEG4 color video, 640 pixels 480 pixels resolution, 30 frames per
second, compressed, 1 hour = 660 MB = 5280 Mbit.

binary 1 and 5 binary 0 values, the compression method
encodes that a binary 1 value is repeated 10 times and binary 0
value 5 times (111111111100000 vs. 10-1 5-0). The destination
computer then decodes the information and expands it back to
the original number of bits. While it is possible to compress dig-
ital EEG signals slightly, in general the savings are small and
may not justify computing overhead (decompressing data) in
these systems. Digital video compression is essential, and this
technology continues to evolve.
Networking
One of the most useful features of digital EEG is the ability to
connect acquisition machines to computer networks for data
transmission and communication. A network is a group of
computers and other devices that are connected together to
facilitate the transfer of information and sharing of resources.
The type of network and its bandwidth, or speed, will deter-
mine how quickly EEG studies can be accessed and reviewed
(Table 7.4). Wired networks are typically faster (100 to 1000
Mbits/sec) and more reliable than wireless networks (band-
width 11 to 100 Mbits/sec), although the latter does not incur
high costs of running cable to all EEG locations and may be
adequate for low-volume laboratories.
An efficient and organized network requires careful plan-
ning, usually in conjunction with trained and certified hospital
network managers. The design of a network for EEG is deter-
mined by a few basic principles: (i) function, (ii) geography
(number of users, distance of the farthest users from the
server), (iii) speed or bandwidth, (iv) budget, (v) expandabil-
ity/flexibility, and (vi) requirements for network administra-
tion. The first step is to identify the purpose(s) of the network,
such as (i) data acquisition, (ii) transmission of data to central
review area, (iii) provision of access to data from multiple loca-
tions, sometimes simultaneously, (iv) interpretation of studies,
(v) generation of reports, and (vi) data archiving.
A server, a computer specialized for storage, transmission,
and security of data, can provide a central repository of files,
most importantly EEG and video files, EEG databases, and EEG
reports, which can then be accessed from multiple network
locations. The EEG can be streamed continuously to the server
138 Part I ■ Basic Principles

as it is acquired, or transferred after the study is completed.
Central location of the data facilitates security measures and
improves the efficiency of data backups and archiving. The
server is typically located locally in the EEG laboratory or in the
hospital’s server farm.
Networks used in hospitals usually consist of a long length of
very high bandwidth cable media that connects the servers,
called the backbone, with branches off the backbone called seg-
ments (Fig. 7.15). Segments (such as to EEG labs) usually con-
sist of lower bandwidth cable media, which again branch to
individual users. EEG machines connected to the network can
be accessed remotely to review ongoing studies, update patient
databases, and adjust recording parameters (e.g., montage,
camera position). The bandwidth (or speed) of the network
will determine the speed of remote review.
Finally, the need to review data from several different sites
(remote EEG labs, physicians’ offices, home, etc.) may
require additional resources, such as remote access servers,
as well as Internet service providers outside the hospital or
health care facility. Users can access the network using nearly
any device with Internet access, such as a home computer,
laptop, wireless personal digital assistant, or digital cellular
telephone. Cellular wireless systems are slower than cable or
DSL connections but may be adequate to quickly review
emergency studies.
Remote access allows users to connect to the network and
perform tasks as if they were directly connected to the net-
work, even when they are far away. Remote access can be
accomplished via a dial-up connection to a remote access
server or via a virtual private network (VPN) tunneled
through the Internet. The VPN tunnel maintains the confiden-
tiality of the data through encryption. Transmission of EEG
data to remote locations may be very slow (see Table 7.4). A
terminal server system such as Windows Terminal Server or
Citrix Metaframe allows multiple remote users to run applica-
tions from anywhere with an Internet connection, nearly iden-
tical to when sitting in front of a computer on the network.
The remote user logs in and is authenticated, then can open
the EEG software application. The terminal server sends
remote keystroke and mouse movements to the application

Research Home Remote Site

Remote
Access
Server
Internet Service Provider
Backup /
Storage

To Remote Sites on Hospital

Backbone: Pediatric Server(s)
EMU, etc.
Backbone

Epilepsy EEG
OR ICU Wada, PET Collectors Collectors

Reading
Room

Fellows

Attending

Figure 7.15 A complex EEG network connected to the hospital backbone. EEG data are collected from patients in vari-
ous locations and sent to a central server. The backbone provides for high-speed transmission of files, so the review loca-
tion can be quite distant from the collection locations. The network also provides remote access for users outside the
hospital. The hospital server can provide a backup for the EEG server.

and sends screenshots of the application to the remote user.
Because only screenshots, not the actual EEG data, are trans-
mitted, review is typically much faster.
An information technology specialist or network adminis-
trator is often helpful for troubleshooting of computer and
software problems and to provide network security and disaster
protection/recovery. Network security is especially important if
patient information is stored in the network, as law protects
patient confidentiality. The security architecture must be com-
pliant with the Health Insurance Portability and Accounting
Act (HIPAA) of 1996 guidelines. Adequate security measures
must protect against unauthorized users, computer viruses,
accidental destruction of files, and other forms of attack.
Sensitive files may need to be encrypted both on the computer
disk and during network transmission. Networks connected to
the Internet or other outside connections are especially vulner-
able to outside attack and need a firewall or proxy to create a
barrier around the internal network. Protecting data from cat-
astrophic data loss includes several considerations: (i) power
backup with surge protectors or universal power supplies, (ii)
data backup on a regular and reliable schedule, (iii) mainte-
nance of backup files in a secure location separate from the
original files, (iv) disk fault tolerance by providing multiple
copies of the same information on a array of hard drives, and
(v) server fault tolerance by providing multiple servers in case
of server malfunction.
Security
HIPAA (http://www.hhs.gov/ocr/hipaa/) instituted a set of
rules (the “Security Standard”) that impact digital EEG storage,
review, transmission, and archiving. The Department of Health
and Human Services has published a guideline document
(http://www.cms.hhs.gov/SecurityStandard/Downloads/Securi
tyGuidanceforRemoteUseFinal122806.pdf) that provides a
concise description of many of the issues that relate to HIPAA
rules on network transmission and storage of protected health
information (PHI) and which can be applied directly to digital
EEG. We review some general issues and recommendations and
describe some of the specifics of the HHS guideline.
One goal intended by HIPAA was protection of patient pri-
vacy and prevention of abuse of patient information. While
the vast majority of data in digital EEG files are the digitized
voltage time series and are fairly useless to identity thieves,
some portions of the files (depending on the vendor) can con-
tain electronic protected health information (ePHI), such as
name, date of birth, address, doctor, and diagnoses. These
pieces of information could be used to harm patients if
obtained by individuals with ill intent. Strategies for protec-
tion of this information must be addressed by the modern
digital EEG laboratory.
The HHS guideline groups electronic processing into three
areas: accessing, storing, and transmitting.
Accessing
With regard to digital EEG, an example relating to accessing
might be the login accounts of acquisition and reviewing
workstations. Unauthorized persons might access patient
Chapter 7 ■ Digital EEG 139
information by directly viewing unsecured workstations, or
obtaining passwords written on the stations themselves, key-
stroke capture malware, or social engineering (requesting
passwords from authorized personnel by using the pretense of
being an Information Technology employee). Some strategies
applicable to digital EEG to mitigate these risks include:
Computers on which EEG data are stored must be password
protected. Automatic logins, in which a login password is
bypassed on computer startup, should be disabled. Each user
should have an individual password. Vendors of EEG acquisi-
tion and reading stations should work toward the goal of
avoiding the practice of a common username and password
for all technicians and physicians using the stations. Ideally,
each user should have individual account, with shared config-
uration settings placed in shared folders and personalized
preferences stored in individual accounts. Antivirus and anti-
malware software should be installed and kept current on
each workstation to mitigate the possibility of keystroke
capture of passwords.
Storing
Digital EEG involves storing/archiving vast amounts of data, and
while PHI is again a small minority, risks remain. Laptops con-
taining digital EEG could be stolen. Portable hard drives could
likewise be stolen. Digital EEG records could be recovered from
local cached files when accessed from computers outside the hos-
pital. Strategies to minimize these risks include: Laptops and
desktops could be stored in locked rooms when not in use.
Laptops could be secured by “locking mechanisms” to fixed walls.
Laptops (and even desktops) could have full hard disk drive
encryption. Without the disk encryption password, the plaintext
information would not be recoverable. Of course, this leads to
additional passwords, so a clear password policy and password
management system becomes even more necessary. Prevention of
cached records on outside computers is more difficult and would
likely involve software implementation by vendors.
Transmitting
This is probably the aspect of security with digital EEG that
requires the most attention. There is no doubt that remote
access (physicians at off-sites, home, conferences, etc.) and dis-
tant second opinions could add much value to patient care.
However, remote access then requires careful consideration of
protection from data security breaches. Electronic PHI could be
intercepted when in network transit, either outside hospital
network (external), or even within (internal). Strategies to min-
imize the risk of network interception include: Digital EEG data
transmitted over networks should be encrypted. For external
network access, most sites use VPN protocols. Ideally, auto-
mated network analysis by the hospital IT staff should prompt
alerts for unusual traffic, such as frequent logins, simultaneous
logins from disparate locations, logins from other continents,
and so on Web-based (HTTP) protocols are also in some use.
For these, at a minimum HTTPS (HTTP-secure) should be
used. For HTTPS, modern browsers use TLS (transport layer
security, which superseded the older SSL/secure sockets layer
protocol) for bidirectional encryption.
140 Part I ■ Basic Principles
CONCLUSIONS
Rapid advances in digital technology have greatly improved the
efficiency and reliability of clinical EEG recordings and offer
new possibilities for research applications. New electrode sys-
tems that use dry electrodes or require limited skin preparation
may allow recordings to be initiated easily in nontraditional
locations, such as the emergency room, ambulances, or even
patients’ homes. EEG data could then be transmitted wirelessly
to central locations for access and review by electroencephalog-
raphers. Small, light, wireless amplifiers may allow recording of
ambulatory patients over weeks to months, analogous to car-
diac monitoring with implantable loop recorders. Such data
could help to elucidate patterns of seizure occurrence or the
development of epileptiform activity or seizures after poten-
tially epileptogenic brain injuries. Current network and remote
access solutions allow rapid EEG review from any location,
making “real-time” monitoring of brain activity (e.g., for criti-
cally ill neurology patients) an achievable goal. Decreased cost
and increased capacity of storage media allows collection and
archiving of vast amounts of EEG data. Modern ADCs can sam-
ple at thousands of times per second from hundreds of chan-
nels, facilitating collection of multichannel EEG from
microelectrodes to reveal pathophysiological mechanisms of
epilepsy and other neuronal activity at the single neuron or
localized network level. Finally, EEG data can be analyzed by an
ever-increasing array of signal analysis techniques, coregistered
with imaging procedures for source localization or used to cre-
ate statistical databases of normal and abnormal findings.
These techniques are described in detail in later chapters.
REFERENCES
1. Berger H. Uber das Elektrenkephalogramm des Menschen IV.
(English translation by P. Gloor: On the Electroencephalogram of
Man, Electroenceph Clin Neurophysiol 1969, Suppl 28). Arch
Psychiat Nervenheil. 1929;97:6–26.
2. Burks AW, Burks AR. The ENIAC: first general-purpose electronic
computer. 1981. MD Comput. 1995;12(3):206–212.
3. Farley B, Clark W, Gilmore J, et al. Computer techniques for the study
of patterns in the electroencephalogram: Lincoln Laboratories, MIT,
1957. Contract No. 337.
4. Cooley JW, Tukey JW. An algorithm for the machine calculation of
complex Fourier series. Math Comput. 1965;19:297–301.
5. Technology and equipment review. Digital electroencephalo-
graphs. J Clin Neurophysiol. 1993;10(3):378–392.
6. Guideline Five: Guidelines for Standard Electrode Position
Nomenclature. American Clinical Neurophysiology Society; 2006
[updated February 10, 2006]. Available from: https://www.acns.org/.
7. Schomer DL. Ambulatory EEG telemetry: how good is it? J Clin
Neurophysiol. 2006;23(4):294–305.
8. Van Drongelen W. Signal Processing for Neuroscientists. Burlington,
MA: Academic Press; 2007.
9. Guideline Eight: Guidelines for Recording Clinical EEG on Digital
Media. American Clinical Neurophysiology Society; 2006 [updated
Februray 10, 2006]. Available from: https://www.acns.org/.
10. Vanhatalo S, Voipio J, Kaila K. Full-band EEG (FbEEG): an emerg-
ing standard in electroencephalography. Clin Neurophysiol. 2005;
116(1):1–8.
11. Horowitz P, Hill W. The Art of Electronics. 2nd ed. Cambridge,
England: Cambridge University Press; 1989.
12. Nuwer MR, Comi G, Emerson R, et al. IFCN standards for digital
recording of clinical EEG. The International Federation of Clinical
Neurophysiology. Electroencephalogr Clin Neurophysiol Suppl.
1999;52:11–14.
13. Blum DE. Computer-based electroencephalography: technical
basics, basis for new applications, and potential pitfalls.
Electroencephalogr Clin Neurophysiol. 1998;106(2):118–126.
14. Gilliam F, Kuzniecky R, Faught E. Ambulatory EEG monitoring. J
Clin Neurophysiol. 1999;16(2):111–115.
15. Budinger TF. Biomonitoring with wireless communications. Annu
Rev Biomed Eng. 2003;5:383–412.
16. Guideline twelve: guidelines for long-term monitoring for
epilepsy. Am J Electroneurodiagnostic Technol. 2008;48(4):265–286.
17. Ives JR. Video recording during long-term EEG monitoring of
epileptic patients. Adv Neurol. 1987;46:1–11.
18. Burgess RC. Design and evolution of a system for long-term elec-
troencephalographic and video monitoring of epilepsy patients.
Methods. 2001;25(2):231–248.
19. Risk WS. Viewing speed and frequency resolution in digital EEG.
Electroencephalogr Clin Neurophysiol. 1993;87(6):347–353.
20. Bresenham J. Algorithm for computer control of a digital plotter.
IBM Systems J. 1965;4(1):25–30.
21. Epstein CM. Aliasing in the visual EEG: a potential pitfall of
video display technology. Clin Neurophysiol. 2003;114(10):
1974–1976.
22. Swartz BE. The advantages of digital over analog recording
techniques. Electroencephalogr Clin Neurophysiol. 1998;106(2):
113–117.
23. Levy SR, Berg AT, Testa FM, et al. Comparison of digital and conven-
tional EEG interpretation. J Clin Neurophysiol. 1998;15(6):476–480.
24. Guideline fourteen: guidelines for recording clinical EEG on
digital media. American Electroencephalographic Society. J Clin
Neurophysiol. 1994;11(1):114–115.
25. Jayakar P, Duchowny MS, Resnick TJ, et al. Localization of epilep-
togenic foci using a simple reference-subtraction montage to doc-
ument small interchannel time differences. J Clin Neurophysiol.
1991;8(2):212–215.
26. Klein SA. Inverting a Laplacian topography map. Brain Topogr.
1993;6(1):79–82.
27. Oppenheim AV, Shafer P, Buck JR. Discrete-Time Signal Processing.
2nd ed. Englewood, NJ: Prentice-Hall, Inc.; 1998.
28. Gustafsson F. Determining the initial states in forward-backward
filtering. IEEE Trans Signal Process. 1996;44(4):988–992.
29. Brigham EO. The fast fourier transform and its applications. In:
Oppenheim AV, ed. Englewood Cliffs, NJ: Prentice Hall; 1988.
30. Cook EW III, Miller GA. Digital filtering: background and tutorial
for psychophysiologists. Psychophysiology. 1992;29(3):350–367.
31. Colditz PB, Burke CJ, Celka P. Digital processing of EEG signals.
IEEE Eng Med Biol Mag. 2001;20(5):21–22.
32. Duffy FH, Hughes JR, Miranda F, et al. Status of quantitative EEG
(QEEG) in clinical practice, 1994. Clin Electroencephalogr. 1994;
25(4):VI–XXII.
33. Nuwer MR, Lehmann D, da Silva FL, et al. IFCN guidelines for topo-
graphic and frequency analysis of EEGs and EPs. The International
Federation of Clinical Neurophysiology. Electroencephalogr Clin
Neurophysiol Suppl. 1999;52:15–20.
34. Gotman J. Automatic detection of seizures and spikes. J Clin
Neurophysiol. 1999;16(2):130–140.
35. Qu H, Gotman J. A seizure warning system for long-term epilepsy
monitoring. Neurology. 1995;45(12):2250–2254.

36. Wilson SB, Turner CA, Emerson RG, et al. Spike detection II: auto-
matic, perception-based detection and clustering. Clin Neurophysiol.
1999;110(3):404–411.
37. Wilson SB. A neural network method for automatic and incre-
mental learning applied to patient-dependent seizure detection.
Clin Neurophysiol. 2005;116(8):1785–1795.
38. Nuwer MR. The development of EEG brain mapping. J Clin
Neurophysiol. 1990;7(4):459–471.
39. Perrin F, Pernier J, Bertrand O, et al. Spherical splines for scalp
potential and current density mapping. Electroencephalogr Clin
Neurophysiol. 1989;72(2):184–187.
40. Soong AC, Lind JC, Shaw GR, et al. Systematic comparisons of
interpolation techniques in topographic brain mapping.
Electroencephalogr Clin Neurophysiol. 1993;87(4):185–195.
41. Ebersole JS, Wade PB. Spike voltage topography identifies two types
of frontotemporal epileptic foci. Neurology. 1991;41(9):1425–1433.
42. Gotman J. Noninvasive methods for evaluating the localization and
propagation of epileptic activity. Epilepsia. 2003;44(suppl 12):21–29.
43. Scherg M, Ebersole JS. Brain source imaging of focal and multifo-
cal epileptiform EEG activity. Neurophysiol Clin. 1994;24(1):51–60.
Chapter 7 ■ Digital EEG 141
44. Alarcon G, Binnie CD, Elwes RD, et al. Power spectrum and
intracranial EEG patterns at seizure onset in partial epilepsy.
Electroencephalogr Clin Neurophysiol. 1995;94(5):326–337.
45. Scheuer ML, Wilson SB. Data analysis for continuous EEG monitor-
ing in the ICU: seeing the forest and the trees. J Clin Neurophysiol.
2004;21(5):353–378.
46. Rapoport DM, Ayappa I, Norman RG, et al. NPSG data inter-
change-dealing with the Tower of Babel. Sleep. 2006;29(5):599–600.
47. Kemp B, Olivan J. European data format ‘plus’ (EDF ), an EDF
alike standard format for the exchange of physiological data. Clin
Neurophysiol. 2003;114(9):1755–1761.
48. Hellmann G, Kuhn M, Prosch M, et al. Extensible biosignal (EBS) file
format: simple method for EEG data exchange. Electroencephalogr
Clin Neurophysiol. 1996;99(5):426–431.
49. Designation AS. E1467-92 specification for transferring digital
neurophysiological data between independent computer systems.
Available from: ASTM, 1916 Race St, Philadelphia, PA 19103. 1992.
50. Guideline TS1: Standard for transferring digital neurophysiological
data between independent computer systems. J Clin Neurophysiol.
2008.

Polarity and Field Determinations
INTRODUCTION
Accurate EEG interpretation is not possible without a thorough
understanding of EEG scalp localization. In contrast to most
other aspects of EEG interpretation, the process of scalp locali-
zation is deductive and does not present the problem of reason-
able disagreement between readers.
The development of digital EEG recording has made it pos-
sible to view the same segment of recording retrospectively
using different frequency filters and spatial filters (montages).
This retrospective analysis greatly increases the ability of the
electroencephalographer to localize the electrical fields of the
scalp EEG and thereby provide more accurate interpretation. It
is not an overstatement to say that it has changed the practice of
electroencephalography.
There are several reasons why localizing EEG activity is an
essential skill. First, the identification and classification of EEG
activity is always based to some degree on localization. For
example, a high amplitude, sharply contoured, anterior tempo-
ral waveform that clearly stands out from the background and
is activated by sleep has a high likelihood of being an abnormal
epileptiform spike. However, the same waveform appearing
over the central head region during sleep is far more likely to be
a benign vertex wave of normal sleep. Second, localization is
often the only feature that allows the electroencephalographer
to determine if an activity is cerebral in origin or an artifact. A
sharp rhythmic pattern that evolves in frequency and morphol-
ogy over a left frontal head region may be likely an electro-
graphic seizure, whereas the same pattern that appears
simultaneously over both the left frontal head region and the
right occipital head region without involvement of electrodes
between these regions is far more likely to be an artifact. Third,
the clinical correlation portion of EEG interpretation always
depends on the cortical localization of the EEG findings. Lastly,
localization is the one skill that greatly increases the likelihood
of correctly interpreting an EEG pattern the reader has never
encountered before.
THE EEG SIGNAL: CORTICALLY GENERATED
SCALP CURRENTS
The skill of localization in routine clinical EEG interpretation
consists of two steps: (1) localizing EEG activity to specific
electrode positions on the scalp and then (2) relating the scalp
localization to the likely source in the underlying cerebral
CHAPTER
BRUCE J. FISCH 8
cortex. To perform the latter requires an understanding of how
the EEG signal arises and is conducted to the scalp. The relation-
ship between the scalp signal and its origin within the brain is a
complex topic that has been approached by a variety of mathe-
matical models (and is covered in greater detail in other chapters
of this book). In routine practice localization is approximated
according to a basic understanding of cortical anatomy and
volume conduction theory.
The spontaneous EEG signal in the routine scalp recorded
EEG arises from cortical pyramidal cell postsynaptic potentials.
Although other sources (e.g., other neurons and glial cells) and
forms of electrical potentials (e.g., action potentials) within the
cortex exist, their contributions to extracranial electrical fields
appear to be very limited. Pyramidal cells that generate postsy-
naptic potentials have the biophysical advantage of firing in
synchrony (i.e., they summate in time) with similar radial ori-
entation (i.e., they are aimed in the same direction toward the
cortical surface) and long duration, making them ideal for pro-
ducing a large enough signal to be conducted through the skull
and recordable at the scalp. As illustrated in Figure 8.1, this usu-
ally means that the recording electrode that detects the highest
amplitude signal is overlying the cortical source (1). However,
there are contributions from the walls of cortical sulci that are
not oriented perpendicular to the scalp and may affect adjacent
electrode positions. More notable exceptions to this simple spa-
tial relationship can occur, particularly if:
(1) the source is in a fissure
(2) the source is primarily in sulci
(3) cortical anatomy and orientation have been altered (most
commonly in cerebral palsy)
(4) skull thickness has been altered (there is relative thinning as
in early infancy, thickening as in certain bone diseases, or
gap in the skull from prior trauma or surgery).
The way in which electrical current is transmitted throughout
the body is referred to as volume conductor theory. Figure 8.1
illustrates an example of volume conductor theory applied to
EEG in which the outer layers of the cortex are producing a
momentary signal that is negatively charged. In the example, the
EEG source involves an entire gyrus. Because the pyramidal cell
postsynaptic potentials generate a circuit of current flow
throughout the length of each cell, one end of the pyramidal cell
always has the opposite electrical polarity from the other (in Fig.
8.1 the outer cortex is negative and the inner cortex is positive).
In Figure 8.1, the addition of all polarities oriented toward the
143
144 Part I ■ Basic Principles
Figure 8.1 This figure illustrates volume conductor theory for a cortical
source that occupies a gyrus. The distribution of the amplitude of the
voltage that reaches the scalp surface is shown in line S. The current that
reaches the electrode from the cortical source is equivalent to the area
contained by the lines that begin at the cortex and converge on the elec-
trodes P1 or P2. For P1, the gyral convexity source projects only negativ-
ity. For P2, the source projects both negativity from the outer cortex and
positivity from the inner cortex of the sulcus (shown by the area between
the dashed lines that converge on P2). The cancellation effect of the inner
cortex and the outer cortex (negativity added to positivity; the area within
the dashed lines subtracted from that within the solid lines converging on
P2) reduces the voltage of the signal detected by P2 as shown in line S.
Although the figure is drawn in two dimensions, it is intended to repre-
sent three-dimensional space with the lines leading from the cortex to
electrodes representing the sides of an approximately conical space. The
amount of current reaching P1 and P2 then depends on the solid angle
(signified by the omega characters in the figure) subtending the volume.
(From Gloor P. 1985 Neuronal generators and the problem of localization
in electroencephalography: application of volume conductor theory to
electroencephalography. J Clin Neurophysiol. 2:327–354.)
recording electrode determines the final voltage at the scalp
(negative in the outer cortex pointing toward the recording elec-
trodes, and positive in the deeper cortex pointing toward the
recording electrodes).
The scalp EEG can be imagined as a constantly changing
relief map in which the map is composed of hills and valleys
whose height or depth represents a particular voltage that is
Figure 8.2 This figure illustrates the steps outlined in the text
(A through F) for measuring the head for the placement of the 10-20
system. (From Fisch BJ. Fisch and Spehlmann’s Digital and Analog EEG
Primer. Amsterdam: Elsevier; 1999.)
determined by changing currents arising from underlying
pyramidal cell postsynaptic potentials. Electrode placement is
used to accurately map the contours of the changing voltages
over the scalp surface. Each electrode, therefore, represents a
spatial voltage sample point. The more electrodes used, the
more faithfully they will map the actual contours of the scalp
potentials. If too few electrodes are used, then the presence or
absence of EEG changes in some areas of the scalp may be
missed. This is somewhat analogous to digital sampling of EEG
or audio signals. If too few samples are taken, then high fre-
quency waveforms cannot be recorded. If too few electrodes are
used, then voltage changes over areas in between will be com-
pletely missed. A certain number of electrodes is necessary to
avoid spatial undersampling.
Routine EEG recording uses 21 scalp electrode positions.
These are placed according to the international 10-20 system of
electrode placement described in Figure 8.2 (2). Although this
is sufficient for routine clinical recording, to accurately sample
the scalp surface for the detection of small signals generated by
evoked potential recordings, or for research in quantitative EEG
analysis and intracranial localization, it is estimated that at least
100 electrodes would be needed (3).
Spatial sampling in neonates also follows the 10-20 system,
but because of the reduced head size fewer electrodes are used.
F7, F3, F8, F4, P7, P3, P8, and P4 are excluded. Fp1 and Fp2 are
replaced by more posterior electrodes, Fp3 located midway
between Fp1 and F3 and Fp4 located midway between Fp2 and
F4, consistent with the relatively more posterior location of the
frontal lobes in relation to skull measurements at that age (4). If

Figure 8.3 The 10% system of electrode placement is shown, with T7,
T8, P7, and P8 replacing the formerly named T3, T4, T5, and T6 elec-
trodes to maintain naming consistency. The added electrodes at 10%
distances are named according to the surrounding 10-20 electrodes.
Note that in recording an anterior temporal source that FT9 and FT10
placement is usually superior to F7 and F8, which are located over the
posterior lateral frontal lobes.
A1 and A2 cannot be applied to the earlobe then the mastoid
position is used (5).
As illustrated in Figures 8.2 and 8.3, each recording electrode
is named with a letter indicating the head region and a sub-
script number indicating the side of the head (odd numbers are
left and even right) or z for the midline. The further the elec-
trodes are placed laterally from the midline, the higher the sub-
script number (e.g., P7 is lateral to P3). Measurements of the
head are made with a measuring tape (calipers or other meth-
ods should never be used). Measurements are all based on four
landmarks: the nasion (lower forehead between the eyes), the
inion (bone protruberance in the midline back of the head),
and the left and right preauricular point (immediately anterior
to the auditory canal). Measurements are made sequentially as
shown in Figure 8.2 and summarized as follows:
(1) The midline distance between the inion and nasion is
measured. Beginning at 10% of the total distance above the
inion (or nasion), a scalp mark is placed with a wax pencil.
Then three more marks are placed at intervals of 20% of
the total nasion to inion distance to establish the positions
for Fz, Cz, and Pz.
(2) The coronal distance between the left and right preauricular
points intersecting with Cz is measured. Points at 20% and
40% of the total distance are then placed from Cz on the left
and right to mark the positions for C3, C4, T7, and T8.
(3) The distance around the head that passes through T7, T8,
and the points 10% above the inion and nasion is measured
by placing the measuring tape around the patient’s head.
Chapter 8 ■ Polarity and Field Determinations 145
Once that measurement is obtained, Fp1 is placed at 5% of
the total distance on the circumferential line to the left of
the midline. Then marks are placed at intervals of 10% of
the total distance around the head with the last mark at the
Fp2 position. These marks establish the positions for F7,
T7, P7, O1, O2, P8, T8, F8, and Fp2.
(4) The front to back distance from Fp1 and Fp2 running
through C3 and C4 to O1 and O2 respectively, is measured
on each side. At half the distance between Fp1 and C3 on
the left, and Fp2 and C4 on the right, F3 and F4 positions
are marked, respectively. Similarly, the midpoints between
C3 and O1 on the left, and C4 and O2 on the right give the
positions for P3 and P4.
(5) Next, the measurements are taken from F7 to F8 through Fz
to define the transverse coordinates for F3 at the midpoint
between F3 and Fz, and for F4 half way between Fz and F8.
Similarly, measurements from T5 to T6 through Pz locate
the transverse coordinates for P3 midway between T5 and
Pz and for P4 midway between Pz and T6.
If an electrode cannot be placed (due to scalp defects, bandages,
or monitors), then the nearest available electrode site of the
“10-10” combinatorial system should be used, and the corre-
sponding electrode on the other side of the head should be
moved to the same mirror image position.
It should be kept in mind by the electroencephalographer
that there is no substitute for the careful approach to measure
the head outlined above. Unfortunately, it is also the procedure
in which short cuts are taken by some EEG technologists. With
the advent of digital EEG and video, it is reasonable to request
that the technologist include the electrode measurement and
application process in the video recording of each EEG as a
quality control measure.
In some clinical situations (e.g., presurgical epilepsy moni-
toring), better localization may require using more than 21
electrodes. In such cases, selected additional electrodes are
placed according to the 10-10 system described in Figure 8.3.
Physiologic electrical currents are detected using differential
amplifiers. Each amplifier has two electrode inputs, named input
terminal 1 and input terminal 2. The amplifiers subtract the volt-
age of the input 2 electrode from the input 1 electrode (input 1 –
input 2 = recorded EEG signal), hence differential amplification.
Differential amplification is used for the recording of all bioelec-
trical signals (ECG, EMG, EEG). The advantage of differential
amplifiers is that they cancel external noise. The most common
sources of noise arise from other electrical devices. In environ-
ments where current is provided at 60 or 50 cycles/sec, the electri-
cal interference nearly equally contaminates both input 1 and 2
electrode wires with 60 or 50 cycle signals. Because the differential
amplifier subtracts the input 2 signal from the input 1 signal, the
electrical interference that is equal in both inputs cancels out and
is removed from the signal. However, because differential amplifi-
cation is used, the absolute voltage value at electrodes attached to
either input 1 or 2 is never known, only the difference between
them is known. Fortunately, by comparing a number of electrode
positions, the EEG can give a good approximation of the activity
at a particular scalp location relative to other locations.
146 Part I ■ Basic Principles
DIFFERENTIAL AMPLIFIER POLARITY
As noted above and illustrated in Figure 8.1, the currents flow-
ing through the scalp have a polarity that is either positive or
negative. The comparison of electrodes by differential amplifi-
cation yields a difference in voltage and assigns the output a
polarity (negative or positive). As described above, the signal
output and its polarity are achieved by subtracting the voltages
between input 1 and 2. The voltage is displayed on the EEG
monitor or recording paper by the distance the signal deviates
from the zero baseline. The polarity is displayed according to
the direction the waveform deflects (above or below the base-
line). If the voltage in the electrode in input 1 is 50 V and
the voltage in the electrode in input 2 is +50 V, then the EEG
signal will be input 1 minus input 2, or in this case ( 50)
(+50) = 100 V. In addition to knowing that input 1 is sub-
tracted from input 2, the electroencephalographer has to
understand how the amplifier orients the direction of the wave-
forms upward or downward to indicate the relative polarity of
the signal.
As shown in Figure 8.4, the convention for biologic recording
is that, if input 1 is more negative than input 2, the output signal
will deflect upward, whereas if input 1 is more positive than
input 2, then the output signal will deflect downward. This is the
same as stating that if input 2 is more positive than input 1, then
the signal will deflect upward, and if input 2 is more negative
than input 1, then the signal will deflect downward. Also, if the
polarity and voltage in the input 1 electrode is the same as in the
input 2 electrode, then the output signal will be a flat line (zero).
At this point it is recommended that the reader test their
understanding of the information presented in Figure 8.4 by
answering the following question:
A potential of 70 V in input 1 and 80 V in input 2 of
an EEG amplifier yields what voltage and what direction of
Figure 8.4 Biological amplifiers use differential amplification in which
the voltage in the electrode plugged into input 2 is subtracted from the
voltage in the electrode(s) plugged into input 1. In EEG, EMG, and
evoked potential recording, by convention if the relative voltage differ-
ence is negative the signal will deflect upward and if positive, down-
ward. However, as shown in the figure, the voltage source in the brain
can produce opposite deflections depending on whether the input 1 or
input 2 electrode is closest to the scalp source.
signal deflection? Which of the following is correct? (A) +10
and downward, (B) +10 and upward, (C) 150 and downward,
(D) 150 and upward, and (E) +/ 0. The answer is located at
the very end of the chapter. Do not look until you have tried to
answer the question on your own.
SCALP CURRENT FIELD LOCALIZATION:
SPATIAL FILTERING
Scalp potential fields overlap and vary in size and location.
Widely distributed fields are often higher in amplitude than
fields that occupy a small area of the scalp. When viewed simul-
taneously, smaller fields are frequently obscured by or lost in the
higher amplitude widespread fields. However, there is a way to
spatially filter out the more widespread fields so that the more
localized ones can be easily seen. The use of spatial filters is
extremely important because many EEG signals, such as epilep-
tiform spikes, often occupy small scalp areas.
The electrodes in inputs 1 and 2 of each line of amplified
EEG signal channel displayed are referred to as the electrode der-
ivation. Derivations are written with the name of the electrode
in input 1 followed by the name of the electrode in input 2, sep-
arated by a minus sign. Fp1-F3 is an example of a derivation. All
derivations are also mathematical expressions. In the case of
Fp1-F3, the output of the differential amplifier displayed on the
monitor screen will be the voltage detected at Fp1 minus the
voltage detected at F3 or Fp1-F3.
A derivation is referred to as bipolar if it consists of adjacent
electrodes on the scalp. Fp1-F3 is therefore a bipolar derivation
because Fp1 and F3 are located next to each other. In contrast
to bipolar derivations, those with longer interelectrode dis-
tances are referred to as referential. An example of a typical ref-
erential derivation is Fp1-A1.
A display that consists of a combination of derivations is
referred to as a montage. Montages that use long interelec-
trode distances are reference montages, whereas those that
use bipolar derivations are referred to as bipolar montages.
The use of bipolar versus reference montages allows the elec-
troencephalographer to visualize localized versus widespread
scalp fields.
The concept of spatial filtering is illustrated in Figure 8.5.
The bipolar derivation (closely spaced electrodes) on the left
side of Figure 8.5 shows the electrode in input 1 immediately
overlying a cortical generator represented by the black oval area,
and the input 2 electrode is outside the area of the black cortical
generator. Both electrodes are within the field of the larger
underlying cortical generator represented by the gray oval area.
As input 2 is subtracted from input 1 the activity in the black
area is accurately represented as the difference between the two
electrodes (black dashed line waveform at the amplifier out-
put). In contrast, the larger gray field produces a similar voltage
at both input 1 and 2 electrodes in the bipolar derivation and is
therefore filtered out by cancellation.
The referential derivation on the right side of Figure 8.5 has a
much larger interelectrode distance. The input 1 electrode is
within the field of both the small black cortical area and the larger
gray cortical area, whereas the input 2 electrode is outside of both
 Chapter 8 ■ Polarity and Field Determinations 147

Figure 8.6 A longitudinal bipolar montage records a left hemisphere

Figure 8.5 This figure illustrates the filtering effects of interelectrode
seizure that is not visible. (From Fisch BJ, Padin-Rosado J. EEG local-
distance on the recording of underlying EEG scalp potential fields. The
ization and interpretation. In: Fisch BJ, ed. Epilepsy and Intensive Care
closer spaced bipolar electrodes (left) filter out the more widespread
Monitoring. New York, NY: Demos Medical Publishing; 2010:57–76.)
field (gray cortical area) because both inputs see the same voltage. The
more localized field (black cortical area) is clearly displayed. The more
widely spaced referential electrodes display a combination of both fields recording. A widespread left frontal seizure is not visible in the
in which the more localized (black cortical area) is obscured by the longitudinal bipolar montage (Fig. 8.6), but becomes easily visi-
larger (gray cortical area) field but the larger field is now seen (see also ble in the transverse bipolar montage (Fig. 8.7) and in the refer-
the discussion in the text). ential montage (Fig. 8.8). The input 2 reference of the referential
montage consists of an electrode placed at the top of the ster-
num and an electrode placed over the C6 vertebral protruber-
fields. This means that it will amplify a summation of both fields. ance and is referred to as a noncephalic neck-chest reference. It
The advantage is that it “sees” the signal from the larger gray cor- has the advantage of eliminating reference contamination by the
tical generator that is eliminated from the bipolar recording. The scalp, but is sometimes overwhelmed by the ECG signal. This
disadvantage is that by combining the two generators, the local- illustration also makes the important point that for some
ized black field is obscured in the larger signal produced by the patients it is not sufficient to review the record using only one
gray field. The spatial filtering effect shown in Figure 8.5 explains montage or important electrographic events can be missed.
how widespread fields can be removed from the recording. It also
illustrates that overall EEG amplitude increases with increasing
interelectrode distances. This tends to occur with distances up to
about 10 cm. This is also the reason why EEGs performed for the
determination of electrocerebral inactivity (brain death) are
recorded with long interelectrode distances.
Early in the history of EEG, clinical schools developed that
strongly stressed the exclusive use of either bipolar or referential
recording. Neither is superior to the other, they are simply viewed
as different spatial filters that emphasize different aspects of the
scalp topography of EEG voltage fields.
Montages are constructed in either a linear anterior to pos-
terior or transverse pattern across the scalp to simplify interpre-
tation. Bipolar montages are constructed in overlapping chains
(e.g., Fp1-F3 followed by F3-C3) in either an anterior to poste-
rior direction (longitudinal bipolar montages) or in a transverse
left to right direction (transverse bipolar montages). Referential
montages are usually displayed with electrodes in successive
channels in an anterior to posterior direction, with the same
reference electrode (or electrodes) in input 2 of every channel.
Longitudinal and transverse bipolar montages are both used
because scalp voltage fields can be asymmetric along either ante- Figure 8.7 The same recorded epoch as in Figure 8.6 using a trans-
rior to posterior or transverse axis. Such asymmetries can only verse bipolar montage now clearly reveals the presence of an electro-
be reliably assessed by either using both a longitudinal and a graphic seizure most easily seen in the fifth channel (F3-Fz). (From
transverse bipolar montage, or by using one bipolar montage Fisch BJ. EEG localization and interpretation. In: Fisch BJ, ed. Epilepsy
and reference montage. This is dramatically illustrated in Figures and Intensive Care Monitoring. New York, NY: Demos Medical
8.6 through 8.8 (6). Each figure shows the same epoch of EEG Publishing; 2010:57–76.)
148 Part I ■ Basic Principles
Figure 8.8 The same recorded epoch as in Figures 8.6
and 8.7 using a noncephalic neck–chest electrode pair
reference reveals the widespread seizure with maximal
amplitude epileptiform spikes at F3 and very little activ-
ity at Fz. (From Fisch BJ. EEG localization and interpre-
tation. In: Fisch BJ, ed. Epilepsy and Intensive Care
Monitoring. New York, NY: Demos Medical Publishing;
2010:57–76.)
SELECTING SPATIAL FILTERS:
MONTAGE DESIGN
There are five commonly used montage designs and each has a
unique spatial filtering characteristic. The montages that can pro-
duce the severest filtering of widespread fields and the greatest
enhancement of highly localized fields are the longitudinal bipo-
lar and transverse bipolar montages (described earlier). These
montages are followed in order of greatest to least in filtering of
widespread fields as follows: the Laplacian montage, the average
reference montage, and the common reference montage.
The Laplacian montage is implemented as a local average ref-
erence montage in which the electrode in input 1 is referred to
a combination of the immediately surrounding electrodes in
input 2, for example, Cz ¼ (Fz + C3 + Pz + C4). Each chan-
nel of the montage has a different reference combination in
input 2 that depends on the location of the electrode in input 1.
The Laplacian montage compliments the bipolar when identi-
fying highly localized fields with low amplitude, such as a focal
electrographic seizure onset. A reduced filtering effect that pro-
duces less suppression of regional fields can be achieved by a
Laplacian montage that incorporates all electrodes on the scalp
except the input 1 electrode into the reference. This is done by
weighting the contributions of the reference electrodes accord-
ing to their distances from the input 1 electrode (the farther
away the input 2 electrode is from the input 1 electrode, the less
it contributes to the combined voltage of the input 2 elec-
trodes). This montage is not commercially available but can be
implemented on many systems using a spreadsheet (7).
The average reference montage consists of a single electrode
in input 1 referred to a combination of all or most of the
remaining scalp electrodes. The average reference montage is
good for viewing highly localized medium to high amplitude
fields and somewhat more widespread regional fields. When a
high amplitude or very widespread activity occurs, the average
reference itself will contain that activity, sometimes producing
a confusing picture. For example, eye movement waveforms
that typically have a voltage field maximum at Fp1 and Fp2 will
often appear upside down simultaneously in the occipital elec-
trodes O1 and O2, not because there are eye movement poten-
tials in O1 and O2, but because the activity is actually coming
from the reference electrodes that include the activity detected
by Fp1 and Fp2. Therefore, it is important to be aware that
waveforms that appear upside down in distant head regions in
an average reference montage are likely to be due to reference
contamination. However, the electrodes with the highest ampli-
tude are almost always closest to the true source of the activity.
The common reference montage consists of the same one or
two electrodes in input 2 in all channels of the montage. The
common reference montage produces the least attenuation of
widespread fields, particularly when the reference electrodes are
not placed on the head (noncephalic neck-chest reference). A
widely used reference montage uses the A1 and A2 ear reference
electrodes. All left-sided electrodes are often referenced to A1
with all right-sided electrodes referenced to A2. A1 and A2 refer-
ence electrodes are particularly useful during sleep when high
amplitude activity is occurring over the central head regions that
would otherwise contaminate the reference electrodes with
excessive activity. In contrast, during wakefulness, a common
reference montage using Cz is often used because central activity
is generally low in amplitude and the relative absence of muscle
at the vertex makes it less likely that Cz will be contaminated
with muscle artifact. The advantage of the common reference
montage, particularly the noncephalic neck-chest reference is
that it produces a simple view of the scalp field, where the input
1 electrode whose channel has the highest amplitude identifies
the scalp location with the field maxima. As will be seen in the
following section, this works well unless the reference electrode
is within the field of interest (reference contamination).
LOCALIZING CURRENT FIELDS: PHASE
CHANGES AND DIFFERENTIAL
AMPLIFICATION
Localization using bipolar montages is accomplished by analyzing
phase orientation (the upward or downward deflection of the
waveform) and amplitude. Localization using reference mon-
tages is accomplished mainly by analyzing amplitude. Bipolar

Figure 8.9 Instrumental phase reversal.
localization depends on changes in phase between channels. Phase
refers to the upward or downward direction of a waveform. A
phase reversal has occurred if the phase has changed direction
from up to down or from down to up. Figure 8.9 shows an exam-
ple of a phase reversal between channels 1 (derivation Fp1-F3)
and 2 (derivation F3-C3).
The rules for localization in bipolar montages are summa-
rized in Table 8.1. Examples of these rules are illustrated in
Figures 8.9 through 8.12. The electrodes in these figures are
shown as black circles and the lines leading from them plug into
input 1 (the higher line going into the derivation) or input 2
(the lower line going into the derivation). They are arranged as
a longitudinal bipolar montage.
Figure 8.9 shows an instrumental phase reversal between
channels with the derivations Fp1-F3 and F3-C3. The field over
the scalp is displayed by the dashed lines with the highest voltage
not covered by an electrode. The field potential is negative.
Therefore, the channel 1 deflection is downward (input 1 is more
positive than input 2) whereas the deflection in channel 2 is
upward (input 1 is more negative than input 2) (see Fig. 8.4). The
third channel deflection is also upward because input 1 is more
negative than input 2. The last channel shows no deflection
because both electrodes, P3 and O1, are the same. The electrode
common to both channels that reverse phase localizes the maximum
field potential to that common electrode. In this case that is F3.
Tabl e 8 . 1
Rules of Localization: Bipolar Montages
The electrode with the highest voltage potential is found by
1. Instrumental phase reversal
2. Instrumental phase reversal with cancellation in an
intervening derivation
3. End of chain phenomenon or the absence of phase
reversal
4. End of chain phenomenon with cancellation
5. True phase reversal (double phase reversal)
Chapter 8 ■ Polarity and Field Determinations 149
Figure 8.10 Instrumental phase reversal with cancellation in an inter-
vening channel.
Figure 8.9 illustrates the concept of instrumental phase reversal
because it is the instrument (i.e., the way the electrodes are
arranged with the input 2 electrode becoming the input 1 elec-
trode in the next channel) that causes the phase reversal. In con-
trast, a true phase reversal would mean that two different polarities
are simultaneously present in adjacent cortical areas. The occur-
rence of a true phase reversal of cortical polarity would produce
two instrumental phase reversals in a linear chain of bipolar elec-
trodes. In the event that there is no phase reversal in a bipolar
chain (i.e., all the waveforms point in the same direction), then the
field maxima is closest to the electrode at the end of the chain that
shows the most rapid change in voltage (highest amplitude) in
successive channels. This is referred to as the end of chain phe-
nomenon. It is localizing to the electrode at the end of the elec-
trode chain and does not suggest the source is deep in the brain.
Figure 8.10 illustrates an instrumental phase reversal with
cancellation in an intervening channel. The field on the scalp
has a negative polarity with maximal voltage in the center of the
dashed lines. The first channel deflection is downward because
input 1 is more positive than input 2. The second channel is flat
because the input 1 and 2 electrodes, F3 and C3, see the same
voltage. The third channel deflects upward and is reversed from
the first channel because input 1 is more negative than input 2.
The intervening channel with cancellation (F3-C3) localizes the
field maximum between the electrodes in that channel.
Figure 8.11 End of chain phenomena with cancellation.
150 Part I ■ Basic Principles

Figure 8.12 Ipsilateral ear reference with the negative field maximum
closest to F3.

Figure 8.11 illustrates the end of chain phenomena with can-

cellation. Channel 1 electrodes detect the highest voltage, but
because it is the same in both electrodes the output is zero, a flat
line. As in Figure 8.10 this localizes the field maximum between
those two electrodes (Fp1 and F3) at the end of the bipolar Figure 8.13 Tangential dipole with cancellation in bipolar derivations
chain. The next channel shows an upward deflection because and true phase reversal in common reference derivations; anterior max-
input 1 is more negative than input 2. The same applies to the imal negativity with posterior positivity. (From Fisch BJ, Padin-Rosado
third channel. The fourth channel is flat because there is no J. EEG localization and interpretation. In: Fisch BJ, ed. Epilepsy and
voltage difference between those electrodes. Intensive Care Monitoring. New York, NY: Demos Medical Publishing;
Without the accompanying head figure it is reasonable to ask 2010:57–76.)
why the changes in this montage are not interpreted as a posi-
tive field in the back of the head instead of negative field in the
front of the head. While that is possible it is unlikely. The rate of input 1 electrode closest to the field maxima. If all channels
change in voltage (i.e., the amplitude of the waveform) between containing a reference show the same signal output then the
two electrodes is usually greatest near the field maxima and activity is coming from the reference, not the input 1 electrode.
least at a distance from it. In this case, the presence of a higher Figure 8.12 illustrates a reference recording of a negative
amplitude in channel 2 than in channel 3 localizes the signal field in which the reference electrode, A1, is outside of the scalp
source (field maxima) to the front of the head. field. In this case, the derivation with the highest amplitude sig-
A true phase reversal is shown in Figure 8.13. True phase nal shows the input 1 electrode that is closest to the field max-
reversals are rarely encountered except in situations where the ima (F3). The deflections are all upward because input 1 is more
cortical anatomy has been disturbed (e.g., cerebral palsy) or more negative than input 2.
often as an identifying feature of benign childhood epilepsy with Figure 8.14 illustrates reference contamination. In this
centromidtemporal spikes (BCECTS). In these patients the example, the reference electrode A1 is within the positive scalp
epileptiform spike arises at an angle that is parallel to the cortical field. The other electrodes are outside of the field. Therefore, the
surface, allowing both the negativity from the outer cortex and deflection seen in all channels is the same and is upward
positivity from the deeper cortex to be seen simultaneously. In because for each channel, input 1 is more negative than input 2.
BCECTS the positivity projects anteriorly, for example, appearing
as an upward deflection in Fp1-F3 and downward in F3-C3. The
negativity appears more posteriorly and in this example would
appear as a downward deflection in C3-P3 and upward in P3-O1.
The anterior positivity is typically lower in voltage than the pos-
terior negativity and may require increasing the gain or the use of
a neck-chest reference to be visualized.
As noted above, common reference montages are most use-
ful for identifying widespread scalp fields and least useful for
identifying highly localized low amplitude fields. Although
localization is usually simple with a reference montage, a con-
fusing picture arises when the reference electrode(s) are within
the field of the activity of interest. In that case, the activity in a
channel may be coming from the reference instead of the input Figure 8.14 Ipsilateral ear reference with reference contamination.
1 electrode. If the reference is not in the scalp field of interest, The signal is from the reference electrode as the other electrodes are
then the channel with the highest amplitude identifies the outside of the scalp field.

Figure 8.15 Reference contamination.
Figure 8.15 illustrates how confusing the picture can become
when the reference is contaminated with activity also detected
by the input 1 electrodes. Channel 1 deflects downward because
the reference is relatively negative (input 1 is more positive than
input 2). Channel 2 is flat because input 1 and input 2 see the
same negative voltage. Channel 3 deflects upward because input
1 is more negative than input 2, and channel 4 deflects down-
ward because input 1 is more positive than input 2.
A field with a single polarity will not produce waveforms of
opposite phase in a referential recording between adjacent
channels unless the reference is contaminated. Two phase rever-
sals as shown in Figure 8.15 (channel 1 vs. channel 3 and chan-
nel 3 vs. channel 4) virtually always assure the reader that
reference electrode contamination is occurring. A single-phase
reversal in a common reference recording without reference
contamination, as shown in the lower three channels in Figure
8.13, represents a true (not instrumental) phase reversal, in that
it indicates the presence of a horizontal (tangential) cortical
dipole.
An example of how to approach localization is shown in
Figure 8.13. In the initial longitudinal bipolar recording
epileptiform spikes are present in only one channel, creating a
complex problem in localization. The problem is quickly
solved by using a reference montage. At the time of EEG inter-
pretation new derivations were selected using a distant refer-
ence electrode unlikely to be contaminated by the field of the
epileptiform activity. In this case the reader selected the most
distant electrode, P8 (opposite side of the head, lateral and
posterior) for the reference to explore the activity in Fp1, F7,
and T7. The added derivations are in the last three channels in
Figure 8.13. They show that there are similar amplitude spikes
in Fp1 and F7. This explains why the first channel is relatively
flat (very low amplitude upward deflections are present), there
is cancellation between Fp1 and F7 (if input 1 and 2 are nearly
equal, the output is nearly zero). T7-P8 demonstrates a phase
reversal with positivity arising from T7. If opposite polarities
are subtracted as shown in the left lower part of Figure 8.4,
then the resulting signal will summate to create a higher
amplitude signal than that in either input alone. This effect
creates the high amplitude spikes in channel 2 (F7-T7). T7
and P7 share a similar amplitude positivity and therefore can-
cel out in the third channel. The positivity of P7 in relation to
O1 can barely be seen in the fourth channel P7-O1.
Chapter 8 ■ Polarity and Field Determinations 151
EEG INSTRUCTION IN LOCALIZATION
Those who teach electroencephalography should be aware that
many, if not most, physicians who read EEGs do not under-
stand how to localize EEG findings. It is a disservice to patients
with neurologic disorders when teachers fail to emphasize the
importance of localization and practicing physicians continue
to believe that EEG interpretation is only a matter of pattern
recognition. Although it is not possible to reach physicians who
have completed their training and are already out in practice, it
is possible to improve the skills of those currently being trained.
When training residents and fellows in EEG localization it is
helpful to ask certain questions that assess their progress. First,
what is the polarity of an upward deflection in a single channel of
recording? Most students initially believe that an upward deflec-
tion in a channel is negative. Of course, the correct answer is that
there is no way of knowing. In a single channel one never knows
if an upward deflection is caused by a negative potential in input
1 or a positive potential in input 2. Any misconceptions can usu-
ally be further resolved by a discussion of positive occipital sharp
transients of sleep (POSTS), because they always deflect upwards
in the longitudinal bipolar montages. Second, ask, is a phase
reversal an abnormal finding? It is a surprising how often stu-
dents believe phase reversals are abnormal. This occurs because
the student never understood the importance of studying local-
ization and the teacher invoked the concept of phase reversal only
in the setting of EEG abnormalities. This situation can be quickly
resolved by discussing phase reversals of vertex waves (or other
normal waveforms) in bipolar montages. Third, ask, what is the
difference between an instrumental phase reversal and a true
phase reversal? If the student can correctly answer these basic
questions then they are well on their way to understanding EEG
localization. Lastly, never allow students to speak in terms of head
regions when localizing EEG findings. Instead insist that they
speak in terms of actual scalp spatial sampling points by using
electrode names. (The answer to the question in the section on
differential amplification and polarity is A).
REFERENCES
1. Gloor P. Neuronal generators and the problem of localization in
electroencephalography: application of volume conductor theory
to electroencephalography. J Clin Neurophysiol. 1985;2:327–354.
2. Fisch BJ. Fisch and Spehlmann’s Digital and Analog EEG Primer.
Amsterdam: Elsevier; 1999.
3. Srinivasan R, Tucker DM, Murias M. Estimating the spatial
Nyquist of the human EEG. Behav Res Methods Instrum Comput.
1998;30:8–19.
4. American Electroencephalographic Society Guidelines in EEG,
1–7 (Revised 1985). J Clin Neurophysiol. 1986:3131–3168.
5. Park KL, Nordli DR. Special considerations in neonatal monitor-
ing. In: Fisch BJ, ed. Epilepsy and Intensive Care Monitoring. New
York, NY: Demos Medical Publishing; 2010:91–103.
6. Fisch BJ. EEG localization and interpretation. In: Fisch BJ, ed.
Epilepsy and Intensive Care Monitoring. New York, NY: Demos
Medical Publishing; 2010:57–76.
7. Lemos ML, Fisch BJ. The weighted average montage. Electroenceph
Clin Neurophysiol. 1991;79:361–370.

Normal EEG and Sleep:
Preterm and Term Neonates
ELI M. MIZRAHI, SOLOMON L. MOSHÉ, AND RICHARD A. HRACHOVY
INTRODUCTION
The interpretation of the neonatal electroencephalogram con-
sists of related phases of analysis most of which are conducted
in parallel. Initially, the EEG is assessed to determine the con-
ceptional age (CA) of the infant. In concert with this, the back-
ground activity is assessed to determine whether it is normal,
and if not, to characterize the abnormal features. In addition,
the determination of CA provides the basis for the understand-
ing of the range of both diffuse and focal abnormalities that
may be anticipated within each age-dependent epoch. Thus,
visual analysis of the neonatal EEG begins with the recognition
of the CA-dependent features characteristic of specific epochs
of development. This is followed by inspection for EEG abnor-
malities based on an understanding of possible age-dependent
findings. Taken together these aspects of visual analysis of the
neonatal EEG provide the basis for the determination of CA of
the infant and the assessment of the degree to which the EEG is
normal.
Two additional principles of interpretation of the neonatal
EEG relate to age-dependent changes over time and to find-
ings of indeterminate significance. Typically, the EEG-deter-
mined age of the infant is given in CA—considered the age
from the time of conception. The gestational age (GA) is the
time from conception to the time of birth. The chronologic or
legal age is the time from birth to the time of the EEG. The CA
is the GA plus the legal age. A basic principle of interpretation
of the neonatal EEG is that the immature brain develops at the
same rate whether in utero or following delivery. Thus, the
EEG recorded several weeks after an infant was born prema-
turely would have similar features to an EEG recorded on the
first day of life of infant whose CA equaled the GA plus the
legal age of the premature infant. Another important consid-
eration is that the clinical significance of all features of the
neonatal EEG has not been determined. Thus, the challenge is
to differentiate the known from the unknown and focus inter-
pretation on those aspects of the EEG for which accurate data
are available.
A number of investigators have, over the years, contributed
to our understanding of the normal neonatal EEG beginning
with the pioneering work of the French school (1–5). More
recent discussions of normal neonatal electroencephalography
include those by Hrachovy et al. (6), Watanabe et al. (7), Clancy
et al. (8), Mizrahi et al. (9), Plouin et al. (10), Vecchierini et al.
(11), and André et al. (12).
Part II Normal EEG
CHAPTER
9
TECHNICAL CONSIDERATIONS
The general principles of recording the electroencephalogram
in neonates are similar to those of older children and adults
with some important additions and exceptions. Guidelines for
the recording of the neonatal EEG have been established by the
American Clinical Neurophysiology Society (13) and the
International Federation of Clinical Neurophysiology (14).
Critical to the recording of neonatal EEG is a well-trained
staff of electroneurodiagnostic technologists (ENDTs) with
expertise in the recording of newborn and young infants. To
insure maximum clinical relevance, the ENDT should obtain
basic information about each neonate to be recorded including
standard demographic data, description of the recording envi-
ronment, documentation of reason for referral, details of the
medical history, a list and timing of medications, and the
specifics of the infant’s general medical condition. In addition,
the technologist must consider the state of each infant and
determine the best method to make the infant as comfortable as
possible in order to obtain a complete recording. This may
require having the infant fed, having diapers changed, adjusting
room temperature and, often, just prolonging the recording
until the infant becomes comforted.
The International 10-20 System of Electrode Placement has
been modified for recording neonates. Typically, a minimum of
nine scalp positions are utilized (F1, F2, C3, C4, CZ, T3, T4, O1,
and O2). In addition, electrodes are placed at A1 and A2 and a
ground electrode is placed either at mid-forehead or on a mas-
toid region. Since digital recordings are fundamentally referen-
tial, an additional reference electrode position may be needed
(typically noncephalic), although some instruments provide a
so-called “internal” reference. The F1 and F2 electrode posi-
tions are 20% of the inion–nasion distance above the nasion
and 10% of the circumferential measurement from the midline.
Their placement allows for the foreshortened position of the
neonatal immature frontal lobes. Placement of all of the stan-
dard electrodes of the International System would result in such
close spacing on the neonate’s scalp that electrode recording
may overlap with redundant electrical fields. Paste, not collo-
dion, is utilized for electrode placement.
Polygraphic measures are integral to the recording of the
EEG in order to assist in characterizing sleep states, eye move-
ments, muscle contractions, cardiac rhythms, and respiratory
patterns and to assist in identification of noncephalic artifact.
The polygraphic measures include the electrooculogram
153
154 Part II ■ Normal EEG

(EOG), submental electromyogram (EMG), electrocardiogram
(ECG), and respiratory monitors such as a strain gauge, bipolar
electrodes, or pneumograph.
The recording of the neonatal EEG includes the documenta-
tion of wake/sleep cycles, the characterization of reactivity of
the record to stimulation and identification of age-dependent
waveforms and other features. These are best achieved utilizing
a single, bipolar montage with broad coverage over the scalp.
Digital recording of the neonatal EEG provides the opportunity
to examine various waveforms with different montages after
recordings are complete.
Instrument settings used at the onset of recording are listed
in Table 9.1. The filter settings and sensitivity of the EEG chan-
nels should be the same as that for EOG in order to allow accu-
rate comparison of waveforms to differentiate cerebral activity
from that of ocular origin. Paper speed is set at 30 mm/sec for
analog recordings or 10 sec/screen or “page” for digital record-
ings. These paper speed settings are utilized in many laborato-
ries in the United States. However, several laboratories,
particularly those with ties to the French School of Recording,
utilize a slow speed: 15 mm/sec or 20 sec/“page.”
The most successful and clinically relevant neonatal EEG
recordings are those in which objectives and strategies are iden-
tified prior to the beginning of each study. Advance planning is
essential. The basic tasks are to obtain historical data, determine
reason for referral, initiate technical recordings, examine the
EEG in real time, observe the infant for clinical behaviors,
record the infant in sleep and wakefulness, and attempt to pro-
voke abnormal paroxysmal clinical events. Close observation of
the infant at all times is particularly important when recording
a neonatal EEG. The record should be annotated when behav-
ioral or autonomic changes occur or when other events happen
that may affect the record.
A final technical consideration relates to the duration of the
neonatal EEG recording. There are two considerations: the time
it may take for the infant to experience the full cycles of sleep
and wakefulness, and a sufficient period of time for the infant
to experience clinical or electrical events. The typical minimum
duration of recording for a neonatal EEG is 1 hour. However,
the recording duration is an interactive process based on clini-
cal circumstances and the unfolding EEG.
AGE-DEPENDENT DEVELOPMENTAL
FEATURES
Initial analysis includes the assessment of the degree of conti-
nuity of background activity and the degree of interhemi-
spheric synchrony of the background activity. Further analysis
consists of inspection for specific age-dependent waveforms
and patterns (so-called “grapho-elements”) (Table 9.2) and the
presence and character of sleep/wake cycles. In addition, there
is recognition of some special waveforms and patterns that
occur in the near-term or term infant that are considered devel-
opmental milestones and variations of the normal EEG.
Continuity
The earliest appearance of electrical activity on the EEG is char-
acterized by a pattern of discontinuity, with long periods of qui-
escence (Fig. 9.1). This pattern may be present in all states of
waking and sleep depending on CA and has been referred to as

Tabl e 9 . 1

Initial Instrument Settings for Recording Neonatal EEG

EEG channels
Sensitivity 7 V/mm
Time constant 0.3 sec
High-frequency filter 70 Hz
Notch filter (60 Hz, 50 Hz) Off
Display time
Paper speed 30 mm/sec*
Screen display 10 sec/“page”*
EOG** EMG*** Respir.*** ECG***
Polygraphic channels
Sensitivity 7 V/mm 7 V/mm 7 V/mm 300 V/mm
Time constant 0.3 sec 0.01 sec 0.3 sec 0.3 sec
High-frequency filter 70 Hz 70 Hz 70 Hz 70 Hz
Notch filter (60 Hz, 50 Hz) Off Off Off Off

*Some laboratories may choose to record at “half paper speed” with paper speed setting of 15 mm/sec or screen display of 20 sec/“page.”
**EOG settings remain the same as EEG settings for comparison of simultaneously recorded waveforms.
***EMG, Respir. (pneumograph), and ECG settings may be adjusted to optimize display.
 Tabl e 9 . 2

Age-Dependent Features of the Neonatal EEG

CA (weeks) Continuity Synchrony Landmarks Wake/Sleep Cycles Reactivity

24–26 Discontinuous Asynchronous Beta–delta complexes may Limited behavioral or No reaction to stimulation
(trace discontinu) be present, although polygraphic changes
absence is not abnormal No cyclical EEG changes
27–28 Discontinuous Asynchronous Beta–delta complexes first Limited behavioral or No reaction to stimulation
Interburst intervals Traditionally appearing in the central polygraphic changes
are CA dependent characterized regions No cyclical EEG changes
and less than as an epoch of Temporal theta bursts
preceding epoch hypersynchrony (rudimentary)
29–30 Discontinuous Asynchronous Beta–delta complexes in Tendency for greater continuity No reaction to stimulation
Interburst intervals Greater than 50% the central regions of background activity
are CA dependent activity Temporal theta bursts during behavioral REM sleep
and less than asynchronous on (most consistent during
preceding epoch the two sides this epoch)
Some brief periods of
relative continuity
31–33 Discontinuity during Asynchronous Beta–delta complexes more Continuity during REM sleep No reaction to stimulation
wakefulness Degree of synchrony prominent in temporal Discontinuous during
Continuity during greater than and occipital regions wakefulness and NREM sleep
behavioral sleep previous epoch Temporal theta bursts
until 32 weeks CA
Temporal alpha bursts
replace theta bursts
at 33 weeks CA
34–35 Discontinuity during Asynchronous Frontal sharp transients EEG and physiologic features Background activity is
NREM, less than the Degree of synchrony (enchoches frontales) of wakefulness, REM, and reactive to stimulation
previous epoch greater than Beta–delta complexes in NREM sleep clearly defined —state dependent
Continuity during previous epoch occipital-temporal region,
wakefulness most often in NREM sleep
and REM sleep
36–37 Continuity during Residual asynchrony Frontal sharp transients persist EEG and physiologic features Background activity is
wakefulness and Degree of synchrony Beta–delta complexes of wakefulness, REM, and reactive to stimulation
REM sleep about 80% on the become less frequent NREM sleep clearly defined —state dependent
Episodes of discontinuity two sides Bifrontal delta may be present
during NREM sleep
38–40 Continuity in all states Synchronous on the Frontal sharp transients persist EEG and physiologic features Background activity is
two sides by 40 Beta–delta complexes of wakefulness, REM, and reactive to stimulation
weeks CA disappear from NREM sleep NREM sleep clearly defined —state dependent
Appearance of trace alternant
41–44 Continuity in all states Synchronous on Frontal sharp transients persist EEG and physiologic features Background activity is
the two sides No other immature of wakefulness, REM, and reactive to stimulation
waveforms are present NREM sleep clearly defined —state dependent
Trace alternant begins to
resolve by 44 weeks CA

155
156 Part II ■ Normal EEG

Figure 9.1 26 to 27 weeks CA. This EEG

demonstrates a trace discontinu pattern with a
burst of bilateral polyfrequency activity. The
burst is somewhat asynchronous and has fea-
tures of beta–delta complexes. (Reprinted with
permission from Mizrahi EM, Hrachovy RA,
Kellaway P. Atlas of Neonatal Electroencephalo-
graphy. 3rd ed. Philadelphia PA: Lippincott
Williams & Wilkins; 2004.)

trace discontinu. As the CA increases, periods of inactivity At approximately 30 weeks CA, continuous activity first
shorten. The identification of interburst intervals (IBIs) is pri- appears, but is state dependent: present only during behavioral
marily based on the degree of attenuation between bursts, rapid eye movement (REM) sleep. At about 34 weeks CA, the
although there is no real consensus among investigators regard- EEG also becomes continuous in the apparent awake state. By
ing a definition of the maximum amplitude of the IBI, ranging 37 to 38 weeks CA, continuity appears in behavioral sleep state
from 30 V (15) to 15 V (16), with 30 V recommended as characterized as non-REM (NREM). However, from that time
the upper limit (17). until about 5 to 6 weeks post-term, during periods of NREM
Several groups of investigators have measured IBIs in order to sleep the EEG demonstrated semi-periodic episodes of gener-
establish normative data at the various CAs. Although all have alized voltage attenuation, not quiescence, lasting from 3 to 15
noted the trend of shortened IBIs with increasing CA, there is seconds; a pattern that has been called trace alternant. These
much variation in mean and maximum IBI values at the earliest periods of attenuation differ from those of trace discontinu
CAs (1,2,15–20). The variation in findings may relate to method- since the trace alternant is characterized by persistence of
ology differences that include amplitude of activity during the waveforms at lower voltage, whereas trace discontinu features
IBI (the lower the allowable amplitude, the shorter the attenuated
interval), localization of attenuated activity of IBI (all channels
simultaneously vs. allowable activity in some channels), method
of measurement of burst onset, and the presence and degree of
underlying central nervous system dysfunction such as hypoxic-
ischemia (which may prolong IBIs, even at subclinical levels).
In most studies of IBI duration, the earliest CA characterized
has been approximately 27 to 28 weeks. There are limited data
concerning the IBIs of younger neonates. Hayakawa et al. (21)
studied infants beginning as young as 21 to 22 weeks CA and
observed that the mean and maximum IBI duration also
decreased with increasing CA. Vecchierini et al. (22) concluded
that in neurologically normal 24- to 26-week-CA infants the
IBIs do not exceed 60 seconds. In addition, Clancy et al. (8) sug-
gested that a starting point for mean IBI in the very premature
infant (24 weeks CA) is 10 seconds.
Figure 9.2 represents a composite of data of maximum IBIs Figure 9.2 Average duration of interburst intervals in NREM sleep in
based on various studies (23) beginning with a CA of approxi- prematurity. Composite of findings from studies of interburst intervals.
mately 26 weeks. Within an individual recording the duration (From Hrachovy RA. Development of the normal electroencephalo-
of IBIs may vary. However, in clinical practice, the IBIs with the gram. In: Levin KH, Luders HO, eds. Comprehensive Clinica l
longest duration are most clinically relevant. Neurophysiology. Philadelphia, PA: WB Saunders; 2000:387–413.)
 Chapter 9
periods of electrical quiescence with virtual absence of electri-
cal activity.
Bilateral Hemispheric Synchrony
It has often been reported that at or prior to 27 to 28 weeks CA,
bursts of activity in the two hemispheres between periods of
quiescence occur as generalized bisynchronous bursts with a
discontinuous background (2,19). However, others have
reported that at 27 to 28 weeks CA the activity is generally asyn-
chronous in homologous regions of the hemispheres (9). The
greater the recording distance from the midline, the greater
the degree of asynchrony (Fig. 9.3). With increasing maturity,
the degree of asynchrony diminishes. The degree of asynchrony
reflects not only maturation but also wake/sleep state of the
infant. Asynchrony is most prominent in NREM sleep and is
least prominent in REM sleep. The only exception to these gen-
eral rules is that from the time frontal sharp waves first appear,
at about 35 weeks CA, they are bilaterally synchronous.
Age-Dependent Waveforms
There is an orderly appearance and disappearance of specific
waveforms and patterns with increasing CA. These age-depend-
ent waveforms have been referred to as “grapho-elements” and
are essential landmarks in assessing CA and determination of
the degree to which the background EEG is normal.
Beta–Delta Complexes
These constitute the principal landmarks of prematurity.
Various names have been given to these complexes: “spindle-
■ Normal EEG and Sleep: Preterm and Term Neonates 157
Figure 9.3 29 to 30 weeks CA. Beta–delta com-
plexes are present predominantly in the left central
region (shaded oval). Temporal theta bursts are
present independently on the left and right (shaded
square). The EEG is discontinuous and asynchro-
nous. (Reprinted with permission from Mizrahi EM,
Hrachovy RA, Kellaway P. Atlas of Neonatal
Electroencephalography. 3rd ed. Philadelphia, PA:
Lippincott Williams & Wilkins; 2004.)
delta bursts,” “brushes,” “spindle-like fast waves,” and “ripples of
prematurity.” Dreyfus-Brisac et al. (24), who first described the
complexes, referred them as “rapid bursts,” emphasizing the fast
component of the complexes.
They are present from about 26 to 38 weeks CA and consist of
randomly occurring 0.3- to 1.5-Hz waves of 50 to 250 V, with
superimposed bursts of low- to moderate-voltage fast activity
(Fig. 9.3). The frequency of the fast activity may vary, even in the
same infant. Two frequencies predominate: 8 to 12 Hz and, more
commonly, 18 to 22 Hz. The voltage of the fast activity varies
throughout each burst but rarely exceeds 75 V. Until 32 weeks
CA, the fast component has a predominant frequency of 18 to 22
Hz; thereafter, the slower frequency is most often present.
The complexes first appear as a dominant feature in the EEG
at about 26 weeks CA. When first present they occur infre-
quently, predominantly in the central regions. During the next
5 to 6 weeks, they become progressively more persistent, and
the voltage of the fast component usually increases. From 29 to
33 weeks CA, the pattern is a prominent feature during behav-
ioral REM sleep. The spatial distribution of the beta–delta com-
plexes is also age dependent, becoming more prominent in the
occipital and temporal areas with increasing age. Beta–delta
complexes typically occur asynchronously in derivations from
homologous areas and show a variable voltage asymmetry on
the two sides.
At 33 weeks CA, beta–delta complexes are maximally
expressed in NREM sleep, rather than REM sleep, and appear
more prominently in the temporal-occipital areas. From 33 to
38 weeks CA, beta–delta complexes continue to occur primarily
158 Part II ■ Normal EEG

Figure 9.4 33 weeks CA. Temporal alpha bursts

are present on the left (shaded square) and
beta–delta complexes are present in the occipital
(shaded oval) and temporal regions (shaded
oval). The EEG is discontinuous and asynchro-
nous. (Reprinted with permission from Mizrahi
EM, Hrachovy RA, Kellaway P. Atlas of Neonatal
Electroencephalography. 3rd ed. Philadelphia, PA:
Lippincott Williams & Wilkins; 2004.)

in NREM quiet sleep. The disappearance of the beta–delta com- relation to other developmental features. The occurrence of
plexes when the infant appears behaviorally awake constitutes temporal alpha bursts defines a more narrow CA range at 33
one marker of 36 to 37 weeks CA. weeks.

Temporal Theta and Alpha Bursts Frontal Sharp Waves

Temporal theta bursts appear at about 26 weeks CA and are
maximally expressed between 30 and 32 weeks (Fig. 9.3). They
then rapidly disappear—replaced by temporal alpha bursts, a
very specific marker for 33 weeks CA. Temporal theta bursts are
characterized by rhythmic 4.5- to 6.0-Hz waves occurring inde-
pendently in short bursts of rarely more than 2 seconds from
the left and right midtemporal areas. Voltage varies from
roughly 20 to 200 V. Some individual waves often have a sharp
configuration. Temporal alpha bursts have similar features
except the frequency of their waveforms is greater than 6.0 Hz
(Fig. 9.4).
In practical terms, strongly expressed temporal theta bursts
can serve as a useful developmental landmark in determining
CA between 30 and 32 weeks—particularly when considered in
These are characterized as isolated sharp, but bluntly config-
ured waves. They usually have an initial surface-negative phase
followed by a surface-positive phase. They have been referred to
as “enchoches frontales.” Frontal sharp waves may first appear
at 34 weeks CA but attain maximum expression at about 35
weeks CA (Fig. 9.5). When they initially appear they are low
voltage and infrequent in occurrence, with a less well-defined
waveform than when they are maximally expressed later. After
44 weeks CA they diminish in number and voltage and are only
rarely seen in infants older than 6 weeks post-term.
Frontal sharp transients are bilaterally synchronous and
symmetrical from the time of their first appearance. The initial
surface-negative component lasts about 200 msec. The succeed-
ing surface-positive component lasts somewhat longer and is

Figure 9.5 36 to 37 weeks CA. The awake EEG

is characterized by frontal sharp transients that
occur bilaterally, synchronously, and simultane-
ously (shaded square). The background activity
is continuous and synchronous, with some per-
sistent beta–delta complexes in occipital regions.
(Reprinted with permission from Mizrahi EM,
Hrachovy RA, Kellaway P. Atlas of Neonatal
Electroencephalography. 3rd ed. Philadelphia, PA:
Lippincott Williams & Wilkins; 2004.)
 Chapter 9
variable. Although typical blunt in configuration, when they are
maximally expressed, the waveforms may be quite sharp in
morphology. Frontal sharp transients typically occur randomly
as single events, predominantly in transitional rather than in
REM or NREM sleep. However, they may also recur in brief
runs and may also be mixed with bifrontal delta activity, which
is another normal feature of near-term infants (see below).
STATE CHANGES
The EEG in Wakefulness and Sleep
Until 36 to 37 weeks CA, distinguishing the various states of the
wake/sleep cycle is based on infant behavior and findings of
polygraphic recordings rather than EEG features, although
there may be EEG changes associated with behavioral state
changes before that epoch. Eye opening is associated with the
■ Normal EEG and Sleep: Preterm and Term Neonates 159
awake state and eye closure is associated with sleep: relatively
unreliable criteria. Regular respiration, random eye move-
ments, and variable muscle tone are associated with NREM
sleep. Irregular respiration, rapid eye movements, and
decreased muscle tone are associated with REM sleep.
Wakefulness
At about 30 weeks CA, the background activity is continuous in
behavioral and polygraphic determined REM sleep and discon-
tinuous during wakefulness and behavioral and polygraphic
NREM sleep. In all states, beta–delta complexes are present with
their age-dependent characteristics of abundance, spatial distri-
bution, and degree of synchrony. By 36 to 37 weeks CA, there is
a clear distinction between the waking EEG and the sleep EEG,
without reliance on clinical or polygraphic recordings
(Fig. 9.6). In the awake EEG age-dependent features persist and
Figure 9.6 34 to 35 weeks CA. A:
Awake: The background activity is
continuous with intermittent
beta–delta complexes in the occipital
regions (same recording as B). B:
NREM sleep: Discontinuity during
NREM sleep with otherwise similar
developmental features as during
wakefulness (same recording as A).
(Reprinted with permission from
Mizrahi EM, Hrachovy RA, Kellaway
P. Atlas of Neonatal Electroencepha -
lography. 3rd ed. Philadelphia, PA:
Lippincott Williams & Wilkins;
2004.)
160 Part II ■ Normal EEG
Figure 9.7 40 to 44 weeks CA.
The EEG is characterized by
NREM sleep. The fluctuating
amplitude of the background
activity is typical of the trace
alternant pattern of quiet sleep.
the background activity consists chiefly of continuous polyfre-
quency activity. This polyfrequency activity is characterized by
random, very slow, low-voltage activity best described as
baseline shifting, with superimposed semirhythmic 4 to 8 Hz
activity in all regions. In addition there may be very low voltage
(18 to 22 Hz) activity and very low voltage (2 to 3 Hz) activity
in frontal regions. At 36 to 37 weeks, during wakefulness, the
beta–delta complexes are no longer present.
NREM Sleep
Prior to about 36 weeks CA, the background activity in behav-
ioral and polygraphic determined NREM sleep is discontinu-
ous. Between 36 and 38 weeks CA, two patterns of NREM sleep
emerge: continuous high-voltage slow wave activity in all
regions and trace alternant (Fig. 9.7). The latter is characterized
by a modulation of activity with alternating periods of high-
and low-voltage activity. This pattern may be present upto 44
weeks CA. Thereafter, NREM sleep is characterized by continu-
ous slow wave activity with the eventual emergence of sleep
spindles after about 46 weeks CA (although rudimentary spin-
dles may occur earlier).
REM Sleep
The character of the EEG during behavioral REM sleep is sim-
ilar to that of the awake recording beginning from about 30
weeks CA. Infant behaviors and polygraphic recordings of res-
piration, eye movements, and submental EMG are used to dis-
tinguish the awake state from REM sleep.
Transitional or Indeterminant Sleep
These terms are utilized to characterize the EEG when the
state of the infant cannot be precisely determined. This may
occur in the transition from wakefulness to sleep and be a rel-
atively brief period or may be more prolonged between
states.
Reactivity to Stimulation
Between 33 and 34 weeks CA, changes in EEG activity in
response to stimuli appear. By 37 weeks CA, these responses
may be easily and more reliably elicited. Typically, the response
to a stimulus is related to the character of the ongoing activity
at the time of the stimulus. If high-voltage, very slow activity is
present an effective stimulus produces abrupt and pronounced
generalized attenuation of voltage lasting upto several seconds
(Fig. 9.8). In addition, there may be spontaneous episodes of
attenuation associated with self-arousal. These may occur in
infants until about 2 weeks post-term and should not be inter-
preted as evidence of immaturity or be confused with the repet-
itive episodes of generalized or regional attenuation that occur
in neonatal encephalopathies.
SPECIAL WAVEFORMS AND PATTERNS
Some special waveforms and patterns are present, particularly
in the near-term and term infant, which are considered to be
within the range of normal variation.
Bifrontal Delta Activity
Bifrontal delta activity appears in the near-term or term infant
as intermittent semirhythmic 1.5 to 2 Hz moderately high to
high voltage activity in the frontal regions bilaterally (Fig. 9.9).
These waveforms may occur in close association with frontal
sharp transients and are most prominent during transitional
sleep. Although the pattern has been referred to as “anterior
 Chapter 9 ■ Normal EEG and Sleep: Preterm and Term Neonates 161

Figure 9.8 38 to 40 weeks. A

transient arousal is characterized
by a brief episode of generalized
voltage attenuation. The back-
ground activity is consistent with
NREM sleep before and after the
arousal. (Reprinted with permis-
sion from Mizrahi EM, Hrachovy
RA, Kellaway P. Atlas of Neonatal
Electroencephalography. 3rd ed.
Philadelphia, PA: Lippincott
Williams & Wilkins; 2004.)

dysrhythmia,” it does not suggest nor has it been associated
with any abnormality.
Temporal Sharp Waves
Temporal sharp waves are typically discussed as abnormalities
or findings of uncertain diagnostic significance (see Chapter
25). However, some temporal sharp waves may appear in oth-
erwise normal neonatal EEG recordings and are considered
findings within the range of normal variation. Temporal sharp
waves may be considered normal, which meet the following
criteria: simple diphasic morphology, with primarily surface-
negative polarity; occur randomly and, usually, asynchro-
nously on the two sides; and occur during sleep (Fig. 9.10).
Criteria for abnormality include complex morphology, posi-
tive polarity, persistent localization, and occurrence during
wakefulness.
CONCLUSION
The interpretation of the neonatal electroencephalogram is
based on the recognition of age-dependent features. For the
neonate with, ultimately a normal EEG, this provides the
basis for determination of CA and the assessment of
expected background and focal features. In addition, the
understanding of CA-dependent features will suggest the
range of both diffuse and focal abnormal features that may

Figure 9.9 38 to 40 weeks.

Bifrontal delta activity is present
(shaded oval), mixed with frontal
sharp transients (shaded square)
in this recording of transitional or
indeterminant sleep. The EEG is
continuous and synchronous on
the two sides. There are no imma-
ture features. (Reprinted with per-
mission from Mizrahi EM,
Hrachovy RA, Kellaway P. Atlas of
Neonatal Electroencephalography.
3rd ed. Philadelphia, PA:
Lippincott Williams & Wilkins;
2004.)
162 Part II ■ Normal EEG
Figure 9.10 38 to 40 weeks. A temporal sharp wave is present on the
left (shaded square). It is randomly occurring, moderate in voltage, sim-
ple in morphology occurring during transitional or indeterminant sleep.
The background activity is within the range of normal for CA.
(Reprinted with permission from Mizrahi EM, Hrachovy RA, Kellaway
P. Atlas of Neonatal Electroencephalography. 3rd ed. Philadelphia, PA:
Lippincott Williams & Wilkins; 2004.)
be present at specific epochs. An orderly approach to the
visual analysis of the neonatal EEG will provide for its great-
est clinical utility.
REFERENCES
1. Dreyfus-Brisac C. The electroencephalogram of the premature
infant. World Neurol. 1962;3:5–15.
2. Dreyfus-Brisac C. The electroencephalogram of the premature
infant and full-term newborn: normal and abnormal development
of waking and sleeping patterns. In: Kellaway P, Petersén I, eds.
Neurological and Electroencephalographic Correlative Studies in
Infancy. New York, NY: Grune & Stratton; 1964:186–206.
3. Dreyfus-Brisac C, Fischgold H, Samson-Dollfus D, et al. Veille,
sommeil, réactivité sensorielle chez le prématuré, le nouveau-né et
le nourrisson. Electroencephalogr Clin Neurophysiol. 1957;6(suppl):
417–440.
4. Dreyfus-Brisac C, Monod N. Sleeping behavior in abnormal new-
born infants. Neuropädiatrie. 1970;1:354–366.
5. Dreyfus-Brisac C, Larroche JC. Discontinuous EEGs in premature
and full-term neonates. Electro-anatomo-clinical correlations.
Electroencephalogr Clin Neurophysiol. 1972;32:575.
6. Hrachovy RA, Mizrahi EM, Kellaway P. Electroencephalography of
the newborn. In: Day DD, Pedley TA, eds. Current Practice of
Clinical Electroencephalography. 2nd ed. New York, NY: Raven
Press; 1990:201–242.
7. Watanabe K, Hayakawa F, Okumura A. Neonatal EEG: a powerful
tool in the assessment of brain damage in preterm infants. Brain
Dev. 1999;21:361–381.
8. Clancy RR, Bergqvist C, Dlugos DJ. Neonatal electroencephalogra-
phy. In: Ebersole JS, Pedley TA, eds. Current Practice of Clinical
Electroencephalography. 3rd ed. Philadelphia, PA: Lippincott
Williams & Wilkins; 2003:160–234.
9. Mizrahi EM, Hrachovy RA, Kellaway P. Atlas of Neonatal
Electroencephalography. 3rd ed. Philadelphia, PA: Lippincott
Williams & Wilkins; 2004:250.
10. Plouin P, Kaminska A, Moutard ML, et al. L’EEG en pediatrie.
United Kingdom: John Libbey; 2005:185.
11. Vecchierini MF, André M, d’Allest AM. Normal EEG of premature
infants born between 24 and 30 weeks gestational age: terminol-
ogy, definitions and maturation aspects. Neurophysiol Clin.
2007;37:311–323.
12. André M, Lamblin MD, d’Allest AM, et al. Electroencephalography
in premature and full-term infants. Developmental features and
glossary. Neurophysiol Clin. 2010;40:59–124.
13. American Clinical Neurophysiology Society. Guideline 2.
Minimum technical standards for pediatric electroencephalogra-
phy. J Clin Neurophysiol. 2006;23(3):92–96.
14. De Weerd AW, Despand PA, Plouin P. Neonatal EEG: recommen-
dations for the practice of clinical neurophysiology: guidelines for
the International Federation of Clinical Neurophysiology.
Electroencephalogr Clin Neurophysiol. 1999;52(suppl):149–157.
15. Biagioni E, Bartalena L, Boldrini U, et al. Background EEG activ-
ity in preterm infants: correlation of outcome with selected mat-
urational features. Electroencephalogr Clin Neurophysiol. 1994;91:
154–162.
16. Hahn JS, Monyer H, Tharp BR. Interburst interval measurements
in the EEGs of premature infants with normal neurological out-
comes. Electroencephalogr Clin Neurophysiol. 1989;73:410–418.
17. Selton D, Andre M, Hascoët JM. Normal EEG in very premature
infants: reference criteria. Clin Neurophysiol. 2000;111:2116–2124.
18. Monod N, Tharp B. Activité electroencéphalographique normale du
nouveau-né du pré`maturé au cours des états de veille et de som-
meil. Rev Electroencephalogr Clin Neurophysiol. 1977;7:302–315.
19. Anderson CM, Torres F, Faoro A. The EEG of the early premature.
Electroencephalogr Clin Neurophysiol. 1985;60:95–105.
20. Connell JA, Oozeer R, Dubowitz V. Continuous 4-channnel EEG
monitoring: a guide to interpretation, with normal values, in
preterm infants. Neuropediatrics. 1987;18:138–145.
21. Hayakawa M, Okumura A, Hayakawa F, et al. Background elec-
troencephalographic (EEG) activities of very premature infants
born at less than 27 weeks gestation: a study on the degree of con-
tinuity. Arch Dis Child Fetal Neonatal Ed. 2001;84:F163–F167.
22. Vecchierini MF, d’Allest AM, Verpillat P. EEG patterns in 10
extreme premature neonates with normal neurological outcome:
qualitative and quantitative data. Brain Dev. 2003;25:330–337.
23. Hrachovy RA. Development of the normal electroencephalogram.
In: Levin KH, Luders HO, eds. Comprehensive Clinical Neurophysi-
ology. Philadelphia, PA: WB Saunders; 2000:387–413.
24. Dreyfus-Brisac C, Blanc C. Électroencéphalogramme et matura-
tion cerebrale. Encephale. 1956;45:205–241.

Normal EEG and Sleep:
Infants to Adolescents
JAMES J. RIVIELLO, JR, DOUGLAS R. NORDLI, JR, AND ERNST NIEDERMEYER
GENERAL PRINCIPLES
Electroencephalogram (EEG) interpretation depends on accu-
rate pattern recognition. One of the first lessons the novice elec-
troencephalographer learns is that EEG pattern interpretation
must take into account the patient age and the level of vigilance,
or state. EEG patterns vary according to central nervous system
(CNS) development and maturation. This process evolves over
time, starting with the early development and maturation of the
nervous system (an evolution) to a peak of maturity, followed
by an involution. Basic differences exist between the ascending
(developmental) and the descending (involutional) portions of
this curve. This chapter discusses the influence of age and thus
the ontogenetic evolution or maturation of the EEG.
The declining curve of life is not determined by a reverse
process such as “dematuration” (certainly not by demyelination),
although there is degeneration of CNS structures. CNS aging is
caused predominantly by a variety of pathologic processes. The
awareness of this basic difference between the ascending and the
descending portions of life is of great significance in practical
electroencephalography. During the maturational phase, the EEG
must be interpreted knowing the conceptual age (CA) and
according to statistical norms, whereas in the aged individual, the
EEG is interpreted using the ideal or optimal norm, because the
optimal situation, even in advanced age, is absence of pathology.
Using a statistical norm, a mentally alert octogenarian with the
frequency pattern of a normal middle-aged adult would have to
be read as abnormal. This would be an absurdity.
Figure 10.1 demonstrates the maturational period, the peak
of maturity, and the decline caused by pathology. The peak
period of EEG maturation lies well into adulthood (after the age
CHAPTER
10
of 30 years), at a time when biologic aging of body tissues has
already started and physical capabilities or athletic performances
have clearly passed their peak. However, cerebral maturation is
not a smooth process of steady growth, since quantum changes
occur in prematurity, as well as in infancy and childhood (1).
HISTORICAL ASPECTS
Berger (2) performed the first EEG studies in children and dis-
covered age-dependent changes. The EEG in normal and
abnormal children has been well studied and has been
expanded with quantified EEG (QEEG). The developmental
aspect of EEG interpretation is especially important in the
interpretation and study of the neonatal EEG. The technical
performance of the EEG and its interpretation in the premature
and full-term newborn requires such specialized knowledge
that neonatal electroencephalography is now a special branch of
electroencephalography, neonatal electroencephalography.
Neonatal EEG is described in Chapter 9.
The EEG evolution from infancy to adolescence in a large
number of healthy adolescents has been extensively studied by
Swedish authors (3–7), whose work set the standards for develop-
mental electroencephalography. In a remarkable study of 1416
healthy subjects with an age range from 6 to 39 years using QEEG,
Matsuura et al. (8) investigated the development of various fre-
quency ranges over the occipital, central, and frontopolar regions.
Gasser et al. (9,10) investigated the development of the EEG in
children and adolescents by frequency analysis and topography.
EEG maturation has also been studied in animals. A study in
the dog by Pampiglione (11) and the work of Caveness (12) in
the rhesus monkey deserve special attention. Table 10.1 provides
Figure 10.1 Cerebral maturation (infancy, childhood, and
adolescence) versus pathology with regard to repercussions
on the EEG.
163
164 Part II ■ Normal EEG

Tabl e 10 . 1

A Condensed View of EEG Maturation

Premature Premature Premature Full-Term Newborn

(24–27 Wk) (28–31 Wk) (32–35 Wk) (36–41 Wk)
Continuity Discontinuous, long flat Discontinuous Continuous in waking state and Continuous except for tracé
stretches REM sleep, discontinuous in alternant in non-REM
non-REM sleep (quiet) sleep
Interhemispheric Short bursts in synchrony Mostly asynchronous Partly synchronous, especially in Minor asynchronies still present
synchrony activity leads occipital
Differentiation of waking Undifferentiated Undifferentiated Waking distinguished from sleep Good
and sleeping early in the period, then
differentiation of non-REM
and REM sleep
Posterior basic (alpha) None None None None
rhythm

Slow activity (awake) Very slow bursts, high Very slow activity Slow (delta) with occipital Slow (delta), mostly of moderate
voltage (state of vigilance predominant maximum voltage
undifferentiated)
Temporal theta Present and increasing Prominent Decreasing and disappearing Disappearing or absent
Occipital theta Prominent Decreasing Decreasing Absent
Fast activity (awake) Very little beta activity Frequent ripples or Frequent ripples or brushes Decreasing ripples, sparse fast
brushes around (16–20 per second) activity
16 per second
Low voltage Long flat stretches Flat stretches, mainly Low-voltage record suspect of Very low voltage records are due
asynchronous serious cerebral pathology to severe cerebral pathology;
prognosis ominous
Hyperventilation Not feasible Not feasible Not feasible Not feasible

Intermittent photic Unknown Unknown Unknown Driving response below 4 flashes/

stimulation second may occur, not easily
elicited
Drowsiness Undifferentiated Undifferentiated Undifferentiated Undifferentiated

Tracé alternant None None Present in non-REM (quiet) sleep Present in non-REM (quiet) sleep

Spindles None None (but ripples present) None (but ripples present) None (but scanty ripples)

Vertex waves and K None None None None

complexes

Positive occipital sharp None None None None

transients of sleep
Slow and fast activity in Slow activity of high Much slow activity, more Irregular slow activity of occipital Much delta and theta activity,
sleep voltage, little slow activity irregular; little fast predominance continuous in REM sleep
(stage of vigilance activity
undifferentiated)
REM sleep Undifferentiated Undifferentiated Continuous slow activity; Continuous slow activity, REM in
oculographically, REM present EOG (more REM or “active”
than non-REM sleep)
Rhythmical frontal theta None None None None
activity (6–7 per second)
14 and 6 per second None None None None
positive spikes
Psychomotor variant None None None None
(marginal abnormality)
Sharp waves, spikes Some intermixed sharp Some intermixed sharp Often prominent sharp Some minor sharp transients
activity in bursts (normal) activity (normal) waves or spikes (normal) (normal) (abnormal spikes
more consistent and
prominent)

Infancy (2–12 mo)
Continuous
No significant asynchrony
Good
Starting at age 3–4 mo at 4 per
second, reaching about
6 per second at 12 mo
Considerable
None
None
Very moderate
Uncommon, usually
abnormal
Not feasible
Improving driving to low
flash rates after age 6 mo
Around age 6 mo, appearance
of rhythmical theta
Disappears in first (seldom
second) month
Appear after second
month; 12–15 per
second, sharp, shifting
Appear mainly at 5 mo,
fairly large, blunt
None
Much diffuse 0.75–3 per
second activity with
posterior maximum;
moderate fast activity
REM portion decreasing;
mostly slow activity
None
None
None
Essential as abnormal
phenomena
Early childhood
(12–36 mo)
Continuous
No significant asynchrony
Good
Rising from 5–6 per second
to 8 per second (seldom
9 per second)
Considerable
None
None
Mostly moderate
Uncommon, usually
abnormal
Mostly not feasible
Often good driving response
to low flash rates
Marked “hypnagogic”
rhythmical theta
(4–6 per second)
None
In second year, sharp and
shifting, then symmetrical
with vertex maximum
Large, becoming more
pointed
Poorly defined
Marked posterior maximum
of slow activity; often a
good deal of fast activity
Mostly slow, starting to
become more
desynchronized
Seldom in third year of life
Rare
None
Spikes in seizure- free
children, mainly occipital
(mild abnormalities)
Chapter 10 ■ Normal EEG and Sleep: Infants to Adolescents
Preschool age
(3–5 yr)
Continuous
No significant asynchrony
Good
Rising from 6–8 per second
to 7–9 per second
Marked admixture of
posterior slow activity
(to alpha rhythm)
None
None
Mostly moderate
Uncommon, usually
abnormal
Often marked delta response
Often good driving response
to low flash rates
Rhythmical theta gradually
vanishing, other types of
slow activity predominant
None
Typical vertex maximum
Large with an increasingly
impressive sharp
component
Poorly defined
Predominant slowing but
less prominent posterior
maximum
Slow activity with some
desynchronization
May occur, not very common
May occur, not very common
Probably none
Spikes in seizure-free
children, mainly
occipital, also
Rolandic (slight
abnormalities)
Older children
(6–12 yr)
Continuous
No significant asynchrony
Good
Reaching 10 per second at
age 10 yr
Varying degree of posterior
slow activity mixed with
alpha
None
None
Mostly moderate
Seldom as variant
of normalcy
Often marked delta response
Often good driving response,
chiefly at medium flash
rates (8–16 per second)
Gradual alpha dropout with
increasing slow activity
None
Typical vertex maximum
Large with a prominent
sharp component
Still poorly defined but
gradually evolving
Much diffuse slowing,
slightly decreasing
voltage
Less slowing and increasing
desynchronization
A bit more common
Fairly common
Uncommon
Spikes in seizure-free
children, mainly Rolandic
(central-mid-temporal),
slight to moderate
abnormalities; physiological
occipital spikes in
congenitally blind children
165
Adolescents
(13–20 yr)
Continuous
No significant asynchrony
Good
Averaging 10 per second
Posterior slow activity
diminishing
None
None
Moderate, except for low
voltage fast records
Occasionally and (at end of
teenage period more often)
as variant of normalcy
Delta responses become less
impressive
Often good driving response,
chiefly at medium flash
rates
Gradual alpha dropout with
low-voltage stretches
(mainly slow)
None
Typical vertex maximum
Not quite as large, sharp
component not quite
as prominent
Often very well developed
Much diffuse slowing with
further attenuation of
voltage
Mature desynchronization
A bit more common,
declining at end of period
Fairly common
More common (although
relatively rare)
Benign Rolandic spikes
usually disappear before
beginning of this period
166 Part II ■ Normal EEG
a condensed presentation of a number of EEG variables and
their developmental aspects.
DISAPPEARANCE OF NEONATAL PATTERNS
Neonatal EEG patterns generally end between 46 and 48 weeks
CA and transition to predominantly infantile EEG patterns end
by 3 months of age. The trace alternant pattern is the EEG
correlate of quiet sleep (non-REM sleep) and is common dur-
ing the first 1 to 3 weeks of life in a full-term newborn. Active
sleep (REM sleep) occurs more commonly in the sleeping new-
born (64% of sleeping time, according to Passouant et al. (13)).
The healthy full-term newborn demonstrates active sleep (REM
sleep) at sleep onset.
Quiet sleep onset is well established 1 month after full-term
delivery (37 days, after Ellingson (14)) and gradually emerges
as the predominant type of sleep with the development of
stages 2, 3, and 4. Trace alternant patterns disappear about 3
to 4 weeks after full-term birth (15), although they have been
observed up to day 47 (14). Intervening CNS insults, either
primary or secondary (systemic), may interfere with CNS
function and temporarily cause developmental regression,
which may be expressed with a return to a less mature EEG
pattern.
With the disappearance of neonatal EEG patterns and the
emergence of patterns such as the posterior basic (the posterior
dominant rhythm), the forerunner of alpha rhythm, and sleep
spindles, the EEG begins to demonstrate the more mature
patterns seen in adults. During the waking state, a generally
rhythmical 3 to 4 Hz posterior activity occurs, a precursor of
the posterior alpha rhythm, or the posterior occipital or domi-
nant rhythm. This posterior rhythm demonstrates reactivity to
eye opening and closing: blocking with eye opening and activa-
tion with eye closing.
EEG IN INFANCY (2 TO 12 MONTHS)
General Considerations and Technical Aspects
It is not easy to obtain a waking EEG recording in an infant. As
infants tend to keep their eyes open, in order to obtain a poste-
rior dominant rhythm, Dreyfus and Curzi-Dascalova (15) sug-
gest gentle passive occlusion of the eyes (passive eye closure) for
short periods to activate the posterior basic rhythm. Temporary
use of a short-time constant (0.1 second) may be helpful to sep-
arate the posterior rhythm from movement artifacts. This
rhythm may also be present in a crying infant (associated with
forceful closure of eyes and concomitant frontal muscle arti-
facts), as well as in a quiet infant with open eyes. The infant usu-
ally closes the eyes as a sign of impending drowsiness.
There are two basic philosophies to EEG recording in
infancy. In the first, the recording is started in a sleeping baby.
Spontaneous sleep is desirable, and this may be achieved when
the recording is scheduled shortly after feeding. In the past,
sedation was commonly used, usually chloral hydrate. Because
of concerns for adverse events with conscious sedation, such as
apnea, the use of conscious sedation for EEG has been disap-
pearing. Many laboratories now favor sleep deprivation the
night before and allow the infant to lie on a bed or in the
mother’s arms. Sweeney et al. (16) compared sleep deprivation
versus sedation. The recordings had fewer artifacts and were off
a higher quality with sleep deprivation. Guidelines have been
established for sleep deprivation (17). The use of sleep depriva-
tion has been confirmed by Ong et al. (18). In a sedated infant,
it is often impossible to secure a waking tracing. Using the sec-
ond approach, the record routinely starts while the baby is
awake. It is extremely difficult to place electrodes under these
circumstances as most infants resent this type of manipulation.
Some EEG technologists have become extremely experi-
enced in the art of handling a waking infant, and often succeed
in obtaining both waking and spontaneous sleep records.
Further information about the management of infants and chil-
dren in North American EEG laboratories is found in the study
of Leonberg (19).
The International 10-20 Electrode System should be used for
electrode placement, using a full set of 21 EEG electrodes, 19
scalp electrodes, and two ear reference electrodes, according to
the American Clinical Neurophysiology Society guidelines (20).
The use of ocular leads, respiratory monitoring, and electrocar-
diogram (ECG) are helpful in the detection of artifacts, as well
as in the assessment of sleep stages.
Although some laboratories use rubber caps in small infant
sizes, this technique is not conducive to sleep, and collodion is
preferable, especially in the larger pediatric hospitals. Bentonite
paste (recommended by Fois (21)) is no longer available. Excellent
new types of commercial paste are now available, but when used,
it is possible that rapid drying and thus deteriorating electrode
function may occur, resulting in high electrode impedance values,
in addition to the natural higher impedance of the scalp in new-
borns and early infancy. Needle electrodes are now rarely used.
EEG Characteristics in the Waking State
In early infancy (around the age of 2 months), irregular delta
activity of 2 to 3.5 Hz and medium to high voltage (50 to 100
V) is widely preponderant. As noted above, rhythmical occip-
ital 3- to 4-Hz activity is often noted at the age of 3 to 4 months;
this activity can be blocked by eye opening.
This posterior basic (dominant) rhythm becomes more sta-
ble at the age of 5 months and increases over time, first to 5 Hz
(15) and subsequently to an average frequency of 6 to 7 Hz
(occasionally 8 Hz) by age 12 months. The amplitudes range
from 50 to 100 V.
Some rhythmical rolandic (central) activity of 5 to 8 Hz may
be present as early as the age of 3 months, with an amplitude
around 25 to 50 V. This activity is the precursor of the mu
rhythm and is stable during the first year of life (15), but may
be dependent on somatosensory stimulation (22). These visual
analysis findings have been confirmed with QEEG (23–25).
EEG Characteristics in Drowsiness
Prior to the age of 5 to 8 months, the transition from the waking
state to sleep is a gradual process characterized by progressive
slowing into the delta frequency range. No specific drowsy state
is identified in this smooth progression of slow activity to the
sleep stage. Drowsiness is recognized by a distinct pattern
 Chapter 10
between the age of 6 and 8 months: a hypersynchronous rhythm
in the lower theta range, around 4 Hz, with gradual acceleration
to 5 and 6 Hz over the ensuing months. This impressive theta
rhythmicity is known as “hypnagogic hypersynchrony” (26).
This drowsy pattern seems to develop from the posterior
basic (dominant) rhythm. The transition from wakefulness to
drowsiness is associated with a change in the amplitude distri-
bution. The maximum rhythmical theta activity moves into the
centroparietal region where amplitudes commonly reach 100 to
250 V. EEG amplitude values depend on the interelectrode
distance and may vary from montage to montage and are best
measured with a reference montage.
According to Dreyfus and Curzi-Dascalova (15), the occipi-
tal rhythm may be somewhat slower than the diffusely predom-
inant theta rhythm (possibly indicating a basic difference
between two coexisting rhythmical theta patterns). In rare
cases, the hypnogogic hypersynchrony may not occur.
Sleep EEG and Non-REM Sleep
During the first 3 months of life, sleep may begin in a peculiar
manner. Curzi-Dascalova et al. (27) have shown that some
infants fall asleep without eye closure, others with half-closed
eyes, and others with brief alternating opening and closing of
the eyes. The EEG and polygraphic data indicate sleep onset
with active (REM) sleep in neonates and gradual evolution of
sleep onset with quiet (non-REM) sleep during the ensuing
weeks. The “slow” sleep of the infant is dominated by diffuse
0.75- to 3-Hz activity with a maximum amplitude (100 to 150
or 200 V) over the occipital area, “occipital delta.” This occip-
ital delta activity may be quite prominent during the first year
of life. The amplitudes increase with deepening slow sleep.
There are some intermixed theta, alpha, and beta frequencies of
smaller amplitudes.
Sleep spindles usually appear during the second month of
life; occasionally, spindle fragments may be seen somewhat ear-
lier (28). The spindle frequency ranges from 12 to 15 Hz, with
14 Hz the most commonly encountered frequency. Throughout
infancy, spindles are maximal over central and parietal areas
with shifting asymmetries. A clear-cut midline (vertex) maxi-
mum does not exist at this age. Spindles of infancy usually show
a negative sharp component, whereas the positive component is
rounded. The sharp spindle configuration (shown by Fois (21))
is a typical hallmark of sleep in infancy. The comb-like shape of
these runs may be erroneously interpreted as 14-Hz positive
spikes, but a careful analysis of polarity clearly shows the nega-
tivity of the spiky components. The spatial distribution also is
different: spindles are rolandic, whereas 14-Hz positive spikes
are predominant in a posterotemporal location. Finally, the
14-Hz or 14- and 6-Hz positive spike pattern is extremely rare
before the age of 2 years and virtually nonexistent during the
first year of life (Fig. 10.2).
Spindles may show sharp negative and positive phases after
the age of 6 months Katsurada (29). This sharp positive spindle
polarity is less common than spindles with a strictly negative
sharp component.
The spindle train duration varies with age. Spindle trains
are of short duration and low voltage during the second
■ Normal EEG and Sleep: Infants to Adolescents 167
month. With the third month, the amplitude increases and the
duration of the runs becomes much longer (30). The intervals
between spindle trains become shorter, but we do not agree
that spindling is almost continuous between the age of 2 and
6 months (24). Spindle bursts may reach a duration of 10 sec-
onds in the second half of the first year, but there is a decreased
duration of each spindle run, while the overall number of runs
increases.
Dreyfus and Curzi-Dascalova (15) report that the complete
absence of spindles at the age of 3 to 8 months represents a
severe abnormality. However, this finding should be interpreted
with great caution, as enough sleep must have been obtained
during the recording in order “to give the baby a chance to pro-
duce spindles.” At this age, there is still a fair chance that the
sleep recording is limited to REM sleep; this demonstrates the
importance of oculographic and pneumographic recording.
An excellent demonstration of the development of sleep
spindles from the age of 10 weeks to 1 year was presented by
Hughes (31) (Fig. 10.3).
Vertex waves and K complexes are usually seen around the age
of 5 months, although rudiments may occur much earlier.
Vertex waves may be quite large at the age of 5 to 6 months. At
this age, K complexes are of considerable voltage, but the initi-
ating sharp component is poorly developed and somewhat
“blunted” (32). In contrast, Metcalf et al. (33) have stressed the
comparatively low voltage of K complexes in infancy. These
authors also noticed the appearance of K complexes in infants
aged 5 to 6 months, but the complexes may be obscured by
background activity (Figs. 10.4 and 10.5). A vertex wave may be
frontally dominant or extend into the lateral frontal regions;
this is called an F-wave (34) (Fig. 10.6).
In normal infants in the first year of life, Ellingson et al. (35)
recorded brief apneic episodes (“respiratory pauses” lasting 3 to
10 seconds) during sleep. These pauses are more common in
REM sleep.
Sleep EEG and REM Sleep
REM sleep abundance decreases during the first year of life
(36,37), evolving from approximately 50% at birth, falling to
40% at 3 to 5 months and 30% between 12 and 24 months.
Dittrichova et al. (38) and Dreyfus and Curzi-Dascalova (15)
reported the occurrence of sharply contoured occipital activity
in the REM sleep of infants. This activity shows a frequency
around 2 Hz at 6 weeks and 2 to 4 Hz at 12 to 16 weeks of age.
There is some intermixed delta and theta activity associated
with the occipital sharp transients.
The REM sleep latency (time span from sleep onset to the
first REM period) gradually lengthens during the first year of
life. Schulz et al. (39) has shown that the REM sleep latency
underlies diurnal rhythmical changes, with the longest latencies
between noon and 4 PM and the shortest between 4 AM and 8 AM.
Reactivity and Evoked Responses
The evolution of evoked responses in newborns and infants are
discussed in Chapter 49.
The blocking response of the posterior basic (dominant)
rhythm (obtained by passive eye closure, see General
168 Part II ■ Normal EEG

Figure 10.2 A: Patient age 10 months. Light non-REM sleep. Typical spindles of infancy,
sharp and shifting. Normal posterior voltage maximum of slow activity (channels 3 to 7
from the top). B: Patient age 9 months. More prolonged trains of infantile spindles.

Considerations and Technical Aspects) is usually present at the
age of 3 to 4 months. The central mu rhythm does not react to
eye opening, which permits differentiation when there is ante-
rior spreading of the posterior dominant rhythm. A true
rolandic mu rhythm cannot be identified during the first year
of life, although forerunners of this activity are likely to be pres-
ent. The authors’ earliest observation of unmistakable rolandic
mu rhythm was made in a 20-month-old child (40).
A photic driving response to flickering light may be obtained
as early as 3 to 4 months after delivery; the responses are most
prominent in the theta band (41,42). Occipital lambda wave
activity occurs rarely during the first year of life.
EEG IN EARLY CHILDHOOD
(12 TO 36 MONTHS)
Waking Record
Readable waking records are not easily obtained in the second
and third years of life. At this age, many children resent being laid
down in a supine position for electrode placement and recording
 Chapter 10
Figure 10.3 Sleep spindles at different ages. At 10 weeks note the very
low amplitude at the beginning and end of the burst (underlined) with
a duration of 5 seconds; at 13 weeks a long lasting spindle of 6.5 sec-
onds; at 17 weeks, 3.5 seconds; at 39 weeks, 2 seconds; and 52 weeks,
1.5 seconds. Time (1 second) and voltage calibration (50 V) are
shown. Montage is bipolar linkage of the C3-P3 electrodes. Arrows des-
ignate the beginning and ending of typical spindles at the given age.
(From Hughes JR. Development of sleep spindles in the first year of life.
Clin Electroencephalogr. 1996;27:107–115.)
and may be too heavy to be comfortably held. These children may
successfully fight sleepiness and sleep, and if sedated, the waking
portion is a postarousal tracing, in which it must be determined
that the child is truly awake. Some children may pull off their
electrodes almost immediately after awakening.
The posterior basic (dominant) rhythm increases from the
upper theta to the lowest alpha range and is most commonly
found between 6 and 7 Hz during the second and 7 and 8 Hz
Figure 10.4 Patient age 9 months. Light non-REM sleep. A large vertex
wave after auditory stimulus. The response is somewhat complex and
resembles a K complex.
■ Normal EEG and Sleep: Infants to Adolescents 169
Figure 10.5 A K complex in a patient age 13 months; quite large with
a blunted sharp component.
during the third year of life. The blocking response to eye open-
ing intensifies, although the degree of this reactivity, a more
impressive response, is considerably determined by darkness or
brightness of the recording room.
The posterior basic (dominant) rhythm frequency is subject
to considerable variation in the second and third year of life.
According to Lindsley (43), a frequency range from 5 to almost
10 Hz can be observed in the second year, whereas Eeg-
Olofsson (3) reported a surprisingly high mean frequency of
8 Hz at this age.
Eye blinks are associated with a biphasic (negative–positive)
potential of medium to high voltage in bilateral synchrony over
occipital areas (44).
The amount of fast activity varies considerably. Unusually
pronounced activity in the 18 to 25 Hz range in unmedicated
children may reach abnormal degrees in the waking state as well
as in sleep; such records suggest a nonspecific abnormality and
may be seen in the study of Gibbs and Gibbs (45).
Frequencies slower than the posterior basic (dominant) rhythm
are practically always noted in the waking state at this age,
mostly in the 2 to 5 Hz range and widely scattered. The amount
of this activity is also quite variable and may be accentuated by
crying due to a hyperventilation effect.
The reactivity of the waking EEG becomes more prominent;
the blocking response of the posterior basic (dominant) rhythm
170 Part II ■ Normal EEG

Figure 10.6 The F wave: a more

frontal (anterior) vertex wave.

is more clearly demonstrable than in the first year. A mu Sleep Record

rhythm over the rolandic region may be identifiable in the Sleep begins with non-REM sleep. Because of the usual rela-
second year of life (40). The photic driving response to the tively short duration of a sleep recording in the EEG laboratory
strobe light remains accentuated in the low-frequency range. in daytime, REM stages are usually not observed. The hypna-
gogic hypersynchrony of drowsiness is replaced by diffusely
Drowsiness
preponderant, irregular, high-voltage (1 to 3 Hz and medium-
Generalized high-voltage (mostly 4 to 6 Hz) theta activity is the to high-voltage 4 to 6 Hz) activity. The maximum amplitude is
hallmark of the drowsy state in early childhood. This drowsy almost invariably found over the occipital area, where 0.75- to
pattern is called hypnogogic hypersynchrony. It is most pro- 2-Hz waves may reach very high voltage (Fig. 10.2). A different
nounced in the central and parietal leads, is less impressive over type of voltage distribution or even an anterior voltage maxi-
the temporal areas, and sometimes least developed over the mum should warn the electroencephalographer; residual dam-
occipital region. The rhythmicity may assume quite monoto- age to the posterior regions is a good possibility, although none
nous character; the in-phase character of the ubiquitous pattern of the typical EEG abnormalities may be present (Fig. 10.8).
makes it tempting to theorize about an enormously extended This anterior to posterior amplitude difference is referred to as
generator area (Fig. 10.7). the frequency amplitude gradient (48).
Gibbs and Gibbs (46) paid special attention to a variant of The varying degree of intermixed fast activity without pre-
the rhythmical theta activity of drowsiness. In a large number medication was mentioned earlier. Spindle activity in the 12- to
of small children, bursts of 4- to 5-Hz activity may be present
instead of the steady rhythmical theta activity. This pattern
extends into the fourth year of life and shows mild paroxysmal
activity without representing a true abnormality. Occasionally,
small sharp or spiky discharges may be interspersed between
the theta waves; the electroencephalographer must refrain from
calling these discharges epileptic or spike–wave complexes.
Brandt et al. (47) determined that these admixed sharp or spiky
components occur in a small percentage (approximately 5 to
10%) of normal children at all ages. This was confirmed by Eeg-
Olofsson et al. (5). Unless unquestionable spikes are detectable,
the presence of such small spiky potentials is harmless, does not
represent a definite abnormality, and should be interpreted as a
normal finding. The mature form of drowsy EEG activity with
dropout of the posterior basic (dominant) rhythm and general Figure 10.7 Patient age 3 years. Drowsiness. Rhythmical hypnagogic
decline of the voltage output is quite common at this age. theta activity, 4 to 5 per second.
 Chapter 10 ■ Normal EEG and Sleep: Infants to Adolescents 171

Figure 10.8 Patient age 9 months. Cerebral palsy, diplegic. Lack of

posterior voltage maximum. Otherwise a normal record with sharp
spindles of infancy.

14-Hz range is now found in a transitional stage; the shifting

spindle runs occurring over centroparietal areas with a nega-
tive spiky component are gradually replaced by symmetrical
spindle activity with a maximum over vertex. A 14-Hz spindle
type appears first over the vertex, and with deepening sleep
(transition stage 2 to 3), spindles in the 12-Hz range appear
over frontal midline and upper frontal regions. Both types of
spindles may occur concomitantly, but have a different spatial
distribution.
Vertex waves, also known as “biparietal humps” (46), are
quite prominent at this age; their amplitude is impressive, but
their rise is not as abrupt and the configuration is not as sharp
as in later years (the age of 4 to 12 years). This same is true for
K complexes, which are abundantly noted in the stages of non-
REM sleep.
Positive occipital sharp transients of sleep (occipital positive
sharp activity, lambdoid activity) are absent or very poorly
developed at this age.
REM sleep at this age rarely occurs during a standard EEG,
but is seen during a standard sleep study; it starts to show signs
of EEG desynchronization, but slow activity (mostly 2 to 5 Hz,
according to Cadilhac (49)) is still preponderant (Fig. 10.9).
Figure 10.9 Patient age 22 months. REM sleep.
The fast spiky spindle variant (51) is a very rare pattern
(about 1 case in 2000 to 3000 EEGs of children). It consists of
spindle-like activity in the range of 16 to 20 Hz and may occur
over a variety of areas (central, parietal, vertex, or mid-
temporal) in stage 2 or 3 of sleep. It is seen mainly in early child-
hood (around the age of 3 years) but also occurs in older chil-
dren and adolescents. This fast spiky spindle variant is usually
seen in children with static or progressive brain dysfunction
and probably indicates a mild abnormality.
The 14- and 6-Hz positive spike discharge is very uncommon
before the age of 3 years and will be discussed in Chapter 26. It
is now called the 14- and 6-Hz positive burst.

Unusual and Variant Patterns in Sleep

Extreme spindles (45,50) represent an unusual variant of sleep
spindle activity with high voltage, a wide frequency range (6 to
18 Hz), and occasional paroxysmal traits. In a diminutive form,
these spindles may even occur while the child is awake (coexist-
ing with a posterior alpha rhythm) (Fig. 10.10).
According to Gibbs and Gibbs (45), this pattern occurs in
0.05% of normal children, but the prevalence reaches 5% to 18%
in children with mental retardation and/or cerebral palsy. This pat-
tern is seen from the age of 1 to 12 years, with a peak at the age of
3 years; after the age of 12 years, the incidence falls to almost zero.
Following the original studies of Gibbs and Gibbs, there have
been very few reports dealing with this pattern. When clearly
separable from typical spindle activity, extreme spindles might Figure 10.10 Patient age 22 months. Extreme spindles over central
represent a mild abnormality. region; a somewhat slower type of spindles over the frontal area.
172 Part II ■ Normal EEG
Arousal from Sleep
Arousal from sleep is characterized by a similar marked and
prolonged high-voltage (4 to 6 Hz) activity in all leads with
some intermixed slower frequencies (52,53). This is called
hypnopompic hypersynchrony. White and Tharp (54) reported
the observation of prolonged rhythmical sharp or spiky activity
over the frontal areas in children between the age of 2 and 12
years. This pattern was seen only in children with minimal cere-
bral dysfunction. In the authors’ experience, this pattern has a
very definite maximum over the frontal midline and is seen
mainly from the second to the fourth year of life. Its mildly
abnormal character is likely, but it is now considered a normal
variant (Fig. 10.11).
However, in a study of 50 children with “frontal arousal
rhythm,” it is rare, occurring in 0.22% of children, and epileptic
seizures were present in 35 (70%) of the cases (55). Hughes (56)
also reported a case in which this frontal arousal rhythm was
associated with a clinical seizure consisting of eyelid flutter and
chewing.
Scarcity of Minor Abnormalities during
the First 2 Years
When abnormalities are carefully and consistently graded as
minimal, slight, moderate, marked, and severe, there is a sta-
tistically significant (p .001) small number of minor (i.e.,
minimal, slight, and moderate) abnormalities prior to the age
of 21 months. In other words, there is no real continuum of
various forms of minor abnormalities between normal
records on one side and markedly abnormal records on
the other side (57). After the age of 21 months (~2 years),
minor abnormalities become much less frequent and the
aforementioned continuum of graded abnormalities is pres-
ent. This data suggests that perinatal brain damage shows
accompanying EEG abnormalities in the neonatal period, fol-
lowed by a silent period with normal EEG findings yet ongo-
ing clinical problems until the end of the second year of life,
when EEG abnormalities tend to reappear.
Figure 10.11 Patient age 16 months. Arousal from sleep. Pronounced
rhythmical 6 per sec activity with a sharp component, maximal over
frontal midline.
We hypothesize that there is a limited number of EEG
responses to cerebral impairment prior to the age of 21 months,
with a lack of fine intermediate nuances of abnormality. It is
possible that minor abnormalities are not so rare after all and
that new valid criteria will evolve for the interpretation of the
EEG infancy and early childhood. Recent progress in neonatal
EEG has shown that classical signs of abnormality such as
spikes, focal or diffuse slowing, or local voltage depression can-
not serve as the only guideline; the sequences of sleep stages and
their adequacy for the gestational age are important criteria for
the newborn. In the older child and adult, an example of a new
criteria for normality is the frequency amplitude gradient:
lower amplitude, faster frequency waveforms in the anterior
regions with higher amplitude, slower frequency waveforms in
the posterior regions, reported in sleep by Slater and Torres (48)
(also see Fig. 10.8). However, this gradient is also present in the
waking state.
PRESCHOOL-AGE EEG (3 TO 5 YEARS)
Waking Record
A waking record is much easier to obtain, and with good han-
dling by the technologist, many children are quite cooperative.
At the age of 3 years, the posterior basic (dominant) rhythm has
reached 8 Hz and thus the alpha range. The alpha amplitude is
almost always higher than in adolescence and adulthood.
Leissner et al. (58) correlated the alpha amplitude with skull
thickness by ultrasound. The alpha amplitude may reach 100
V and tends to increase from the age of 3 to a peak at 8 or 9
years. Amplitudes are usually higher over the nondominant
hemisphere (6).
The train of posterior alpha waves is frequently interrupted
by admixed slow waves, mostly in the range of 1.5 to 4 Hz,
extending from occipital into the posterior temporal and, less
impressively, into the parietal regions. These admixed slow
waves, fused with overriding alpha, are called posterior slow
waves of youth. Alpha voltage, distribution, and admixture of
slow activity create a picture that makes the tracing almost
immediately identifiable as a record of childhood.
Posterior slowing may show a variety of forms. Most com-
mon is the irregularly interspersed type of slow activity.
These slow waves are often preceded by a sharply contoured
potential that blends together with the ensuing slow wave
“slow fused transients” (1). The slowing is more often later-
alized to the left hemisphere, according to Petersén and
Eeg-Olofsson (6); in the author’s experience, however, later-
alization to the right seems to be more common. Posterior
slow activity can reach abnormal degrees, especially with
towering amplitudes or abnormal rhythmicity such as pro-
longed trains of rhythmical slow waves. The differentiation of
such abnormal forms from physiologic posterior slow activ-
ity is usually not difficult. Posterior slow waves of youth react
to eye opening and closing, as does the posterior dominant,
or alpha, rhythm.
In the absence of posterior slow activity, the posterior alpha
rhythm assumes a more “mature” character with unbroken
prolonged stretches. Eeg-Olofsson (4) and Petersén and Eeg-
 Chapter 10
Olofsson (6) list such tracings as “supernormal,” which is a
term comprehensible to every electroencephalographer. Such
records suggest a hastened process of maturation, and a
5-year-old child may show the adult features of posterior alpha.
Anterior rhythmical 6- to7-Hz theta activity is not uncom-
monly found at preschool age but reaches its peak somewhat
later. The clinical significance of this pattern is discussed in the
next section.
Rolandic mu rhythm is clearly on the rise at this age before
peaking early in the second decade of life.
At this age, low-voltage records with activity consistently
below 25 to 30 V are not seen as a variant of normalcy. Such
tracings may be associated with states of cerebral palsy (45).
Excessively fast records usually represent drug effect; otherwise,
such records fall into the mildly to moderately abnormal range.
Activations
At the age of 4 years, children usually become very cooperative
for hyperventilation, although one seldom finds a cooperative 3-
year-old child. Four-year-olds seem to enjoy the test and are
sometimes hard to stop. The use of a pinwheel is very helpful.
High-amplitude slowing to hyperventilation can be very
pronounced at this age and must be considered normal unless
unequivocal epileptic discharges or marked asymmetries are
found. Protracted slowing (either as rhythmical delta bursts or
as continuous irregular delta–theta activity) after termination
of hyperventilation usually indicates that the child continues
overbreathing. In such cases, the effect of hypoglycemia must
also be taken into consideration. It may be important to
document the time of the last meal. Special studies in this field
were done by Gibbs et al. (59) and by Petersén and Eeg-
Olofsson (6). A prolonged response to hyperventilation may
indicate vascular disease, as has been specifically reported in
Moyamoya syndrome (60).
The occipital driving response to intermittent photic stimula-
tion is quite often best obtained with a flash frequency in the
lower flicker frequency range; responses are most prominent
below 8 Hz (6). Paroxysmal responses to flicker (even in the
normal child) are more likely to occur in school-age children.
EEG Findings in Drowsiness
Past 3 years of age, the previously mentioned hypnagogic
hypersynchrony disappears; it is seldom seen in healthy chil-
dren at the age of 6 years. Paroxysmal theta bursts are also on
the decline. The admixture of posterior slow activity tends to
increase in early drowsiness, and increased delta and theta
activity soon occurs in all leads.
Anterior rhythmical 6- to 7-Hz theta activity increases in early
drowsiness but disappears in deep drowsiness. This pattern occurs
more often at school age and is discussed in Waking Record.
The 14- and 6-Hz positive burst manifests itself in deep
drowsiness and vanishes in light sleep. This pattern is not very
common at this age and is discussed in Chapter 26.
Sleep EEG Records
The posterior maximum of diffusely predominant delta (1 to
3 Hz) activity is not as pronounced as in infancy and early
■ Normal EEG and Sleep: Infants to Adolescents 173
childhood. Sleep spindles have lost their sharp and bilaterally
shifting feature of infancy and now show a well-defined
maximum over the vertex. Frequencies around 14 Hz are seen
before more anterior spindles in the 10- to 12-Hz range occur
with deepening sleep.
Vertex waves and K complexes show a more prominent sharp
component than in earlier years. Positive occipital sharp transients
of sleep (lambdoid activity) are absent or very poorly delineated
at this age. REM sleep is rarely recorded under regular laboratory
conditions. The EEG shows little desynchronization at this age.
EEG IN OLDER CHILDREN (6 TO 12 YEARS)
General Considerations
Normal children are fairly cooperative. Complete records with
a resting awake portion, hyperventilation, intermittent photic
stimulation, and sleep are easily obtained.
This is a well-explored age range. The most authoritative
study in this field was carried out in Gšteborg, Sweden, by
Eeg-Olofsson (3) and Petersén and Eeg-Olofsson (6) (also
see (7). To secure a population of truly healthy children, a
remarkable set of strict criteria was used, excluding those
with even minute signs of neurologic, autonomic, and psy-
chological dysfunction.
Waking Record
The posterior alpha rhythm gradually reaches a mean frequency
of around 10 Hz, which equals the mean frequency of the
mature adult EEG. This frequency is reached around the age of
10 years. In the work of Petersén and Eeg-Olofsson (6), the
mean value of 9 Hz was reached at the age of 7 and 10 Hz at
15 years. Interestingly, girls showed a statistically significant
faster acceleration of the posterior alpha frequency. These
authors reported an increase of the alpha amplitude during
early childhood until a peak was reached at the age of 6 to
9 years, with subsequent gradual decline. There was no gender
difference. The alpha amplitude was determined by a bipolar
occipitotemporal array, and a mean of 56 V was found; such
precise values are completely dependent on electrode selection
and interelectrode distance.
The posterior alpha rhythm is usually of higher amplitude
over the nondominant hemisphere and thus larger on the right
side. According to Petersén and Eeg-Olofsson (6), these asym-
metries seldom exceed 20 V. These authors were also unable to
correlate alpha asymmetries with handedness.
Although the mature alpha frequency of 10 Hz may be
reached at the age of 10 or at least at age 15 years, there is still a
considerable admixture of posterior slow activity interspersed
between trains of alpha waves. The maturational process of the
second decade of life is thus characterized by the gradual disap-
pearance of the admixed posterior slow activity, excluding a few
other minor changes in EEG evolution (Fig. 10.12A).
Posterior slow activity is still quite prominent between the
age of 6 and 12 years; “slow fused transients” (61) with sharp
transients preceding single large slow waves are quite common.
In addition to the common type of posterior slowing as randomly
174 Part II ■ Normal EEG
Figure 10.12 A: “Supernormal” posterior alpha rhythm in a patient aged 9 years. The only intermingled slow waves are
indicated by an arrow. B: Patient age 8 years. Considerable intermixed posterior slowing, mostly 2.5 to 3.5 per second.
intermixed slow potentials, there are also other and more rhyth-
mical forms of posterior slow activity.
Waveforms of posterior slow activity are broken down into
arrhythmical and rhythmical slowing. This division also applies
in adolescence and adulthood, although certain forms of poste-
rior slowing are more common in children and others in adults.
This subject has been extensively studied and discussed by Aird
and Gastaut (62), Petersén et al. (7), and Kuhlo (63).
In the age range from 6 to 12 years, posterior slow waves of
youth, or the arrhythmical type of “slow, posterior waves found
predominantly in youth” (62), are most commonly encountered.
This pattern consists of waves in the delta and theta range, of vari-
able form, lasting 0.35 to 0.5 second or longer without any consis-
tent periodicity. Alpha waves are almost always intermingled or
superimposed. Posterior slow waves of youth react to eye opening
with blocking, similar to the posterior dominant rhythm.
Among the rhythmical forms, the slow alpha variant repre-
sents a subharmonic of the alpha frequency (mostly 4 to 5 Hz),
often with intermixed alpha waves. Unilateral alpha subharmon-
ics (left side, 4.5 Hz) were noted (64) in a neurologically normal
10-year-old girl. Forms of pronounced rhythmical high-voltage
(3- to 4-Hz) activity over occipital areas and adjacent regions are
not uncommon at this age; these waves are often sharply con-
toured, and the moderately paroxysmal character of these slow
runs places this activity into the mildly abnormal range.
According to the extensive work of Kuhlo et al. (65), the 4- to
5-Hz rhythm, which is a special entity of posterior slowing,
is not demonstrable during the first decade. Very rhythmical
high-voltage 3- to 4-Hz waves may occur in prolonged runs in
children with absence epilepsy. This is referred to as occipital
intermittent rhythmic delta activity (OIRDA) (66). This pattern
is clearly abnormal and is discussed in Chapter 26.
Figure 10.12B shows slow waves often preceded by a sharp-
contoured large alpha wave (fused forms). Record is within nor-
mal limits of variability for age.
Fast activity usually does not play a major role at this age.
Rolandic mu rhythm steadily increases until a peak is
reached between the age of 11 and 15 years (40). Occipital
lambda waves are demonstrated more easily in older than in
younger children but still lack the crisp and unmistakable
contour of adulthood.
The prevalence of anterior rhythmical 6- to 7-Hz theta activ-
ity increases between the age of 6 and 12 years before reaching
a peak at the age of 13 to 15 years (67). These authors consider
this pattern not an abnormality as such, but “of clinical interest
because of its epileptologic implications” (Fig. 10.13).
This view differs from the opinion of Petersén et al. (7), who
found a relatively high prevalence of anterior theta rhythm in
healthy young individuals (21%). This figure is much larger than
that of Palmer et al. (67), who found up to 4%. Interestingly,
Petersén et al. (7) found a gender difference in the occurrence of
this rhythm, with 24.9% in boys and 16.5% in girls.
 Chapter 10 ■ Normal EEG and Sleep: Infants to Adolescents 175

Drowsiness
Drowsiness is characterized, at this age, by increasing theta and
delta frequencies along with gradually fading posterior rhyth-
mical alpha activity. The mature type of onset of drowsiness
with gradual alpha dropout and mixed low-voltage slow and
fast activity usually does not occur in the first decade and slowly
makes its appearance with early adolescence. Alternatively, hyp-
nagogic hypersynchrony seen in early childhood usually disap-
pears around the age of 6 years.

Sleep
Under regular EEG laboratory conditions (leaving aside noc-
turnal or all-day sleep studies), non-REM sleep in its various
stages is almost exclusively found unless the recording is pro-
longed to 1 hour or more.
Prior to the appearance of the first spindle trains, vertex waves
are noted (transition from stage 1 to 2). These vertex waves are of
remarkably high voltage and may have a very sharply contoured
morphology; at times, their sharp character may be poignant,
bordering on spikes. These potentials are also known as “bipari-
etal humps” because, in the montages of Gibbs and Gibbs (46),
the maximum was over the parietal region, because neither vertex
Figure 10.13 Rhythmical 6 to 7 per second activity over frontal
nor strictly central-rolandic electrodes were used (Fig. 10.14).
midline and vicinity in a 12-year-old child.
In school-age children, vertex waves may show moderate (sel-
dom considerable) amplitude asymmetries with asymmetrical
A 4- to 6-Hz theta rhythm in children with predisposition to
primary generalized epilepsy (spike–wave absences, etc.) has
been described by Doose et al. (68–70). This activity was seen
mainly below the age of 8 years and was common in seizure-free
siblings of children with petit mal absences. The predominantly
centroparietal theta activity is considered abnormal (after care-
ful differentiation from physiologic activity and drowsiness).
Low-voltage records as a variant of normalcy are very rare at
this age. Petersén and Eeg-Olofsson (6) did not find a single
case in their series. Schauseil-Zipf et al. (71) reassessed the EEG
at the age of 7 to 12 years in children who were regarded as
“small-for-date” babies. Out of a total of 56 children, the EEG
was normal in 36 and slightly to moderately abnormal in the
remaining cases.

Activations
Hyperventilation shows particularly pronounced high-ampli-
tude slowing at this age. The slowing may start over posterior
areas and gradually becomes diffusely distributed with a frontal
maximum; rhythmical slow activity usually ranges from 1.5 to
4 Hz. A variety of responses has been reported by Daute et al.
(72) (with concomitant evaluation of pH and pCO2 changes)
and by Petersén and Eeg-Olofsson (6). The normal character of
the rhythmical and arrhythmical slow responses cannot be
overemphasized (see Activations [Adolescents], below, for more
details).
Intermittent photic stimulation shows a more mature type of
occipital driving response, less prominent at low flash rates, and
more impressive in the medium range (6 to 16 Hz). The work
of Herrlin (73), Doose et al. (74), and Petersén and Eeg-
Olofsson (6) has been essential in the investigation of these Figure 10.14 Patient age 10 years, non-REM sleep stage 2. Well-
responses. developed spindles, vertex waves, and K complexes.
176 Part II ■ Normal EEG

spread into the vicinity. Under these circumstances, the physio-

logic vertex activity may be confused with cerebral (rolandic)
spikes; with greater experience, however, such errors can be
avoided. The distinction between vertex waves and rolandic
spikes, usually picked up by electrodes C3 or C4 over central
areas, may occasionally arise from the vertex itself. When such a
coexistence of vertex potentials exists, the physiologic vertex
waves are usually of longer duration and of somewhat higher
voltage than the abnormal spikes.
Prior to onset of stage 2 (light sleep), positive occipital sharp
transients of sleep may already be noticeable. This pattern, how-
ever, is less common and less prominent in children than in
adolescents and adults. A more detailed discussion is presented
in the section dealing with the sleep EEG in general.
Sleep spindles show the features of maturity, with a well-
defined vertex maximum. Its initial frequency of 14 Hz or 12
to 14 Hz may decrease to 10 to 12 Hz with deepening sleep;
the superior frontal electrodes (F3, F4) may show the spindles
somewhat better than the central electrodes (C3, C4). The spin-
dle activity maximum remains over the midline; a vertex
maximum may or may not be replaced by a maximum over
frontal midline (FZ) with deepening sleep (transition from
stage 2 to 3). The spindle trains are usually shorter than 1 second. Figure 10.15 Evolutional and involutional phases of the K complex.
The waves are rounded; the physiologic negative spiky spindle Left tracing, patient age 5 years. Prominent initiating sharp component.
character has disappeared much earlier (at the age of 2 years) Upper right tracing, patient age 8 years. Note very pronounced sharp
(Fig. 10.14). component of K complex. Far right, a single vertex wave is shown.
K complexes are often seen in association with spindles. The Lower right, patient age 47 years. K complex of moderate voltage; the
complex wave morphology of this pattern is extensively discussed initiating sharp component is markedly blunted. The slow component
in conjunction with sleep in general and again in Chapter 26. is inconspicuous; superimposed spindles are barely visible. This decline
K complexes and vertex waves represent fascinating responses to in the K complex suggests some cerebral pathology.
intrinsic or extrinsic stimuli discussion (Fig. 10.15).
REM sleep (rarely seen in a regular EEG laboratory) shows
less slow activity and increasing desynchronization with mixed dominantly seen in adolescents and young adults, but it is not
activities in the theta, alpha, and beta frequency range. uncommon in older children. The wave morphology of this dis-
Arousal from sleep is characterized by an increasingly shorter charge is presented in other sections. The psychomotor variant
transition from sleep to the waking state and decreasing length occurs almost exclusively in very light drowsiness and disap-
of high-voltage theta activity. pears with deepening level of awareness. This pattern has also
been called rhythmic mid-temporal discharges (80). This pat-
Variant Patterns of Uncertain Clinical tern does not have the evolution of frequencies and waveforms
Significance in Drowsiness and Sleep seen with an actual electrographic seizure discharge (76).
There are EEG waveforms that are now considered normal vari-
ants. These waveforms typically occur in a rhythmic manner,
with an admixed “epileptiform” morphology. These may repre- EEG IN ADOLESCENTS (13 TO 20 YEARS)
sent a superimposition of background frequencies (75). These
patterns are now referred to as normal variants, benign vari- General Considerations
ants, or patterns of uncertain clinical significance (76,77). Although the transition from childhood to puberty is a major
The 14- and 6-Hz positive burst is a common finding at this step in biologic development, EEG maturation shows no dra-
age. The sections on the sleep EEG and on abnormal (epileptic) matic changes in these years. In other words, there are no
patterns in the EEG discuss more extensively this type of striking differences between the record of a healthy teenager
activity, which was widely considered a genuine abnormality and a healthy older child. Nevertheless, there are a number of
after the first report of Gibbs and Gibbs (78). Subsequently, subtle changes.
accumulating evidence demonstrates that this pattern is not Eeg-Olofsson (4) carried out a special study of the EEG
categorically abnormal and is considered a normal variant development of normal adolescents and selected 185 persons
unless it appears in enormous abundance. from the age of 16 to 21 years, using the same strict criteria that
The rhythmic temporal theta bursts of drowsiness, previ- were applied to the younger children. Earlier studies mostly
ously known as the psychomotor variant pattern (45,79), is pre- included mixed ages, either adolescents or children.
 Chapter 10
Waking Record
The posterior alpha rhythm shows a mean frequency of around 10
Hz; according to Petersén and Eeg-Olofsson (6), the mean
frequency lies at 10.2 Hz in the male and 10.3 Hz in the female.
The entire alpha range, as reported by Petersén and Eeg-
Olofsson, stretches from 8 to 12 Hz; these authors observed a case
of 8 Hz alpha rhythm in a healthy 17-year-old boy (this would
raise the suspicion of a minimal abnormality, in the opinion of
some electroencephalographers). With the use of frequency
analysis, Samson-Dollfus and Goldberg (81) found mean alpha
frequencies from 9.64 to 9.94 Hz in subjects at the age of 15 years.
The posterior alpha rhythm is of moderately higher ampli-
tude than in adulthood, but a slight decline from the height of
alpha waves in childhood is noticeable. Alpha waves may show
a negative sharp component (classically present in mu rhythm)
that is due to a hidden or overt admixture of fast activity. Such
sharp alpha waves are common in older children, adolescents,
and young adults; this pattern is perfectly normal, a fact already
stressed by Gibbs and Gibbs (46). Amplitude asymmetries are
the rule, with amplitudes usually somewhat higher on the right
side but, according to Petersén and Eeg-Olofsson (6), the differ-
ence seldom exceeds 20% of the greater amplitude.
The admixture of posterior slow activity constantly dimin-
ishes during the period of adolescence. In general, the amount
of intermixed delta–theta activity (between alpha waves) lies
around 10% to 15%; slightly or moderately higher figures are
still acceptable as compatible with normalcy. Healthy adoles-
■ Normal EEG and Sleep: Infants to Adolescents 177
cents’ slow frequencies may show a wider distribution and their
peak over the posterior regions is not quite as impressive as in
older children. “Slow fused transients” (61) are not necessarily
normal at this age.
A special form of posterior slow activity is the posterior 4- to
5-Hz rhythm (Fig. 10.16), which probably constitutes a mild
abnormality. This pattern is uncommon, with probably only
one case occurring in 1000 or 2000 tracings from a large EEG
laboratory. The mean age at first examination was 24.15 years
(65), which indicates that this is a pattern of adolescence and
young adulthood. Acquired cerebral pathology (especially cran-
iocerebral trauma), as well as genetic predisposition (65), may
play a determining role.
Another form of rhythmical posterior slow activity consists
of rhythmical 3- to 4-Hz waves of high voltage over the occipital
region and its immediate vicinity. This is referred to as OIRDA,
and it is activated by hyperventilation, eye closure after a latency
period of 300 to 500 msec and lasts for about 1.5 to 3 seconds,
drowsiness, and photic stimulation. The rhythmical 3- to 4-Hz
waves exhibit a more or less obvious sharp component,
sometimes a “notch” in the slow wave. This type of posterior
slowing has been investigated by Belsh et al. (82). These authors
found this pattern only in children and adolescents from the age
of 6 to 16 years. This pattern occurs in children and juveniles
with or without epileptic seizure disorder. One must agree with
Belsh et al., who think that “such a discharge should not be
interpreted as epileptiform activity,” although Riviello and
Figure 10.16 Patient age 21 years. An example of the
4- to 5-Hz rhythm over the posterior region. Note that
eye closure is followed by a few 12- to 13-Hz alpha
waves before a prominent posterior 4.5- to 5-Hz rhythm
emerges. This rhythm apparently vanishes shortly
before eye opening. The record was otherwise normal.
The patient was referred because of suspected psy-
chogenic seizures.
178 Part II ■ Normal EEG
Foley reported OIRDA as an ictal waveform (66). This pattern
is also known as the “phi rhythm” (83).
Fast activity is more widely seen in adolescents than in older
children. The maximum fast activity is most commonly found
over the frontal areas. Well-defined and circumscript beta waves
over the rolandic regions are seen, usually in conjunction with
the central mu rhythm.
The rolandic mu rhythm is fairly common in adolescents and
gradually declines from a peak reached between the age of 11
and 15 years (40). Occipital lambda waves show their typical
mature configuration. Anterior rhythmical 6- to 7-Hz theta
activity gradually decreases from its peak prevalence (the age of
13 to 15 years; see Fig. 10.16). A rare phenomenon is rhythmical
4-Hz activity over the vertex (“4-Hz vertex spindles”) described
by Van Huffelen and Magnus (84); this occurs in the waking
state only, mostly in adolescents, and especially in patients with
vasomotor instability, such as syncope.
Low-voltage records, very rare in childhood, are more com-
mon in adolescents and may occur in about 5% of patients;
these records are usually dominated by beta frequencies. This
type of low-voltage fast record has been found more often in
females (6,85,86). In healthy adults, the prevalence of low-
voltage records increases to 11.6% (46).
Activations
Hyperventilation does not yield the dramatic high-amplitude
slowing that is seen in children. Marked bilateral synchronous
Figure 10.17 Phantom spike–wave complex. The high-amplitude wave is the prominent component of this spike–wave
complex.
delta activity was found in 22% of the healthy adolescents stud-
ied by Petersén and Eeg-Olofsson (6). It should be noted that
hyperventilation is a poorly standardized procedure in which
motivation and effort play a paramount role.
Intermittent photic stimulation shows essentially the mature
type with a peak driving response in the medium and fast rate
of flicker (mostly 6 to 20 per second). In adolescents, the predis-
position to paroxysmal flicker responses must not be underrated;
such responses of the photomyoclonic as well as of the photo-
convulsive type (87) may occasionally occur in apparently
healthy individuals. Further investigation of such persons
sometimes unearths signs of mental, emotional, and autonomic
nervous system instabilities.
Drowsiness
The transition from wakefulness to sleep (stage 1) shows, in
general, the mature adult type with gradual alpha dropout and
long stretches of a diffuse low-voltage pattern with mixed slow
and fast activity. With deepening drowsiness, vertex waves
appear, and positive occipital sharp transients of sleep may be
present.
Sleep
In the commonly observed non-REM sleep onset recording in
the EEG laboratories, there is little difference from the activity
found in adults. The sleep spindle amplitude is somewhat
greater than in adulthood, and this is also generally true for this
 Chapter 10
amplitude of vertex waves and K complexes. Their sharp
component is not quite as sharp as in older children. Positive
occipital transients of sleep may be abundant in stage 2. REM
sleep and arousal from sleep show the mature type with very
rapid transition from sleep patterns to waking activity.
Variant Patterns of Uncertain Clinical
Significance
The 14- and 6-Hz positive burst is not uncommon at this age.
The incidence starts to decline, as demonstrated by Gibbs and
Gibbs (45) (15.8% at the age of 5 to 9, 20.8% at 10 to 14, and
1.65% at 15 to 19 years; there is marked further decline to
8.75% between the age of 20 and 24 years and 1.4% between 25
and 29 years). Eeg-Olofsson confirmed these findings (3). Deep
drowsiness and light sleep facilitate this pattern, which is most
prominent over the posterior temporal and occipital regions.
Lombroso et al (88) found 14- and 6-Hz positive bursts in 58%
of 13- to 15-year-old normal high school boys. They referred to
them as ctenoids, after the Greek, for comb.
The 6-Hz spike–wave complex, now typically referred to as the
phantom spike and wave, is occasionally seen in adolescents;
Gibbs and Gibbs (45) found this pattern in 2.8% of their 3476
control subjects, a higher figure than at any other age.
Considered a mild or moderate abnormality in the past, it is
now considered a normal variant or a pattern of uncertain clin-
ical significance. It is called the phantom spike and wave
because the wave component has a much higher amplitude
than the spike (Fig. 10.17). It is discussed in Chapter 26 with
epileptic seizure discharges.
The rhythmic temporal theta bursts of drowsiness, or the
psychomotor variant pattern, is much less common and reaches
a slight peak at the age of 15 to 19 years (0.4%), according to
Gibbs and Gibbs (45).
CONCLUDING STATEMENTS
This chapter provides a condensed description of EEG matura-
tion from birth through young adulthood. The basic rules of this
EEG maturation are not carved in stone and must be viewed with
a certain degree of flexibility. It is strongly recommended that
terms such as within broad (very broad) normal limits of variabil-
ity for age be used in mildly deviant but not definitely abnormal
tracings, rather than making rigid distinctions between the “nor-
mal” and “abnormal” character of a given record.
The “clinician within ourselves” will demand statements
concerning the clinical significance of mild deviations or defi-
nite abnormalities in the maturational process. The aforemen-
tioned normal variant patterns, such as the 14- and 6-Hz
positive spike discharge, rhythmic temporal theta bursts of
drowsiness, and 6-Hz spike–wave discharges, must be inter-
preted in a prudent and circumspect manner.
It is also very difficult to make a firm and solidly founded
statement as to whether deviations from the rules of EEG mat-
uration (for instance, excessive posterior slowing at a given age)
are simply a sign of immaturity or a sign of possible sustained
structural insult to the brain.
■ Normal EEG and Sleep: Infants to Adolescents 179
REFERENCES
1. Kellaway P. Ontogenetic evolution of the electrical activity of the
brain in man and animals. Fourth International Congress of
Electroencephalography and Clinical Neurophysiology. Acta Med.
Belg. 1957:141–154.
2. Berger H. Über das Elektroenzephalogramm des Menschen. 5.
Mittlg. Arch. Psychiatr. Nervenkr. 1932;98:231–254. [English trans-
lation: Gloor, P. 1969. Hans Berger. On the Electroencephalogram of
Man. Amsterdam: Elsevier (see pp. 157–160).]
3. Eeg-Olofsson O. The development of the EEG in normal children
from age 1 to 15 years. The 14 and 6 Hz positive spike phenome-
non. Neuropaediatrie. 1971;4:405–427.
4. Eeg-Olofsson O. The development of the electroencephalogram in
normal adolescents from the age of 16 through 21 years.
Neuropaediatrie. 1971;3:11–45.
5. Eeg-Olofsson O, Petersén I, Selldén U. The development of the
electroencephalogram in normal children from the age of 1
through 15 years. Paroxysmal activity. Neuropaediatrie. 1971;
2:375–404.
6. Petersén I, Eeg-Olofsson O. The development of the electroen-
cephalogram in normal children from the age of 1 through 15
years. Neuropaediatrie. 1971;2:247–304.
7. Petersén I, Selldén U, Eeg-Olofsson O. The evolution of the EEG in
normal children and adolescents from 1 to 21 years. In: Lairy GC, ed.
Handbook of Electroencephalography and Clinical Neurophysiology.
Vol 6B, ed.-in-chief, A. Remond. Amsterdam: Elsevier; 1975:31–68.
8. Matsuura M, Yamamoto K, Fukuzawa H, et al. Age development
and sex differences of various EEG elements in healthy children
and adults. Quantification by a computerized wave form recog-
nition method. Electroencephalogr Clin Neurophysiol. 1985;60:
394–406.
9. Gasser T, Verleger R, Bacher P, et al. Development of the EEG of
school-age children and adolescents. I. Analysis of band power.
Electroencephalogr Clin Neurophysiol. 1988;69:91–99.
10. Gasser T, Jennen-Steinmetz C, Sroka L, et al. Development of the
EEG of school-age children and adolescents. II. Topography.
Electroencephalogr Clin Neurophysiol. 1988;69:100–109.
11. Pampiglione G. Development of Cerebral Function in the Dog.
London: Butterworths; 1963.
12. Caveness WF. Atlas of EEG in the Developing Monkey (Macaca
mulatta). Reading, MA: Addison-Wesley; 1962.
13. Passouant P, Cadilhac J, Delange M. Le sommeil du nouveauné.
Considérations sur la période de mouvements oculaires. Arch Fr
Pediatr. 1965;22:1087–1092.
14. Ellingson RJ. EEGs of premature and full-term newborns. In: Klass
DW, Daly DD, eds. Current Practice of Electroencephalography. New
York: Raven Press; 1979:149–177.
15. Dreyfus C, Curzi-Dascalova L. The EEG during the first year of
life. In: Lairy GC, ed. Handbook of Electroencephalography and
Clinical Neurophysiology. Vol 6B, ed.-in-chief, A. Remond.
Amsterdam: Elsevier; 1975:24–30.
16. Sweeney D, Beckman D, Calmese F, et al. The use of modified sleep
deprivation to facilitate pediatric electroencephalographic record-
ings. Am J END Technol. 1997;37:218–230.
17. Brown V, Haak N, Egel RT. Technical tips: Age-specific sleep
deprivation guidelines. Am J END Technol. 1997;37:231–235.
18. Ong H-T, Lim KJL, Low P-C, et al. Simple instructions for partial
sleep deprivation prior to pediatric EEG reduces the need for seda-
tion. Clin Neurophysiol. 2004;115:951–955.
19. Leonberg SC. Management of infants and children for electroen-
cephalography. Clin Electroencephalogr. 1984;15:202–207.
180 Part II ■ Normal EEG
20. Epstein CM. Guideline Two: Minimum technical standards for
Pediatric Electroencephalography. J Clin Neurophysiol. 2006;23: 92–96.
21. Fois A. The Electroencephalogram of the Normal Child. Springfield,
IL: Charles C Thomas; 1961.
22. Stroganova TA, Orekhova EV, Posikera IN. EEG alpha rhythm in
infants. Clin Neurophysiol. 1999;110:997–1012.
23. Hagne I. Development of waking EEG in normal infants during
the first year of life. In: Kellaway P, Petersén I, eds. Clinical
Electroencephalography of Children. New York: Grune & Stratton
(Stockholm: Almqvist & Wiksell); 1968:97–118.
24. Hagne I, Persson J, Magnusson R, et al. Spectral analysis via fast
Fourier transform of waking EEG in normal infants. In: Kellaway
P, Petersén I, eds. Automation of Clinical EEG. New York: Raven
Press; 1972:3–48.
25. Ohtahara S. Development of electroencephalogram during
infancy and childhood. Proceedings of the Japanese EEG Society.
1964:18–23.
26. Samson-Dollfus D, Nogues B, Delagree E. EEG recording during
drowsiness in normal babies from 2 to 12 months of age.
Electroencephalogr Clin Neurophysiol. 1981;52:58P[abstract].
27. Curzi-Dascalova L, Monod N, Guidasci S, et al. Waking-sleeping tran-
sition in the newborn baby and in infants before the age of 3 months.
Electroencephalogr Clin Neurophysiol. 1981;52:57P [abstract].
28. Metcalf DR. The effect of extrauterine experience on the ontogen-
esis of EEG sleep spindles. Psychosom Med. 1969;31:393–399.
29. Katsurada ML. Electroencephalographic study of sleep in infants
and young children. I. Development of spindle waves (in
Japanese). Ann Paediatr Jpn. 1965;2:391–394.
30. Lenard HC. The development of sleep spindles during the first two
years of life. Neuropaediatrie. 1970;1:264–276.
31. Hughes JR. Development of sleep spindles in the first year of life.
Clin Electroencephalogr. 1996;27:107–115.
32. Niedermeyer E. The Generalized Epilepsies. Springfield, IL: Charles
C Thomas; 1972.
33. Metcalf DR, Mondale J, Butler FK. Ontogenesis of spontaneous K
complexes. Psychophysiology. 1971;8:340–347.
34. Westmoreland BF. Electroencephalography: Adult, Normal, and
Benign Variants. Clinical Neurophysiology. 3rd ed. Oxford: Oxford
University Press; 2009:119–136.
35. Ellingson RJ, Peters JF, Nelson B. Respiratory pauses and apnea
during daytime sleep in normal infants during the first year of life:
longitudinal observations. Electroencephalogr Clin Neurophysiol.
1982;53:48–59.
36. Roffwarg HP, Muzio JN, Dement WC. Ontogenetic development
of the human sleep dream cycle. Science. 1966;152:604–619.
37. Stern E, Parmelee AH, Akiyama Y, et al. Sleep cycle characteristics
in infants. Pediatrics. 1969;43:65–70.
38. Dittrichova J, Paul K, Pavlikova E. Rapid eye movements in para-
doxical sleep in infants. Neuropaediatrie. 1972;3:248–257.
39. Schulz H, Salzarulo P, Fagioli I, et al. REM latency: development in
the first year of life. Electroencephalogr Clin Neurophysiol.
1983;56:316–322.
40. Niedermeyer E, Koshino Y. My-Rhythmus: Vorkommen und klin-
ische Bedeutung. Z EEG-EMG. 1975;6:69–78.
41. Hrbek A, Vitova Z, Mares P. Optic evoked potentials in children
at different stimulation frequencies. Physiol Bohemoslov. 1966;15:
201–204.
42. Vitova Z, Hrbek A. Developmental study on the responsiveness of
the human brain to flicker stimulation. Dev Med Chid Neurol. 1972;
14:476–486.
43. Lindsley DB. A longitudinal study of the alpha rhythm in normal
children: frequency and amplitude standards. J Genet Psychol. 1939;
55:197–213.
44. Westmoreland BF, Stockard JE. The EEG in infants and children:
normal patterns. Am J EEG Technol. 1977;17:187–206.
45. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 3.
Reading, MA: Addison-Wesley; 1964.
46. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 1.
Cambridge, MA: Addison-Wesley; 1950.
47. Brandt H, Brandt S, Vollmund K. The electroencephalographic
patterns in young healthy children from 0 to five years of age. Acta
Psychiatr Scand. 1955;30:77–89.
48. Slater GE, Torres F. Frequency-amplitude gradient. A new param-
eter for interpreting pediatric sleep EEGs. Arch Neurol.
1979;36:465–470.
49. Cadilhac J. Ontogenesis and phylogenesis of sleep. In: Passouant P,
ed. Handbook of Electroencephalography and Clinical Neurophysio-
logy. Vol 7A, ed.-in chief, A. Remond. Amsterdam: Elsevier; 1975:
18–25.
50. Gibbs EL, Gibbs FA. Extreme spindles: correlation of electroen-
cephalographic sleep pattern with mental retardation. Science.
1962;138:1106–1107.
51. Niedermeyer E, Capute AJ. A fast and spiky spindle variant in chil-
dren with organic brain disease. Electroencephalogr Clin
Neurophysiol. 1967;23:67–73.
52. Hess R. The electroencephalogram in sleep. Electroencephalogr
Clin Neurophysiol. 1964;16:44–55.
53. Kellaway P, Fox BJ. Electroencephalographic diagnosis of cerebral
pathology in infants during sleep. I. Rationale, technique and the
characteristics of normal sleep in infants. J Pediatr. 1952;41:
262–287.
54. White JC, Tharp BR. An arousal pattern in children with organic
cerebral dysfunction. Electroencephalogr Clin Neurophysiol. 1974;
37:265–268.
55. Hughes JR, Daaboul Y. The frontal arousal rhythm. Clin
Electroencephalogr. 1999;34:13–14.
56. Hughes JR. The frontal arousal rhythm (FAR) is an ictal pattern: a
case report. Clin Electroencephalogr. 2003;34:13–14.
57. Niedermeyer E, Yarworth S. Scarcity of minor EEG abnormalities
during the first two years of life. Electroencephalography. 1978;
9:20–28.
58. Leissner P, Lindholm LE, Petersén I. Alpha amplitude dependency
on skull thickness as measured by ultrasound technique.
Electroencephalogr Clin Neurophysiol. 1970;29:392–399.
59. Gibbs FA, Gibbs EL, Lennox WG. Electroencephalographic
response to overventilation and its relation to age. J Pediatr. 1943;
23:497–505.
60. Sunder TR, Erwin CW, Dubois PJ. Hyperventilation induced
abnormalities in the electroencephalogram of children with
Moya-Moya disease. Electroencephalogr Clin Neurophysiol. 1980;
49:414–420.
61. Kellaway P, Crawley JW, Kagawa N. A specific electroencephalo-
graphic correlate of convulsive equivalent disorders in children.
J Pediatr. 1959;55:582–592.
62. Aird RB, Gastaut Y. Occipital and posterior electroencephalographic
rhythm. Electroencephalogr Clin Neurophysiol. 1959;11:637–656.
63. Kuhlo W. Slow posterior rhythms. In: Handbook of Electroenc-
ephalography and Clinical Neurophysiology. Vol 6A, Ed.-in-chief, A.
Remond. Amsterdam: Elsevier; 1976:89–104.
64. Attarian HP, Pacquiao PA, Erickson SM. Unilateral alpha subhar-
monics: a case report. Clin Electroencephalogr. 2001;32:32–35.
65. Kuhlo W, Heintel H, Vogel F. The 4–5/sec rhythm. Electroen-
cephalogr Clin Neurophysiol. 1969;26:613–618.
66. Riviello JJ, Foley CM. The epileptiform significance of intermittent
rhythmic delta activity (IRDA) in childhood. J Child Neurol. 1992;
7:156–160.
 Chapter 10
67. Palmer FB, Yarworth S, Niedermeyer E. Frontal midline theta
rhythm. Electroencephalogr Clin Neurophysiol. 1976;7:131–138.
68. Doose H, Gerken H, Horstmann T, et al. Genetics of centren-
cephalic epilepsy in childhood. Epilepsia (Amsterdam). 1968;9:
107–115.
69. Doose H, Gerken H, Horstmann T, et al. On the genetics of EEG
anomalies in childhood. I. Abnormal theta rhythms. Neuropaediatrie.
1972;3:386–401.
70. Doose H, Gerken H, Horstmann T, et al. Genetic factors in spike-
wave absences. Epilepsia (Amsterdam). 1973;14:57–75.
71. Schauseil-Zipf U, Hamm W, Mandl-Kramer S. EEG findings in
“small-for-date babies” at 7–12 years of age. Electroencephalogr
Clin Neurophysiol. 1984;57:33P–34P[abstract].
72. Daute KH, Frenzel J, Klust E. Über den unspezifischen
Hyperventilationseffekt im EEG des gesunden Kindes. I. Stärkegrad.
Z Kinderheilkd. 1968;104:197–207.
73. Herrlin KM. EEG with photic stimulation: a study of children with
manifest or suspended epilepsy. Electroencephalogr Clin Neurophysiol.
1954;6:573–589.
74. Doose H, Gerken H, Hien-Volpel KIF, et al. Genetics of photosen-
sitive epilepsy. Neuropaediatrie. 1969;1:56–73.
75. Klass DW, Westmoreland BF. Nonepileptogenic epileptiform elec-
troencephalographic activity. Ann Neurol. 1985;18:627–635.
76. Westmoreland BF. Benign EEG variants and patterns of uncertain
clinical significance. In: Ebersole JS, Pedley TA, eds. Current
Practice of Clinical Electroencephalography. 3rd ed. Philadelphia:
Lippincott Williams & Wilkins; 2003:235–245.
77. Tatum WO, IV, Husain AM, Benbadis SR, et al. Normal adult EEG
and patterns of uncertain significance. J Clin Neurophysiol. 2006;
23:194–207.
■ Normal EEG and Sleep: Infants to Adolescents 181
78. Gibbs EL, Gibbs FA. Electroencephalographic evidence of thala-
mic and hypothalamic epilepsy. Neurology (Minneapolis). 1951;1:
136–144.
79. Gibbs FA, Rich CL, Gibbs EL. Psychomotor variant type of seizure
discharge. Neurology. 1963;13:991–998.
80. Lipman IJ, Hughes JR. Rhythmic mid-temporal discharges. An
electro-clinical study. EEG Clin Neurophysiol. 1969;27:43–47.
81. Samson-Dollfus D, Goldberg P. Electroencephalographic quantifi-
cation by time domain analysis in normal 7–15 year old children.
Electroencephalogr Clin Neurophysiol. 1979;46:147–154.
82. Belsh JM, Chokroverty S, Barabas G. Posterior rhythmic slow activ-
ity in EEG after eye closure. Electroencephalogr Clin Neurophysiol.
1983;56:562–568.
83. Silbert PL, Radhakrishnan K, Johnson J, et al. The significance of the
phi rhythm. Electroencephalogr Clin Neurophysiol. 1995;95: 71–76.
84. Van Huffelen AC, Magnus O. 4c/sec vertex spindles. Electroen-
cephalogr Clin Neurophysiol. 1973;34:543–546.
85. Mundy-Castle AC. Theta and beta rhythm in the electroencephalo-
gram of normal adults. Electroencephalogr Clin Neurophysiol.
1951;3:477–486.
86. Vogel F, Fujiya Y. The incidence of some inherited EEG variants
in normal Japanese and German mates. Hum Genet. 1969;7:
38–42.
87. Bickford RG, Sem-Jacobsen CW, White PC, et al. Some observa-
tions of the mechanisms of photic and photo-metrazol activation.
Electroencephalogr Clin Neurophysiol. 1952;4:275–282.
88. Lombroso CT, Schwartz IH, Clark DM, et al. Ctenoids in healthy
youths: controlled study of 14 and 6 per second positive spiking.
Neurology. 1966;16:1152–1158.

Normal EEG and Sleep:
Adults and Elderly
BERNARD S. CHANG, DONALD L. SCHOMER, AND ERNST NIEDERMEYER
INTRODUCTION
There is perhaps nothing as compellingly beautiful to a clinical
neurophysiologist as the appearance of a normal adult elec-
troencephalogram (EEG) recording, in wakefulness and sleep.
But the relative order, symmetry, and transitions that populate
such a recording and are so welcome to the clinically trained eye
can lead us to forget the fundamentally remarkable phenomena
that electrical activity from the human brain can indeed be
measured routinely by electrodes placed only on the scalp; that
defined, well-catalogued abnormalities in the resulting meas-
urements can be identified and interpreted with clinical mean-
ing; and that our entire understanding of this field has
essentially arisen within only the past century.
A chapter on the normal EEG is actually more difficult to
organize than it might seem. Many clinical electroencephalog-
raphers would define normality on an EEG as being simply the
absence of identifiable abnormalities. In other words, there is
no single rhythm, feature, or characteristic of an EEG that
must be present to make it normal (except, perhaps, reactiv-
ity), and, indeed, no combination of any such findings either.
Such is the nature of human EEG, and this must be kept in
mind in the clinical interpretation of tracings. Having said
that, decades of EEG study, beginning with Berger’s early fun-
damental descriptions and continuing now with advanced
power spectral analyses, have shown us that there are certain
basic recognizable patterns, seen at least some of the time in
almost all healthy individuals, that together constitute what
we would consider to be representative of a normal EEG. It is
the goal of this chapter to define and describe these recogni-
zable elements, with attention to their clinical implications and
their significance for our understanding of underlying cerebral
physiology.
EEG FREQUENCIES
The EEG contains a wide frequency spectrum, but it is not
simply a hodgepodge of frequencies. Rhythmicity seems to
create some law and order among waves of various lengths
and amplitudes. The impression of prevailing rhythmicity and
organization, however, is not a yardstick for the normality of
an EEG. Pronounced rhythmicity may be a sign of abnorma-
lity, and a chaotic appearance does not necessarily imply
abnormality. Reactivity may be the magic word in such cases;
an EEG of mixed frequencies may be quite responsive to cer-
tain stimuli.
CHAPTER
11
The frequency range of the EEG has a fuzzy lower and upper
limit. There are ultraslow and ultrafast frequency components
that play no significant role in the clinical EEG, with the
exception of ultraslow activity in profound coma and near-ter-
minal states. For these reasons, the frequency-response curve of
an EEG apparatus concentrates on the clinically relevant fre-
quency range, which is also the most important from the psy-
chophysiologic viewpoint. This range lies between 0.1/sec (or
cycles per second [cps] or Hz) and 100/sec and, in a more
restricted sense, between 0.3 and 70/sec. In the normal adult,
the slow ranges (0.3 to 7/sec) and the very fast range (above
30/sec) are sparsely represented; medium (8 to 13/sec) and fast
(14 to 30/sec) ranges predominate.
These frequencies are broken down into the following bands
or ranges:
Delta below 3.5/sec
Theta 4 to 7.5/sec
Alpha 8 to 13/sec
Beta 14 to 30/sec
Gamma above 30/sec
The sequence of these Greek letters is not logical and can be
understood only in the historical view. The terms alpha and
beta rhythm or waves were introduced by Berger (1); the term
gamma rhythm was subsequently used by Jasper and Andrews
(2) to designate frequencies above 30 or 35/sec; these were
essentially 35 to 45/sec and superimposed on the occipital alpha
rhythm (see Ref. 3). This term was temporarily abandoned, and
gamma frequencies became a part of the beta range.
The use of the term gamma rhythm or gamma frequency
range made an impressive comeback during the 1990s. The use
of a fast beta range and a very fast gamma range might have
been convenient for those who utilize frequency analysis with
power spectra. Furthermore, EEG rhythm research in the 1990s
unearthed the all-but-forgotten term gamma rhythm (4–7). In
recent EEG rhythm research, rhythmic activities of the brain are
often conceived mainly as induced rather than as spontaneous
rhythms.
The term delta rhythm was introduced by Walter (8) to des-
ignate all frequencies below the alpha range. Walter himself,
however, found a need to introduce a special designation for the
4 to 7.5/sec range and used the letter theta. He thus bypassed
the Greek letters epsilon, zeta, and eta; he chose theta to stand
for thalamus because he presumed a thalamic origin of these
waves (also see Ref. 9).
183
184 Part II ■ Normal EEG

The term pi rhythm has been used for the designation of
posterior slow rhythms (3 to 4/sec) without harmonious rela-
tionship to the posterior alpha rhythm according to Dutertre
(3) who recommended the preferable (although certainly less
precise) term posterior slow rhythms. Hardly anyone has used
the term pi rhythm since the 1990s.
The term phi rhythm was suggested (according to Ref. 10) for
the designation of monorhythmic posterior delta waves (less
than 4/sec), distinct from the background and occurring within
2 seconds of eye closure. This rhythm was also described by
Belsh et al. (11) as “posterior rhythmic slow activity after eye
closure.”
The alpha-like anterior temporal kappa rhythm (12) is a con-
troversial pattern that is discussed later in this chapter. Kugler
(13) has been using the term sigma activity instead of sleep spin-
dles and, furthermore, the term sigma rhythm for activity in the
11 to 15/sec range. The term rho waves was used for the activity
known as positive occipital sharp transients of sleep (POSTS)
(14), but it has disappeared.
Other Greek letters have been proposed for the designation
of distinct EEG activities. Mu rhythm and lambda waves are
discussed in this chapter. The term tau rhythm denotes a phys-
iologic alpha rhythm of the temporal region (perhaps identical
with the “third rhythm” discussed in this chapter). “Zeta wave”
simply denotes a certain type of delta wave (rather than a
rhythm) with some sharp configuration (15,16).
Thus, 14 of the 23 letters of the Greek alphabet are being
used in the EEG terminology, and this number could be even
higher. In our opinion, it might be better to limit the Greek
terms to the classical EEG frequency ranges retaining solely the
letters alpha, beta, gamma, delta, and theta.
whereas the vast majority of low-voltage records are “desyn-
chronized” (discussed later) and are a variant of normalcy.
BASIC EEG PATTERNS SEEN
IN WAKEFULNESS
Alpha Rhythm
Definition
The International Federation of Societies for Electro -
encephalography and Clinical Neurophysiology (IFSECN)
(committee chaired by G. E. Chatrian; see Ref. 18) proposed the
following definition of alpha rhythm (Fig. 11.1):
Rhythm at 8 to 13 Hz occurring during wakefulness over the
posterior regions of the head, generally with higher voltage over
the occipital areas. Amplitude is variable but is mostly below 50
µV in adults. Best seen with eyes closed and under conditions of
physical relaxation and relative mental inactivity. Blocked or
attenuated by attention, especially visual, and mental effort (18).
This committee also has pointed out that the term alpha
rhythm must be restricted to rhythms fulfilling all of the above
criteria. Rolandic mu rhythm may have the same frequency
range, but its topography and reactivity are different (also
see Ref. 19).
Frequency
EEG maturation shows the gradual frequency increase of a pos-
terior basic rhythm that is detectable around the age of 4 months
with a frequency of approximately 4/sec. This posterior
dominant rhythm shows a progressive frequency increase with

EEG AMPLITUDES
The EEG denotes voltage plotted against time. The voltage of
the EEG signal determines its amplitude. The passage of the
cortical EEG signal through leptomeninges, cerebrospinal fluid,
dura mater, bone, galea, and scalp has a strongly attenuating
effect on the original signal (17). Corticographic discharges
show amplitudes of 500 to 1500 µV (0.5 to 1.5 mV) and several
millivolts in prominent spiking. The amplitudes of the scalp
EEG are markedly reduced and lie between 10 and 100 µV (in
adults, more commonly between 10 and 50 µV).
The EEG amplitudes are measured from peak to peak.
Precise determination of the voltage of each wave is unneces-
sary and should be discouraged as pseudoaccuracy; too many
variables are involved (above all, the interelectrode distance and
the type of montage, whether bipolar or referential recording).
Electroencephalographers may indicate in their reports a cer-
tain amplitude range, such as “alpha rhythm from 20 to 30 µV,”
or, even better, limit themselves to statements such as “of
medium voltage” or “of low to medium voltage.” Figure 11.1 Normal record in a patient aged 27 years. Posterior 10 to
A given frequency can be rendered abnormal by excessive 12/ sec alpha rhythm. Good blocking response to eye opening and reac-
voltage. This is true for all frequencies, and it is particularly tivation of alpha rhythm with eye closure. Note low-voltage fast activ-
important for the fast (beta) band. The problem of low voltage ity and some return of alpha rhythm while eyes are kept open. Also
will be thoroughly discussed, because low amplitudes can indi- note typical artifacts with eye opening, eye closure, and eye blink
cate a life-threatening decline of cerebral voltage output, (caused by eye potential shifts).
 Chapter 11
average values of around 6/sec at the age of 12 months and 8/sec
at the age of 3 years. At that time, the alpha frequency band is
reached, and there is justification for the use of the term alpha
rhythm. The frequency reaches a mean of about 10/sec at the age
of 10 years. This is essentially the mean alpha frequency of adult-
hood; in other words, the progressive alpha rhythm acceleration
usually ends around the age of 10 years, but the second decade
of life (and to some degree also the third decade) features a con-
stant decline of intermixed posterior slow activity that is usually
present in considerable quantity at the age of 10 years.
The frequency of the alpha rhythm tends to decline in eld-
erly individuals. This decline apparently reflects some degree of
cerebral pathology, which may be vascular or degenerative, in
many instances. Healthy and vigorous elderly people may show
little or no alpha frequency decline, even in the ninth decade.
The figure of 10.2 ± 0.9/sec has been indicated as the mean
adult alpha frequency (20). An element of instability of the
alpha frequency must be taken into consideration; according to
Townsend et al. (21), the alpha rhythm frequency can be stabi-
lized by sinusoidally modulated light. Extreme upward gaze
tends to facilitate the posterior alpha rhythm (22,23). Lateral
eye deviations may have similar effects (24).
An alpha rhythm frequency shift to the faster portion of the
band is not uncommon and is essentially within normal limits,
as will be discussed later. The similarities between the frequen-
cies of alpha rhythm and the physiologic finger tremor have
been discussed by Isokawa and Komisaruk (25). Immediately
after eye closure, the alpha frequency may be accelerated for a
moment (“squeak effect,” after Ref. 26).
Amplitude
Alpha rhythm amplitudes vary considerably from individual to
individual and, in a given person, from moment to moment.
The electroencephalographer, therefore, should look for
stretches of optimal alpha output. A referential montage to the
ipsilateral ear is usually most suitable for the determination of
the alpha rhythm amplitude, but the interelectrode distances
must always be considered. The maximum alpha voltage is usu-
ally over the occipital region as such, but a bipolar montage
with a parasagittal array may obscure rather than reveal the true
alpha maximum. The alpha amplitude may be quite small in
the channels displaying P3–O1 and P4–O2 because of massive
homophasic activity, which results in canceling out.
The alpha amplitudes tend to show constant waxing and
waning. For this reason, trains of alpha waves show a typical
spindle shape with a belly and a thin portion. However, the
term spindle has been reserved for a classical pattern of sleep
and should not be used in this context.
Berger (1) found alpha rhythm voltages of 15 to 20 µV; these
are small values when one considers his fronto-occipital record-
ing technique; the smallness was probably due to the limitations
of his Edelmann string galvanometer. According to Cobb (27),
the alpha rhythm voltage fluctuates between 0 and 40 to 50 µV
in the individual record; values above 100 µV are uncommon in
the adult, whereas maximums of 5 to 10 µV are frequently seen
(27). The work of Simon (also known as Simonova) and her
coworkers has shed more light on this subject; Simonova et al.
■ Normal EEG and Sleep: Adults and Elderly 185
(28) found amplitudes between 20 and 60 µV in 66% of their
subjects; values below 20 µV were found in 28% and above 60
µV in 6% (also see Ref. 29).
Higher alpha amplitudes are more likely to be found in asso-
ciation with slower alpha frequencies (30,31). There is good
evidence of a mild-to-moderate alpha amplitude asymmetry
with a higher voltage on the right (20,27,29,31–33). This seems
to indicate that the alpha rhythm is of greater amplitude over
the nondominant hemisphere, but no convincing conclusion
concerning handedness can be derived from this asymmetry
(20). This physiologic asymmetry has been confirmed by
Matousek et al. (34), who also found a reversal of this rule (i.e.,
higher voltage on the left occiput in patients with endogenous
depression).
Amplitude asymmetries must be demonstrated in both ref-
erential and bipolar montages from two or more posteriorly
placed electrodes (such as parietal and posterior temporal)
before they are considered significant (19). It may be difficult to
distinguish between physiologic and truly abnormal alpha
amplitude asymmetries.
Morphology
The alpha rhythm is usually characterized by rounded or sinu-
soidal wave forms. However, a sizable minority of individuals
have sharp alpha configuration. In such cases, the negative
component appears to be sharp and the positive component
appears to be rounded, similar to the wave morphology of
rolandic mu rhythm.
The sharp configuration of posterior alpha waves is by no
means an abnormality. It is a common finding, especially in
young adults, adolescents, and older children. An admixture of
beta waves is usually the cause of the sharp configuration; drug
effects from sedatives or minor tranquilizers must sometimes
be suspected in such cases.
Spatial Distribution
The alpha rhythm is clearly a manifestation of the posterior half
of the head and is usually found over occipital, parietal, and
posterior temporal regions. This observation of Adrian and
Matthews (35) was doubted by Berger (36), whose concept of
alpha rhythm as a global cerebral rhythm was an erroneous
conclusion from his fronto-occipital bipolar recording tech-
nique. The alpha rhythm may extend into central areas, the ver-
tex, and also the midtemporal region. When the central region
is strongly involved, the alpha rhythm must be distinguished
from possibly coexisting rolandic mu rhythm. This is usually
easily demonstrable with eye opening, which should lead to
attenuation of the alpha rhythm but not the mu rhythm.
The alpha rhythm may occasionally extend slightly into the
superior frontal leads (F3, F4). Extension into the frontopolar
region (Fp1, Fp 2) is practically unheard of. Apparent alpha
rhythm in the frontopolar leads may be very prominent in ref-
erential (unipolar) montages if the referential ear electrode
picks up the posterior alpha rhythm. This is particularly com-
mon when the mastoid region is used instead of the ear lobe
(the mastoid being a preferred place with paste technique).
186 Part II ■ Normal EEG
Another source of confusion is eyelid flutter, with closed eyes
giving rise to frontal artifacts in alpha frequency.
In depth electroencephalography and with occipital
implants, posterior alpha rhythm can be demonstrated
throughout the depth of the occipital lobe and even in the
vicinity of the lateral geniculate body. According to Albe-
Fessard (37), alpha rhythm may be recorded from the medial
pulvinar but not from the remaining thalamic nuclei in the
human (also see Ref. 38). In the dog, Lopes da Silva et al. (39)
demonstrated that alpha rhythm of the same peak frequency,
bandwidth, and reactivity can be recorded from the visual cor-
tex as well as from the visual thalamus (lateral geniculate body,
pulvinar nuclei).
Reactivity
The posterior alpha rhythm is temporarily blocked by an
influx of light (eye opening), other afferent stimuli, and men-
tal activities. The degree of reactivity varies; the alpha rhythm
may be completely blocked, suppressed, or attenuated with
voltage reduction. The alpha blocking response to eye opening
was discovered by Berger and described in his first report
(1) on the human EEG; it came as a great surprise for investi-
gators who were searching for action potentials and hence
would have expected enhancement of EEG voltage with influx
of light. Berger’s own explanation was the concept of a zone of
inhibition surrounding the area of excitation by the afferent
stimulus (40,41).
It is noteworthy that in some subjects the alpha suppression
is more pronounced than in others. These persons may have an
alpha-free stretch of desynchronized and chiefly fast EEG activ-
ity for the duration of the eye opening, even in rather dim light.
In others, the alpha blockage lasts for less than 1 second.
According to Gibbs and Gibbs (42) “after the eyes have been
held open for a few minutes, low voltage alpha waves … usually
reappear, unless the subject continues to look at something
which holds his interest.” The amplitude ratio between eyes
closed (well-developed alpha) and eyes open (beta of much
smaller voltage) declines with advancing age (43).
Alpha attenuation due to auditory, tactile, and other
somatosensory stimuli or heightened mental activity (such as
solving difficult arithmetical problems) is usually less pro-
nounced than the blocking effect with eye opening. Indeed,
the blocking or attenuating effect of mental arithmetic on the
posterior alpha rhythm may not be present in most cases (44).
In a study done in 1280 patients (598 with normal EEG
tracings, 682 with various degrees of EEG abnormality), alpha
suppression or attenuation with calculation (serial sevens)
was noted in only 21 (1.6%) of the patients. On the other
hand, Berger (40,45–48) showed fine examples of arithmetic-
induced alpha blocking. A possible explanation for the
discrepancy lies in the difficulty of the task: the arithmetical
problems given to Berger’s two teenage children, Klaus
and Ilse, were more challenging. Other factors may also be
at work: even simple arithmetic may cause alpha blocking if
the subject shows great motivation trying to please the exam-
iner. With a nonchalant approach, there is usually no alpha
blocking.
No EEG is complete without certain reactivity tests.
Although alpha rhythm disappears with the earliest approach
of drowsiness, there are exceptional cases of persisting posterior
alpha rhythm in profoundly comatose patients with pontine
vascular lesions; their alpha rhythm shows no reactivity, even
with strong nociceptive stimuli. According to Kiloh et al. (33),
“complete unresponsiveness of the alpha rhythm to visual stim-
uli is a rare and unequivocally abnormal finding.”
Interindividual Differences
When Adrian and Matthews demonstrated their own EEGs on
May 12, 1934, at a meeting of the astonished members of the
Physiological Society in Cambridge, England, it was found
that Adrian’s 10/sec alpha rhythm was quite impressive,
whereas Matthews produced “no regular waves” (49). Studies
of further subjects showed that Adrian’s alpha development
was that of the majority, whereas Matthews, the ingenious
engineer and designer of Adrian’s instrumentation, belonged
to a minority of persons with little or no organized alpha
rhythm.
Davis and Davis (50) distinguished four types of records: (a)
dominant alpha (found in 20% of healthy adults); (b) subdom-
inant alpha (35%); (c) mixed alpha (20%); and (d) rare alpha
(25%). Golla et al. (51) distinguished three alpha types: M for
minus or minimal, P for persistent, and R for responsive. The P
type shows no real persistence of alpha, but has a very short
blocking response to eye opening (also see Ref. 33). Another
type of alpha rhythm is the “monotonous high voltage alpha,”
which shows little or no waxing and waning of the amplitude
(52). These alpha traits may be, to some degree, genetically
transmitted (50,53,54); marked similarities have been reported
in identical twins. Small differences in the EEG of males and
females have been reported by Veldhuizen et al. (55).
A detailed review of historical EEG literature on the alpha
rhythm inevitably uncovers multiple publications suggesting a
relationship, or lack thereof, between alpha rhythm traits and
particular personality features (56–61), with little evidence of
reproducible findings. Similarly, attempts at simple correlations
between alpha rhythm and intelligence have largely been futile
or unconvincing (36,45,62–67), but with the use of power spec-
tral analysis, Gasser et al. (68) found good correlations between
alpha rhythm (plus other EEG criteria) and intelligence, and
individuals who are faster in retrieving information from mem-
ory tend to have higher alpha rhythm frequencies (69).
Intraindividual Differences
A person’s EEG traits and alpha rhythm development must not
be considered as permanent and unchangeable features like fin-
gerprints. Every EEG tracing of some length gives testimony to
a certain degree of variability, even within the state of relaxed
vigilance. This is probably the effect of a large number of phys-
iologic and psychophysiologic variables. To mention just one
example, the waxing phase of alpha amplitudes has been related
to the waning phase of afterimages (70). With advancing age,
the alpha frequency tends to decrease (71) but this is probably
the effect of cerebral pathology (vascular or degenerative)
occurring at old age.
 Chapter 11 ■ Normal EEG and Sleep: Adults and Elderly 187

Circadian studies of the EEG have been done sparingly
(72,73) and with inconclusive results (for further data, see
Ref. 74). The menstrual cycle has been thought to influence the
EEG and especially the alpha rhythm frequency; Harding and
Thompson (74) have summarized the work in this field as fol-
lows:
Preovulatory phase (days 5 to 14): Alpha frequency increased,
amount of beta increased, photic driving reduced.
Postovulation or luteal phase (days 15 to 23): Alpha slower,
amount of alpha increased, less beta and more theta activity,
photic driving increased.
Premenstrual phase (days 23 to 28): Alpha frequency increased,
amount of alpha reduced, more beta, less theta activity.
Menstrual phase (days 1 to 5): Alpha frequency slowed and
amount increased, less beta, more theta activity.
Body temperature also modifies the alpha frequency; the
alpha rhythm accelerates with increasing body temperature
(75). Therapeutic hyperthermia (up to 41°C) in cancer patients
slows down the EEG to the delta range and depresses the over-
all EEG output (76).
Relationship to Vigilance and Anxiety
Alpha rhythm is the classical EEG correlate for a state of relaxed
wakefulness best obtained with the eyes closed. A degree of higher
alertness attenuates or suppresses the alpha rhythm, which is then
supplanted by “desynchronized” low-voltage fast activity.
The earliest stage of drowsiness is characterized by “alpha
dropout” (Fig. 11.2). The trains of alpha waves become less and
less continuous, and the last alpha fragments finally give way to
a low-voltage pattern of mixed slow (mostly theta range) and
fast frequencies. This type of alpha dropout is a hallmark of a

Figure 11.2 A–C: Three examples of alpha dropout at the earliest stage of
drowsiness in normal adults. The close-up shows clearly the mixture of low-
voltage fast and slow activity following alpha dropout. This is different from
the low-voltage fast activity in the alpha blocking response to eye opening and
heightened alertness.
188 Part II ■ Normal EEG
normal adult EEG; in children and infants, various types of
slow patterns appear. In adults with neurodegenerative disease,
a posterior alpha rhythm of normal appearance may be
replaced by activity in the theta and delta range; a special pat-
tern with rhythmical widespread (but mainly parietotemporal)
5 to 7/sec activity has been individualized recently by
Westmoreland and Klass (77), who consider this pattern
“benign.”
Voluntary control of the alpha rhythm and the use of alpha
feedback methods have been widely discussed topics since the
late 1960s. This work was presumably prompted by the obser-
vation of well-modulated alpha during meditation practiced by
yogis (78,79) or zen buddhists (Hirai, 1968, and Kasamatsu and
Hirai, 1966, both articles cited in Ref. 80). Nowlis and Kamiya
(81) and Brown (82) associated alpha rhythm enhancement
(on a “voluntary” basis) with a pleasant mood. Further work in
this field was done by Wallace (83), Wallace et al. (84), and
Banquet (85), who studied states of transcendental meditation.
Alpha amplitudes increased or decreased; in some subjects
there were periods of low-voltage theta activity. Knott (67) feels
that such states are in essence periods of drowsiness and doubts
the validity of the view that the high alpha state is a desirable
condition. According to Stigsby et al. (86), there is no consistent
EEG pattern associated with successful or unsuccessful tran-
scendental meditation. In the end, the alpha rhythm as studied
here may simply reflect the level of vigilance (see Ref. 87).
Because a relaxed waking state is the optimal condition for the
posterior alpha rhythm, however, it is hence reasonable to
assume that emotional tension attenuates or blocks the alpha
rhythm. This seems to hold true for the state of emotional ten-
sion in patients or subjects with pending litigation, after a head
injury, or after other forms of physical damage, for example
(88).
Cerebral Generators of the Alpha Rhythm
The alpha rhythm is of cortical origin, but the theory of a thal-
amic pacemaker function has frequently surfaced since the
work of Berger (40), who presumed cortical genesis but thala-
mic governance of the alpha phenomenon. Bishop (89) pro-
posed the concept of corticothalamic reverberating circuits,
and Andersen and Andersson (90) are the proponents of a thal-
amic theory that is based on presumed similarities between
human alpha rhythm and experimental barbiturate spindle
activity in animals. According to this theory, the alpha rhythm
is driven by presynaptic input to cortical neurons from the thal-
amic level (also see Refs. 91 and 92). This concept has been
challenged by Lopes da Silva et al. (39,93). Watanabe (94) has
postulated a “somewhat loose but stable oscillator system” sub-
serving the generation of alpha rhythm.
Adrian and Yamagiwa (95) regarded the alpha rhythm as
cortical with maximal involvement of the visual area.
Important new vistas were opened with the demonstration of
some degree of interhemispheric asynchrony between alpha
waves (96). It has been presumed that there is more than one
alpha generator within the posterior regions of the cerebrum
(97); this was further substantiated by depth EEG studies in the
human (98). The technique of chronotopography has added
further insight into the possibility of multiple sources of alpha
generation (99). Further work on interhemispheric phase dif-
ferences between alpha waves has been carried out with topo-
scopic analysis (100) and cross-correlation technique (101).
Posterior alpha generation and spread in posteroanterior
direction (97) is generally acknowledged, but has been chal-
lenged by Inouye et al. (102), who feel that alpha spread occurs
in an anteroposterior direction in the dominant as well as in the
nondominant hemisphere, based on the method of entropy
analysis.
Our comprehension of alpha-rhythm genesis has not strik-
ingly increased in past several decades. It may be assumed that
there are corticocortical and thalamocortical systems that inter-
act in the generation of cortical alpha rhythms (103). From the
experimenter’s as well as from the clinical electroencephalogra-
pher’s viewpoint, there is good reason to presume that alpha
rhythm is most definitely a cortical phenomenon but there has
been, thus far, no evidence of a synchronizing mechanism at the
cortical level. No single neurophysiologic alpha rhythm theory
has yet found general acceptance, and there are still uncertain-
ties about the origin and psychophysiologic significance of this
remarkable phenomenon. And yet, our insights into the nature
of the alpha rhythm (and other EEG rhythms) have been deep-
ening.
Mu Rhythm
Rolandic (central) mu rhythm is in frequency and amplitude
related to the posterior alpha rhythm, but its topography and
physiologic significance are quite different (Fig. 11.3).
Historically, the existence of a special central rhythm was pre-
sumed by Jasper and Andrews (2) (“precentral alpha rhythm”),
Maddocks et al. (104) (“alphoid activity”), and Schütz and
Müller (105) (“high voltage rolandic alpha”). The features of
mu rhythm were first described in detail by Gastaut et al. (106)
and Gastaut (107), who also included electrocorticographic
tracings; these authors introduced the term rhythme rolandique
en arceau. The epithet en arceau alludes to the arch-shaped wave
morphology, which has also prompted the term wicket rhythm
(108). Magnus (109) used the term central alpha (see also Ref.
2). Other terms are arcade rhythm (110), comb rhythm (27), and
somatosensory alpha rhythm (111).
Mu rhythm is not detectable in every mature subject; its
prevalence is limited unless the “hidden” mu rhythm is visual-
ized with special methods. Mu stands for motor; this rhythm is
strongly related to functions of the motor cortex, but the con-
tribution of the adjacent somatosensory cortex must not be
ignored.
Mu rhythm and associated beta activity over the sensorimo-
tor cortex have become a topic of special interest in the past two
decades. Indeed, mu rhythm as an interesting clinical EEG phe-
nomenon has transcended into a powerful contributor to the
understanding of motor activity in general. This development
started with the work of Pfurtscheller on event-related desyn-
chronization (ERD) (112,113). The ERD pertains even more
strongly to the rolandic mu rhythm (and furthermore to vari-
ous cognitive tasks and their cortical EEG accompaniment)
(114–119). In other words, the ERD has become a new criterion
 Chapter 11
Figure 11.3 A: Rolandic mu rhythm, in a patient aged 13 years. Stretches
of central mu rhythm while eyes are kept open. These runs occur asyn-
chronously on both sides and are blocked with clenching of right fist.
Some posterior alpha rhythm persists after eye opening. Possible spread
of mu rhythm into the parietal area makes differentiation of alpha and
mu difficult in right parieto-occipital lead (eighth channel from the top).
B: The close-up view of the same patient shows independent bilateral
central mu rhythm.
for the assessment of cortical functioning, not only over the
motor cortex but also over various cortical regions of strong
afferent input related to neurocognitive activities. Further detail
on the ERD is found in the work of Pfurtscheller and Lopes da
Silva (120,121).
Age and Prevalence
Central mu rhythm used to be considered scarce. The introduc-
tion of the International Electrode System (10-20 system) has
contributed to a much greater awareness of this pattern. The C3
and C4 electrodes are located over the precentral gyrus in an
optimal location for picking up central mu rhythm. The use of
a different electrode system revealed a prevalence of only 3.2%
(122) and 2.9% (123), figures that lie well below those found in
other studies. Gastaut et al. (108) found mu rhythm in 10% of
their adult patients; Beek (124) observed this rhythm in 13% of a
predominantly psychiatric patient population. In 500 essentially
■ Normal EEG and Sleep: Adults and Elderly 189
healthy young male adults, Gastaut et al. (125,126) found mu
rhythm in 14.4%. Dongier and Dongier (127) noted mu
rhythm in 18% of a population of neurotic patients. Figures of
12% were reported by Picard et al. (128) and Simonova et al.
(129). In the patients of Niedermeyer and Koshino (130), the
prevalence of mu rhythm was 8.1% (182 of 2248); broken down
into age ranges, there were 9.0% between ages 0 and 10 years,
13.8% between 11 and 20 years, 8.4% between 21 and 40 years,
and 4.5% above 41 years. These authors demonstrated rolandic
mu rhythm in a 20-month-old child; this was thought to be an
exceptionally early manifestation of mu rhythm, although the
data of Stroganova et al. (131) demonstrated central mu
rhythm (during a state of attention) in the tracing of an 8-
month-old baby with a frequency of about 6 to 8.8/sec. The
authors speculated that mu rhythm tends to appear before the
occipital alpha-equivalent because, unlike visual stimulation,
somatosensory stimulation is present within the uterus. With
the use of frequency analysis, the prevalence of mu rhythm
reaches values close to 100% (132). Familial occurrence of mu
rhythm has been reported by Koshino and Isaki (133).
Morphology, Frequency, and Spatial Distribution
Older synonyms such as rhythme en arceau or wicket rhythm
pertain to the wave morphology; mu rhythm shows in most
instances a sharp (or spiky) negative and a rounded positive
phase. There are a few patterns with a monophasic spiky
appearance. Posterior alpha rhythm may have a similar config-
uration: “wicket spikes” have a similar appearance. The same is
true for 14 and 6/sec positive spikes, except for the spike posi-
tive and rounded negative component. All these patterns show
a spatial distribution that is different from the mu rhythm or
occurs at levels of vigilance that are usually incompatible with
mu rhythm. The amplitudes of mu rhythm are comparable to
those of the posterior alpha rhythm.
Mu rhythm usually occurs in short stretches. Persons with
pronounced mu rhythm show long trains of mu rhythm. The
most common frequency of the mu rhythm is 10/sec.
Frequencies may lie below 9/sec and above 11/sec; a mu rhythm
of less than 8/sec is probably a mild abnormality. Mu rhythm is
often mixed with local activity around 20/sec. According to
Storm van Leeuwen et al. (134), mu frequencies are slightly
higher than alpha frequencies. The spatial distribution is essen-
tially confined to the precentral–postcentral region; some spread
into parietal leads is not uncommon. The C3 and C4 electrodes
are mostly involved; occasionally, a vertex (Cz) maximum is
noted. A special vertex mu rhythm with special reactivity (lower
limb) was described in aged patients by Farnarier et al. (135).
When recorded through bone defects, the mu rhythm is more
widely distributed. According to Kuhlman (111), the maximum
lies over the postcentral rather than the precentral cortex.
The alphoid (around 10/sec) and the fast (around 20/sec)
component of the mu rhythm seem to be inseparably intermin-
gled but, with more advanced technology, spatial separation is
possible. According to Salmelin and Hari (136,137) the beta
component arises from the motor cortex and the alphoid com-
ponent from the sensory cortex. These magnetoencephalo-
graphic findings have been confirmed by Nashmi et al. (138).
190 Part II ■ Normal EEG
In most persons with mu rhythm, the activity is bilateral but
tends to shift from side to side. Coherence function studies have
shown lack of bilateral coherence of mu rhythm in normal sub-
jects (134). Strictly unilateral mu rhythm must be scrutinized
for the possibility of an ipsilateral rolandic disturbance as, for
instance, an early stage of parasagittal meningioma, an arteri-
ovenous malformation, or other types of neoplasms (139). In
such cases, even the possibility of a contralateral rolandic lesion
must be taken into consideration (107,140,141). Above all, the
possibility of a local cranial bone defect, surgical or traumatic,
must be ruled out; a single burr hole in the rolandic region
enables otherwise hidden mu rhythm to become manifest on
the scalp (Fig. 11.4). Thus, mu rhythm represents at least a por-
tion of rhythmical activity beneath bone defects (“breach
rhythm,” Ref. 142).
Reactivity
Mu rhythm is blocked by movements. These movements may
be active (voluntary), passive, or reflexive (106,107,143–145).
The blocking effect is bilateral but more pronounced on the
Figure 11.4 Patient aged 51, 2 years after evacuation of a right-sided
subdural hematoma. A right-sided central mu rhythm shows admixture
of fast frequencies and spiky potentials. All components of this rhythm
show a prominent and persistent blocking response to contralateral
hand movements. (From Niedermeyer E, Koshino Y. My-Rhythmus:
vorkommen und klinische bedeutung. Z EEG-EMG. 1975;6:69–78.)
rolandic region contralateral to the site of movement; the effect
appears prior to the onset of muscular contraction
(106,107,122,144,145). According to Chatrian et al. (145), there
are delays of 50 msec to 7.5 seconds (average, 1.5 seconds) at the
onset of the mu blocking effect after the initiation of the spon-
taneous flexion of the contralateral thumb, with the ipsilateral
lagging behind the contralateral response. Gastaut et al. (106)
and Ciganek (146) demonstrated a somatotopic distribution of
the blocking effect (according to the functional anatomy of the
rolandic cortex), but this observation could not be confirmed
by Chatrian et al. (145).
With the use of a gamma band (38 to 40/sec) frequency
analysis focusing on the event-related synchronization (ERS) of
the gamma component, Pfurtscheller et al. (119) beautifully
demonstrated the locus for right and left index finger move-
ment, right toes, and the rather broad and bilateral area for
tongue movements.
The mu blocking response is likely to be related to the con-
ceptual design of the movement to be executed. Mere thoughts
about performing movements and readiness to move block the
mu rhythm (106,107,122,143–145). Mu blocking responses
have also been demonstrated in persons with amputations of
extremities; these mental activities concern movement of a
phantom limb (122,147).
Light tactile stimuli also produced a mu rhythm blocking
effect that is most evident over the contralateral rolandic region
(106,107,109,143–145). Kuhlman (111) has placed even greater
emphasis on the sensory component and regards mu rhythm as
the “idling” of the cortical sensory region, in analogy to the
alpha rhythm as “idling” of the visual cortex and its vicinity. His
work was carried out with power spectral analysis. This tech-
nique shows that central mu rhythm is a more common phe-
nomenon than one would expect from the visual analysis of the
scalp EEG (also see the power spectra study of Ref. 112). These
authors also showed simultaneous posterior alpha desynchro-
nization (112,113). Coherence functions have also been used
for the demonstration of rolandic mu rhythm (132,134).
It has been noted that rolandic mu rhythm is enhanced dur-
ing intermittent photic stimulation (148) and pattern vision
(149). The latter observation would contradict the basic law of
mu rhythm as a rhythm of immobility, but would be consistent
with extraocular movement control being neuroanatomically
distinct from the rolandic motor cortex.
Cerebral Generators of the Mu Rhythm
Jasper and Penfield (150) found evidence of a strictly localized
beta (around 20/sec) over the human motor cortex with
electrocorticographic recording technique in the locally
anesthetized patient. This rhythm could be blocked like mu
rhythm with movement, especially contralateral movement,
and also with thinking about the execution of movement. This
rhythm is obviously the cortical equivalent of rolandic mu
rhythm on the scalp. The mu waves on the scalp often show
some notching or are imbedded in local fast activity, which is
also blocked by movement. Gastaut (107) confirmed the corti-
cal precentral fast activity in electrocorticographic tracings.
These data conflict with Kuhlman’s (111) observation of a
 Chapter 11
postcentral maximum of rolandic mu rhythm (scalp recordings
with power spectral analysis). On the other hand, Graf et al.
(151) demonstrated precentral 8 to 10/sec mu rhythm in the
electrocorticogram of an adult epileptic.
As mentioned earlier, mu rhythm around 10/sec on the scalp
and cortical rolandic 20/sec activity appear to be harmonically
related; such a relationship, however, does not exist between mu
rhythm and rolandic beta activity on the scalp according to
Pfurtscheller (114) who used power spectral analysis.
Quantitative EEG studies have also established a clean separat-
ing line between mu rhythm and posterior alpha rhythm (152).
A study of the functional significance of the mu rhythm
from the human cortex was carried out with subdural elec-
trodes by Arroyo et al. (153). This work also showed the rela-
tionship of the cortical mu rhythm to the well-known
somatotopic arrangement of the sensorimotor cortex.
Three historical theories have governed the theoretical
approach to the mu rhythm: (a) the hypothesis of “neuronal
hyperexcitability” (“hyperexcitabilité neuronique”), either dif-
fuse or locally restricted to the rolandic cortex
(106,107,110,124); (b) the hypothesis of superficial cortical
inhibition (“inhibition corticale superficielle”), comparable to
aforementioned theories of “cortical idling”; this theory would
explain the suppression of mu rhythm with motor activity
(154–156,147); and (c) the hypothesis of somatosensory “corti-
cal idling,” turning mu rhythm into an afference-dependent
phenomenon (111).
Conditioning, Feedback Training,
and Relationship to Vigilance
An impressive international study on mu rhythm and the pha-
sic evolution of conditioning was presented by Gastaut et al.
(157). Mu rhythm feedback training has been used for various
purposes, including the treatment of epileptic seizure disor-
ders (158–161). Roldan et al. (162) observed the development
of mu rhythm-like activity in the course of Hatha yoga exer-
cises. This activity was called chi rhythm. It was believed to be
a specific rolandic pattern of 11 to 17/sec waves and theta
frequencies. Covello et al. (163) felt that several entities of mu
rhythm exist, even in one subject, and that their underlying
physiologic mechanisms may differ. Personal impressions
lend some support to this idea. Earlier observations (130) led
to the conclusion that even the slightest degree of lessened
vigilance (i.e., earliest drowsiness) was incompatible with
mu rhythm; this view has been supported by Schoppenhorst
et al. (132). However, there have been some personal observa-
tions of patients demonstrating their best stretches of central
mu rhythm in the light drowsy state after posterior alpha
dropout. Yamada and Kooi (164) even described mu rhythm
in sleep (stages 1 and 2, as well as rapid eye movement [REM]
sleep) in 20 individuals (19 patients with various complaints).
The concept of Covello et al. (163) is certainly in need of fur-
ther study.
Clinical Significance
Pure rolandic mu rhythm as such is categorically within normal
limits. In some cases, however, spiky discharges may be present
■ Normal EEG and Sleep: Adults and Elderly 191
and even single spikes may stand out. Children with benign
rolandic epilepsy may gradually develop mu rhythm when
rolandic spike activity subsides and the seizures appear to be
under control. Interestingly, in some children with benign
rolandic epilepsy, cerebral spikes can be blocked by contralat-
eral movements in the same manner as mu rhythm is blocked
(165,166). This response is not present in all children with
rolandic spikes; this explains negative responses to motor activ-
ity reported by Isch-Treussard (167).
As with many other EEG patterns considered to be within
the range of normal, mu rhythm has been associated in the
historical literature with a predisposition to any number
of neurologic and non-neurologic pathologic conditions,
including headache, hypertension, hyperthyroid states, psy-
chiatric disorders, and even eczema (108,123,127,154,168), all
with varying degrees of correlation and plausibility. It is
known that essentially all healthy adolescents and adults
harbor central mu rhythm when power spectra and coherence
tests are utilized (132). Why then is mu rhythm evident in the
conventional EEG in just a minority of persons? It may be that
their mu rhythm is more powerful than that of average
persons in order to be picked up from the scalp. There is no
good evidence that this represents a clinically pathologic sign,
however.
In cases of strictly unilateral mu rhythm, a careful search for
bone defects (burr holes) and/or local pathology is indicated. In
the wake of ischemic hemiplegic insults, central mu rhythm
may disappear or its reactivity may vanish (169).
The “Third Rhythm” (Independent
Temporal Alphoid Rhythm)
Rhythmical activity in the alpha and upper theta range can be
picked up from epidural electrodes over the midtemporal
region (170–174). This rhythm is not detectable in the scalp
EEG unless there is a local bone defect in the midtemporal
region where it has been termed (together with coexisting cen-
tral mu rhythm in the vicinity) “breach rhythm” by Cobb et al.
(142), who regarded the temporal activity as an abnormal
rhythm. There is now good reason to presume that the rhyth-
mical temporal lobe activity, which is clearly independent from
the posterior alpha and rolandic mu rhythm, constitutes a
physiologic rhythm.
The function of this rhythm is still debatable, however. Its
presence as an independent temporal alpha rhythm has been
demonstrated with the use of magnetoencephalography in
healthy adults by Hari (personal communication, 1990, 1991),
who felt that this rhythm is strongly related to the cortical audi-
tory function (“auditory alpha rhythm”). Personal data
(170–173) do not provide good evidence for a relationship to
auditory function. In some individuals, cognitive activities
result in blocking responses of the third rhythm. Rhythmical
activity in the alpha and upper theta range over the anterior
temporal–midtemporal region in conventional scalp EEGs can
be found in patients with cerebrovascular problems (175) and
other types of temporal lobe changes, but this abnormal
rhythm can be differentiated from the physiologic “third
rhythm” (173).
192 Part II ■ Normal EEG
Beta Rhythm
The Greek letter beta stands for the frequency band above
13/sec. In customary EEG recording, frequencies above 70/sec
are considerably attenuated due to filter effects. This estab-
lishes a natural upper limit of discernible EEG activity some-
where between 50 and 100/sec. The upper limit of the beta
band is either considered open-ended or defined as 30/sec
(with gamma frequencies above). Rhythmical beta activity is
encountered chiefly over the frontal and central regions. A
central beta rhythm is related to the rolandic mu rhythm and
can be blocked by motor activity or tactile stimulation, as
described above. The amplitude of beta activity seldom
exceeds 30 µV. Beta activity may be locally enhanced over
bone defects and shows considerable increase (in quantity and
voltage) after the administration of barbiturates, some non-
barbituric sedatives, and minor tranquilizers. With very large
amounts and high amplitudes of beta activity, abnormal pro-
portions may be reached; in general, however, fast activity
alone is seldom a cogent reason for calling a tracing abnormal.
Historically, the discovery of beta activity is closely linked to
the first description of the alpha rhythm by Berger (1). His first
report on the human EEG of 1929 contains the following pas-
sage: “The electroencephalogram represents a continuous curve
with continuous oscillations in which … one can distinguish
larger first order waves with an average duration of 90 sigma
(msec) and smaller second order waves of an average duration
of 35 sigma (msec)” (after Ref. 41). In his second report, Berger
(176) pointed out: “For the sake of brevity I shall subsequently
designate the waves of first order as alpha waves, the waves of
second order as beta waves.” Berger (62) reversed earlier views
and pointed out “that the beta waves and not the alpha waves of
the EEG are the concomitant phenomena of mental activity.”
This concept is now provoking renewed interest.
Prevalence of Beta Activity
Beta activity is found in almost every healthy adult. Simon’s
(29) statement that beta frequencies have been found in 22% of
normal adults probably applies to more conspicuous forms of
beta activity. Depth EEG studies and electrocorticograms show
a remarkably large share of the beta band; such recordings have
been obtained mostly in epileptics or in patients with neuro-
logic or psychiatric problems. Although obtained in patients
and not in healthy persons, such depth records are quite
informative concerning the spatial distribution of fast frequen-
cies. Every scalp recording performed over a bone defect gives
testimony to the wealth of cortical beta activity.
The prevalence of beta activity in the normal person can also
be derived from various studies of normal volunteers
(30,177–185). The results of these studies have been reviewed
by Kuhlo (186) in his thorough EEG handbook article on beta
rhythms. Later work has been presented by Fortuin and Künkel
(187) and Kozelak and Pedley (188).
While Gibbs and Gibbs (42) found varying degrees of fast
activity to be abnormal, depending on age, electroencephalog-
raphers have shown a more lenient philosophy toward fast
tracings over the ensuing decades. Beta activity must be partic-
ularly abundant in quantity and of rather high voltage to be
termed abnormal. Even such fast records are often considered
only slightly abnormal unless they occur in unresponsive
patients, where such fast activity may represent a severe abnor-
mality. This view is essentially congruent with the study of
Drake (189).
Spatial Distribution
In certain individuals, the posterior alpha rhythm is unusually
fast and its frequency exceeds the upper limit of the alpha
rhythm (13/sec). Such a 14 to 15/sec rhythm shows a good
blocking response to eye opening and enhancement with eye
closure; it may be considered a fast equivalent of the alpha
rhythm. Gradual acceleration of the alpha frequency may be
caused by hyperthyroidism. Vogel and Götze (190) separated
the occipital 14 to 19/sec rhythm from the rest of beta activities,
which were divided into frontocentral bursts of 20 to 30/sec
waves, and normal occipital alpha rhythm and diffuse beta
activity generally mixed with alpha frequencies. Vogel
(191–193) subsequently divided the anterior beta activity into
the 25 to 30/sec activity of the frontal region, occurring in brief
trains, and the 20 to 25/sec activity of comparatively larger
amplitudes occurring for variable periods.
Frontal runs of beta activity often exceed 30/sec and reach 35
to 40/sec when the patient is allowed to fall asleep. This activity
may be found in unmedicated persons; sleep-inducing medica-
tion, which increases the activity in the 18 to 25/sec band, also
augments activities in the faster beta range.
The region of fastest EEG frequencies and highest levels of
cerebral blood flow (as well as oxygen and glucose uptake) is the
frontocentral area. This observation has prompted the concept
of “cerebral hyperfrontality” (194). This investigator has
pointed out that (a) prefrontal activity constitutes an integral
part of consciousness and (b) abnormal prefrontal activity is
accompanied by various forms of altered consciousness.
The physiologic beta frequencies may be broken down as
follows:
1. Frontal beta: fairly common, may be very fast, no relation-
ship to physiologic rhythm.
2. Central beta: partly but not generally the basis of rolandic
mu rhythm often mixed with mu rhythm.
3. Posterior beta: often a fast alpha-equivalent, reactive-like
alpha rhythm.
4. Diffuse beta: no linkage with any special physiologic rhythm.
By contrast, “vertex beta” or “fast central midline rhythm”
(195)—vertex activity in the 10 to 25/sec range—appears to be
the expression of polyetiologic focal abnormality in the fronto-
central region.
Other Influences on Beta Activity
One of the most commonly encountered influences on the
abundance of beta activity in clinical practice is the use of cen-
tral nervous system (CNS)–acting medications, particularly
benzodiazepines and barbiturates. This topic is discussed in
more detail elsewhere in this volume.
 Chapter 11 ■ Normal EEG and Sleep: Adults and Elderly 193

It is not surprising that beta activity in its various incar nations slightly contaminated with EEG activity. The definition of
has, over the decades, been associated with various personality Chatrian et al. (18) reads as follows:
traits and psychiatric disorders (42,177,178,183,196–199), but
Low voltage record. A waking record characterized by activity of
rarely with any degree of clinical or statistical certitude (200).
amplitudes not greater than 20 µV over all head regions.
During the entire maturational period (including early
With appropriate instrumental sensitivities this activity can
adulthood), beta activity has been found to be relatively more
be shown to be composed primarily of beta, theta, and, to a
prominent in females than in males (201).
lesser degree, delta waves, with or without alpha activity,
Ultrafast and Ultraslow EEG Activity over the posterior areas.
This definition serves as a solid guideline. One could subdivide
Interest in normal EEG activity within the ultrafast range from
low-voltage records into two or three degrees; occasionally, one
80 to 1000/sec has been surging over the past two decades. The
may see tracings in which 10 µV are not exceeded. Even such
advent of digital EEG recording has flung open the gate to this
“very low voltage records” may be due to desynchronization
new domain of promising work. The study of ultrafast activity
and thus occur in a waking and alert patient or even in a con-
can offer new epileptologic insights (202–205) but also, above
trol subject (Fig. 11.5). Low-voltage records are categorically
all, improved understanding of normal cortical perception,
within broad normal limits of variability and do not represent
motor activity, and, in particular, neurocognitive processes
an abnormality unless the frequency spectrum shows abnor-
(206–209). Similarly, there has been renewed interest in ultra-
mal local or diffuse slowing, asymmetries, or paroxysmal
slow activity, from near zero (DC) up to 0.3/sec. Again, while
events. In a comatose patient, one must infer that the low volt-
activity in this range may improve focal epilepsy diagnosis
age is due to true decline of cerebral activity and not merely
(210), there are also insights to be gained in neurocognitive sci-
caused by desynchronization. Such records must be regarded
ence, motor initiation, and sleep (211).
as extremely abnormal. This indicates that the record reader
The Low -Voltage Record should depend on behavioral observation (patient comatose,
on respirator, etc.). On the other hand, the EEG as such reveals
A below-average voltage output is compatible with perfect CNS
signs of telltale character for the experienced reader, such as
functioning. In such individuals, the small amplitudes are the
presence or absence of eye blinks, types of artifacts, etc., even
result of a lesser degree of synchronization of electrical activity
when he or she must read without any behavioral information.
in the neuronal level. Brainstem reticular formation systems
strongly influence the degree of cortical neuronal synchroniza-
tion; arousal mediated through the ascending mesodiencephalic Prevalence and Clinical Significance
reticular formation has a desynchronizing effect (212), whereas Early studies of low-voltage tracings were carried out by Davis
synchronizing effects are presumed to originate from the pon- and Davis (50), Jasper et al. (217), and Finley (177). Jasper et al.
tine portion of the ascending reticular formation (213,214). (217) used the term flat EEG, which is not reserved for truly iso-
This type of brainstem regulation of the cortical EEG syn- electric records.
chronization is likely to account for numerous fluctuations in
the EEG amplitudes, especially those related to the level of
vigilance. It is not clear whether these reticular effects account
for a low-voltage EEG as a personal trait (with genetic deter-
mination). Under pathologic conditions, low-voltage tracings
in vertebrobasilar artery insufficiency are probably due to
some degree of pontine ischemia acting on the reticular for-
mation. This could also be true for the frequently encountered
low-voltage tracings of chronic alcoholics in advanced stages,
with the possibility of early stages of pontine myelinolysis;
this explanation, however, is but a working hypothesis (215).
True decline of the cerebral EEG voltage output (i.e., low-volt-
age activity independent of desynchronizing mechanisms) is a
grave danger sign of a preterminal state unless it occurs for a
few seconds only, as in certain syncopal attacks. It may also
indicate diffuse chronic cortical degeneration (as described by
Ref. 216).

Definition
The term low-voltage EEG has always proved to be quite
flexible; it is in need of a strict definition. It was pointed out
earlier in this chapter that precise figures concerning ampli- Figure 11.5 Low-voltage record with poor alpha organization, in a
tudes are not truly accurate, because they depend on interelec- patient aged 31 years. Note that recording was done with above-stan-
trode distances and the montage; even an ear electrode may be dard gain. Finer details demonstrable with further gain increase.
194 Part II ■ Normal EEG
Gibbs et al. (179) (also see Refs. 42 and 218) found low-volt-
age fast records in 11.6% of 1000 normal adult subjects. The
study of Adams (219) revealed a prevalence of 8% (in 427 nor-
mal individuals); the breakdown into age ranges is most impor-
tant, because low-voltage tracings were found in 1% between
ages 0 and 20 years, 7% between 20 and 39 years, and 11%
between 40 and 69 years. Pine and Pine (220) reported 7.25%
low-voltage records in 2000 neurologic and psychiatric patients
between 17 and 70 years. Vogel and Götze (221) found low-
voltage tracings in 7.09% of 1540 patients. In 7000 patients
including a large pediatric segment, only 3.6% were found to
have a low-voltage tracing (222).
According to Gibbs and Gibbs (42), the prevalence of low-
voltage tracings increases sharply after the age of 13 years. This
is congruent with the observations of Panzani and Turner
(223), Adams (219), and Lucioni and Penati (224). Low-voltage
fast records in children below the age of 10 years are suspect and
clearly abnormal if neither hyperventilation nor non-REM
(NREM) sleep changes the low-voltage character of the tracing.
In adults, persistence of low-voltage activity with hyperventila-
tion and sleep is not categorically abnormal, although this lends
more support to the presumption that such patients have an
organic cerebral disorder. Among neurologic disorders that
have been associated with a higher prevalence of low-voltage
EEG activity are traumatic brain injury (225), basal ganglia dis-
orders (216), and endocrine and psychiatric disorders (219).
There is some evidence that genetic factors play an impor-
tant role in determining a low-voltage tracing in a healthy per-
son (221). This genetic predisposition, however, obviously does
not manifest itself in infancy or childhood, a time when almost
all low-voltage records bear the mark of abnormality. With ado-
lescence (age 13, according to Gibbs and Gibbs Ref. 42), an
unknown maturational change apparently starts to desynchro-
nize the cortical EEG activity.
Theta Rhythm
The term theta was introduced by Walter and Dovey (226) and
denotes the frequency range of 4 to 7/sec or 4 to 7.5/sec. This
frequency band was a part of the delta range until Walter and
Dovey felt that an intermediate band should be established. The
term theta was chosen in order to allude to its presumed thala-
mic origin. The intermediate character of the theta band has
been stressed in the German terminology Zwischenwellen,
meaning intermediate waves. According to the international
nomenclature, the theta band is the “frequency band from 4
to under 8 Hz” and the theta rhythm is the “rhythm with a
frequency of 4 to under 8 Hz” (18).
Theta Activity in the Waking Adult
The normal adult waking record contains but a small amount
of theta frequencies and no organized theta rhythm. Theta fre-
quencies and theta rhythms, however, play an important role in
infancy and childhood, as well as in states of drowsiness and
sleep (all discussed elsewhere in this volume). Larger contin-
gents of theta activity in the waking adult are abnormal and are
caused by various forms of pathology. Studies of young adults
such as army personnel and navy pilots have shown a sizable
amount of theta activity (178,183) because these slow frequen-
cies tend to linger on through the third decade of life; the com-
pletely mature aspect of the human EEG cannot be expected
before the age of 25 to 30 years.
On the basis of power spectra from normal adults, Rugg and
Dickens (227) presume that alpha and theta activity is gener-
ated by separate mechanisms. Slow (theta and delta) EEG activ-
ity has been correlated with cholinergic activities and central
cholinergic pathways (103). The observation of slowing
induced by the action of cholinergic substances on the ascend-
ing brainstem reticular formation dates back to Rinaldi and
Himwich (228). Further information is found in the overview
of Riekkinen et al. (229).
Spatially Restricted Theta Rhythms
With the use of a stress paradigm, Nowak and Marczynski (61)
produced in six of 24 healthy volunteers a rhythmical 5 to 6/sec
theta response originating from either the occipital region or
vertex during a state of maximal alertness.
A rhythmical 4/sec pattern occurring over the vertex solely
in the waking state has been described as “4/sec vertex spindles”
by Van Huffelen and Magnus (230). This very rare pattern may
represent a mild abnormality and is found mainly in adoles-
cents with syncopal attacks and other signs of vasomotor insta-
bility. The use of the term spindles appears to be somewhat out
of place when one considers the slow frequency and the occur-
rence in the waking state. Hence, Daoust-Roy (231) proposed
the term 4/sec vertex rhythm.
Rhythmical theta activity in the 6 to 7/sec range over the
frontal midline region has been correlated with mental activities
such as problem solving (232–238). In personal studies of men-
tal EEG activation, we were unable to produce this pattern (44);
this indicates the technical difficulties in the demonstration of
the task-related frontal theta rhythm. This vexing problem has
been clarified by the work of Takahashi et al. (239), who found
that individuals with rhythmical frontal midline 6 to 7/sec activ-
ity in light drowsiness also had the same type of activity during
mental tasks: “These two frontal theta rhythms closely resembled
each other in frequency (94.6%) and distribution (83.8%)”
(237). One wonders if these tasks become boring enough to pro-
duce light drowsiness rather than increased alertness (one is even
tempted to utilize the old Pavlovian term internal inhibition, i.e.,
sleep induction in a certain repetitious experimental setting).
Lambda Waves
Definition and Historical Aspects
Lambda waves are “sharp transients occurring over the occipi-
tal region of the head of waking subjects during visual explo-
ration. Mainly positive relative to other areas. Timelocked to
saccadic eye movements. Amplitude varies, but is generally
below 50 µV” (18).
This visually induced occipital activity was described by
Gastaut (240); there was an earlier oral presentation by Evans
(EEG Society in London, 1949). Subsequent EEG studies on
lambda waves were presented by Cobb and Pampiglione (241),
Evans (242,243), and Roth and Green (244). A very extensive
 Chapter 11 ■ Normal EEG and Sleep: Adults and Elderly 195

review of this subject is found in the handbook article of

Chatrian (245).

Prevalence
Lambda waves are unmistakably present in some records (Figs.
11.6 and 11.7) and are not readily demonstrable in others. They
are most prominent in waking patients intently viewing an illu-
minated visual field (245). The prevalence essentially depends
on the thoroughness of the EEG evaluation and the emphasis
placed on the demonstration of this special phenomenon.
According to Tsai and Liu (1965; quoted in Ref. 245), lambda
waves are most frequent between ages 3 and 12 years (82.3%);
their prevalence declines to 72% between 18 and 30 years and
to 36.4% between 31 and 50 years.

Morphology
Lambda waves have been described as biphasic or triphasic; the
most prominent phase is positive. Their form has been
described as triangular or sawtooth shaped. The amplitude is
usually below 20 µV and may exceed 50 µV in some persons.
The overall duration of lambda waves lies between 200 and 300
msec. These waves repeat themselves, usually at intervals from
200 to 500 msec. Marton et al. (246) have demonstrated the
complexities of lambda waves; their predominant positive com-
ponent is preceded and followed by a negative component.

Spatial Distribution
Lambda waves are most prominent in (or confined to) the
occipital leads. Spread into parietal and posterotemporal areas
is common and, in certain cases, the maximum may be found
in areas adjacent to the occipital lobe (243,244). Lambda activ-
ity is strictly bilateral synchronous. With the use of depth elec-
trodes, Perez-Borja et al. (98) demonstrated multiple foci of
lambda waves, either in or near the calcarine region or more
laterally in the occipital lobes.
Figure 11.7 Prominent occipital lambda waves during eye opening.
Patient, aged 54 years, looking ahead in the dimly lit laboratory. There
is no evidence of scanning ocular artifacts. Note instability and fast
character of alpha rhythm (1 to 13/ sec and even faster). The bottom part
shows continuation of the upper part.
Precipitating Factors
Voluntary scanning eye movements (exploratory saccades) play
a very important role. It was found that most lambda waves
follow an exploratory eye movement with a latency of 67 to 85
msec (mean, 78 msec) (247). The preceding eye movement has
been subsequently used as a trigger for computer techniques
(143,248–253). Further interesting details on this subject are
found in the handbook article of Chatrian (245).
Lambda waves are best found in brightly lit laboratories and
cannot be elicited in darkness. The size of a given pattern and its
distance and color are further variables. Binocular viewing of a
picture may or may not produce larger lambda waves than
monocular viewing (254,255). Presentation of a new picture
enhances lambda production (245,256). Lambda activity does not
seem to be related to recognition of objects (247). Hypnotic sug-
gestion of a test picture failed to produce a lambda response (255).

Relationships to Alpha Rhythm

Figure 11.6 Occipital lambda waves while viewing pictures, in a patient
aged 15 years. Note ocular artifacts caused by scanning eye movements.
and Visual Evoked Responses
There appears to be considerable independence between poste-
rior alpha rhythm (with eyes closed) and lambda waves (with
eyes open); no interaction or complementary action of these
two activities has been found. Marked similarities between
lambda waves and visual evoked responses have been reported
196 Part II ■ Normal EEG
by Remond et al. (252) and Lesèvre (257). Relationships
between lambda waves and occipital photic driving were origi-
nally presumed by Gastaut (240) and Evans (243) but could not
be confirmed by Chatrian et al. (258), Perez-Borja et al. (98),
and Scott et al. (255).
Pathologic Significance
No correlation between lambda waves and any neurologic or
psychiatric disorder has been demonstrated.
Theories on the Basic Neurophysiologic Mechanism
Lambda waves have been thought to be related to (a) visual
evoked responses, (b) oculomotor–visual integration, (c) ocu-
lomotor potentials, or (d) arousal mechanisms. These theories
have been thoroughly discussed by Chatrian (245).
Billings (259) has proposed that there are two different
occipital lambda waves in the human. The first one is the result
of the release of peripheral visual inhibition during the braking
phase of the primary saccade. This wave is transmitted via the
faster conducting Y-type fibers of the optic nerve (subserving
peripheral vision). The second wave is elicited by the return of
normal central vision, during or slightly before the braking
phase of the secondary corrective saccade. This wave is trans-
mitted via the slower conducting X-type fibers of the optic
nerve (subserving central vision). According to Billings,
“lambda waves, long considered to be of rather trivial import by
clinical electroencephalographers, are of considerable func-
tional significance in the visual system.”
THE RECORDING OF EEG DURING SLEEP
Sleep EEG recordings have been performed in regular EEG lab-
oratories for at least 60 years in order to secure information not
available in the usual waking–resting record with such activa-
tions as hyperventilation and intermittent photic stimulation.
The purpose of these EEG sleep studies is to search for abnor-
malities that may be hidden in the waking state. The more this
type of combined waking and sleep recording is carried out in
the regular EEG laboratory, the greater will be the insight into
the sleep patterns of normal individuals.
Nevertheless, the term sleep EEG may be somewhat pre-
sumptuous when one considers that only a short segment of
sleep can be recorded under daytime recording circumstances.
Furthermore, the recorded sleep is basically a nap, either sponta-
neous or induced by a sedative, and is thus different from noctur-
nal sleep. In the regular EEG laboratory, one can observe the
process of falling asleep during the daytime. The ensuing sleep is
relatively short and, for the most practical purposes, a duration of
10 to 30 minutes will suffice. The result is a tracing that shows the
transition from the waking state to drowsiness and from drowsi-
ness (stage 1) to light NREM sleep (stage 2). These stages of falling
asleep and light sleep are usually the most informative stages for
the electroencephalographer who is searching for clinically rele-
vant abnormalities and especially for epileptic discharges not
demonstrable in the waking state. In certain cases, deeper sleep
(stage 3 or possible stage 4) may yield additional information;
this may require a more prolonged sleep recording. REM stages
usually emerge about 60 to 90 minutes after sleep onset; this is
closer to 90 minutes in sedated sleep. This necessitates unusually
prolonged recording and, in the hustle of a regular laboratory,
blocks one of the recording units for an unduly long period.
Here the need for nocturnal sleep recording becomes obvi-
ous. There are now a growing number of specialized nocturnal
sleep laboratories that offer 24-hour recordings of a given
patient. These laboratories have concentrated on (a) nocturnal
sleep research in healthy volunteers, (b) the exploration of the
sequential nocturnal sleep phases in patients with sleep disor-
ders, and (c) the study of poorly understood (and possible
epileptic) attacks of strictly nocturnal occurrence.
Specialized sleep laboratories are polygraphically oriented
and use EEG in combination with a variety of other physiologic
parameters. It is not unusual to find only two channels reserved
for EEG recording. This polygraphic approach has made enor-
mous progress over the past decades but will not be discussed
further in this chapter.
The scoring of sleep stages (as it is being done in sleep labo-
ratories) is essentially based on the rules laid down by
Rechtschaffen and Kales (260). These rules have been strongly
criticized and major changes in the nomenclature are being
introduced and are becoming more widely accepted (see the dis-
cussion on polysomnography in Chapter 41). Kubicki and his
coworkers have been the most consistent proponents of criticism
(261–263). Most of this criticism has been directed toward (a)
lack of insufficient EEG information (limitation to central lead)
and (b) misleading delta activity criteria resulting in erroneous
scoring. This is particularly true with stage 1 and 2 segments
occurring within REM sleep and falsely scored as REM sleep.
The emphasis on sleep sections in routine laboratory work
varies greatly and depends on the influence of major centers
and schools. Most major American schools of EEG traditionally
use sleep in routine EEG work to a great extent. Some laborato-
ries try to attempt sleep in every or almost every patient at any
age. In this author’s laboratory, the proportion of sleep records
was consistently around 60%. Debates about the usefulness of
sleep can be heated; many electroencephalographers use sleep
only as a method to obtain a tracing in uncooperative and oth-
erwise unmanageable children. The informative value of short
sleep recordings has been stressed by Gastaut et al. (264,265)
and Niedermeyer and Höller (266).
There is no doubt that the overall information derived from
a wakefulness record is greater and hence clinically more signif-
icant. There are, however, a number of conditions, especially in
the domain of epileptic seizure disorders, in which sleep pro-
vides essential information. (This excludes the sleep disorders
themselves, which ought to be evaluated in a specialized sleep
laboratory.) For the sake of this important minority, sleep por-
tions should be secured in most patients.
The drowsy state and light sleep (stage 2) are usually the most
informative phases; for this reason, the sleep-onset portion with
a length of 5 to 30 minutes in stage 2 sleep may be sufficient. In
infants and children, a sleep section (nap recording) is almost a
necessity, not simply because of easier management of the sleep-
ing child but also because of the needed information. In infants
and in small and preschool children, the EEG recording often
 Chapter 11
cannot start until the child has fallen asleep. In these cases, great
efforts must be made to secure a readable waking portion after
awakening.
It can be understood readily that pad electrodes pressed
against the scalp by caps or bands are counterproductive in a
sleep laboratory. The slight discomfort caused by such elec-
trodes is enough to keep the patient from falling asleep. For this
reason, electrodes suitable for sleep recording ought to be fixed
to the head by an adhesive.
In some laboratories, the technologist and the apparatus are
placed in the same room where the patient lies. In the absence
of video monitoring, this setup has advantages such as closer
observation, especially in patients with clinical seizures, over
the system of having patient and machine in separate rooms
with a window permitting observation of the patient. With this
latter setup, the technician can more effectively request help
from colleagues or the medical director. Beds or carts are supe-
rior to reclining chairs.
SPONTANEOUS SLEEP, SLEEP DEPRIVATION,
AND SEDATED SLEEP
Spontaneous Sleep
Sleep without medication is undoubtedly preferable to sedated
sleep. In an appropriately darkened room, many children and
adults fall asleep quite readily. Outpatients from distant areas
and those who have experienced strenuous travel in the early
morning hours also are likely to sleep without medication.
Sleep Deprivation
Sleep deprivation prior to recording has been suggested over the
past decade by a number of electroencephalographers. This
method has been described by Bennett (267), Bennett et
al. (268), Mattson et al. (269), Oller-Daurella (270), Pratt et
al. (271), Wittenbecker and Kubicki (272), Rumpl et al. (273),
Deisenhammer and Klingler (274), Degen (275), Jovanovic
(276), Kubicki et al. (277), and Klingler et al. (278). It consists of
sleep loss for 24 to 36 hours. Its goal is the detection of epileptic
discharges that could be missed otherwise. According to the view
of Pratt et al. (271), a good deal of the activation of paroxysmal
patterns may be ascribable to drowsiness and sleep as such, but
some other factors specific for sleep deprivation have proved to
be potent activators. In the research of Pratt et al., activation of
epileptic patterns was obtained in 41% of 114 patients.
Sleep deprivation may have drawbacks. Many individuals
and especially children may be so fatigued that they fall asleep
almost instantly in the laboratory with rapid evolution of stages
3 and 4; with the possible exception of temporal lobe epilepsy,
these are much less informative electroencephalographically
than stage 2. This is especially true in cases of generalized
epilepsies (primary generalized epilepsy, Lennox–Gastaut syn-
drome) and benign focal spikes in children.
Naitoh and Dement (279) have summarized the current sta-
tus of sleep deprivation as follows:
1. One night of sleep loss is sufficient. Sleep loss can be per-
formed at home with the help of a family member.
■ Normal EEG and Sleep: Adults and Elderly 197
2. Anticonvulsant medication does not have to be discontinued.
3. Use sleep deprivation with other techniques of activation
(see also Ref. 277).
4. Patients must be kept awake during the test period.
5. Expect very few false-positive cases but many false-negative
cases.
6. Sleep deprivation is a genuine activation method. Its effi-
cacy in provoking abnormal EEG discharges is not due to
drowsiness.
The insistence on keeping the patient awake during the
recording may be debatable and its value should be thoroughly
assessed. There is reason to presume that bilateral synchronous
seizure discharges are triggered by short fluctuations of the
level of awareness (light drowsiness followed by an arousing
stimulus; see Ref. 165); for this reason, the significance of short
and minor drowsy periods should be emphasized rather than
minimized.
As to the search for EEG abnormalities, a differentiation
should be made between patients with primary generalized
epilepsy and temporal lobe epilepsy. In the former, the strict
routine suggested by Naitoh and Dement (279) might be highly
effective; these patients should by all means be kept awake dur-
ing recording (also see Ref. 275). In temporal lobe epilepsies,
the ensuing deep sleep (stage 3) could be more informative.
Rodin et al. (280) have extended the period of sleep loss in
normal young volunteers to 120 hours of total sleep depriva-
tion. There was evidence of frequent generalized bursts similar
to those seen in primary generalized epilepsy. These changes,
however, could not be confirmed by other investigators
(281–284), who used sleep deprivation for periods as long as
264 hours. Welch and Stevens (285) also felt that sleep depriva-
tion did not activate the EEG of healthy subjects.
The suggestion of Gibbs and Gibbs (218) to keep the
patients awake until after midnight is helpful as a means to
obtain sleep more readily the following morning in the labora-
tory. However, this widely used practice almost completely
dilutes the basic principle of sleep deprivation. In practice, good
sleep recordings in stage 2 are only slightly less informative than
tracings obtained after sleep deprivation for 24 hours.
Sleep deprivation is a true activator, it is a specific stress
imposed on the CNS to activate otherwise hidden epileptic
activity. Sleep recordings in general are done in a physiologic
state and should not be listed as activation. The sleep depriva-
tion stress can be augmented by the effect of anticonvulsants
with sedative effect (phenobarbital or primidone) against
which the patient must fight. According to Klass and Fischer-
Williams (286), the benefits derived from sleep deprivation out-
weigh the risks of clinical seizures. These authors, however,
point out that in sleep-deprived persons photic stimulation
requires greater caution. It appears that the emphasis placed on
sleep deprivation in most clinical neurophysiology laboratories
has declined over the past two decades.
Sedated Sleep
Whenever a sleep recording is desirable and cannot be obtained
naturally, sedated sleep is dictated by necessity. The most
198 Part II ■ Normal EEG
Tabl e 11 . 1
Guidelines for Sedation to Obtain
EEG Sleep Records
Recommended sedative: chloral hydrate (oral); chloral
hydrate syrup U.S.P.
A. Adult and adolescents above 15 years
Average weight (56–80 kg): 1.75–2.5 g
Below average weight (40–55 kg): 1.25–2 g
Above average weight (above 80 kg): 2.5–3.5 g
B. Children (5–14 years)
1.0–2.0 g (depending on weight)
C. Infants and small children (4 months to 4 years)
Below 1/2 year: 0.125–0.25 g
1/2–1 year: 0.25–0.5 g (max 0.75 g)
1–4 years: 0.5–1.0 g (max 1.5 g if close to 4 years and above
average weight)
Advantages of chloral hydrate: induced fast EEG activity less prominent than
with barbiturates and rapidly metabolized. Side effects (paradoxical excitation
with irrational behavior) rare.
important advice to a novice in this technique is patience; the
patient must not be overpowered by strong and rapidly effective
sedatives. Therefore, the intravenous and intramuscular routes
must be avoided; with these the patient’s sleep becomes very
unnatural and contains elements of an early state of general
anesthesia. The highly informative stages of drowsiness (stage 1)
and light sleep (stage 2) are quickly passed, and the ensuing deep
sleep yields very little information in most cases. Enormous
amounts of fast activity may also obscure important details.
Orally or rectally administered sedatives act more gently and
slowly. The ensuing sleep contains many or most features of
spontaneous sleep. Most laboratories have the greatest experi-
ence with chloral hydrate (Table 11-1). The oral route is stan-
dard; rectal administration is usually confined to infants
reluctant to take the medication by mouth.
Terminology of Sleep Stages
The discovery of human REM sleep has been attributed to
Aserinsky and Kleitman (287), who were able to demonstrate reg-
ularly recurrent periods of altered ocular motility during sleep.
Further insight into the significance of the REM stages prompted
a new terminology of sleep stages that emphasized the dichotomy
of two basically different types of sleep: non-REM (“slow sleep”)
and REM sleep (also “paradoxical sleep” or “fast sleep”).
This dualistic approach subsequently proved to be appropri-
ate because of impressive differences of neurobiochemistry and
neuronal circuitry (288–292).
A commonly used classification (essentially following that of
Dement and Kleitman Ref. 293) divides sleep into the following
stages:
1. Drowsiness: from alpha dropout to vertex waves
2. Light sleep: spindles, vertex waves, K complexes
3. Deep sleep: much slowing, K complexes, some spindles
4. Very deep sleep: much slowing, some K complexes
REM sleep: desynchronization with faster frequencies
The transition from stage to stage may be somewhat imprecise,
but in general, staging of sleep periods is not difficult for elec-
troencephalographers who are used to reading records in both
the waking state and sleep.
In the clinical EEG laboratory, most sleepers reach only stage 2,
stage 3 is reached occasionally, and stage 4 seldom reached
unless unusually long sleep recording time is allotted or in cases
of profound tiredness after sleep loss. REM stages are seldom
seen in the clinical laboratory except in infants, young children,
and adults with sleep loss. Their occurrence at sleep onset may
denote a primary sleep disorder (especially a narcolepsy–cata-
plexy complex), but REM sleep onset may also occur in patients
with severely disturbed sleep due to central disturbances such
as delirium tremens or in deep-seated cerebral pathology.
Reviews of the EEG characteristics (294) and the neurobio-
chemical basis of sleep (295) are recommended sources of fur-
ther information. The current classification scheme has been
slightly modified and is outlined in Chapter 41.
THE “MICROSTRUCTURE” OF SLEEP:
CYCLIC ALTERNATING PATTERNS
Sleep is not solely determined by macrostructural shifts of
NREM and REM sleep stages. Evidence of a “microstructure” of
sleep has been provided over the past 20 years. Transient activa-
tion phases (“phases transitoires”) were reported by Schieber et
al. (296). “Cyclic alternating patterns” of sleep were observed by
Terzano and his coworkers, first under severely abnormal condi-
tions such as Creutzfeldt–Jakob disease (297), later in normal
individuals (298,299). Cyclic alternating patterns (CAPs) occur
mainly in NREM sleep, are generally associated with transient
lightening of sleep depth, and are independent of afferent stim-
uli, thus constituting modality-independent responses (299,300).
The cyclic alternating pattern consists of an A phase of enhance-
ment of electrical activity and a subsequent B phase characterized
by attenuation of EEG activity. Each phase lasts only between
2 and 60 seconds.
A somewhat different type of periodic activity of light sleep
has been demonstrated by Evans (301), who observed alternat-
ing periods of alpha and theta activity with a usual interval of
about 16 seconds. This activity is thought to be related to inter-
mittent arousing mechanisms involving cortex and brainstem
activating systems (301).
THE STATE OF DROWSINESS
Early Drowsiness
The level of vigilance tends to fluctuate, and fleeting brief
episodes of very light drowsiness are very common in waking
records. Boredom, fatigue, and monotony may quickly induce
 Chapter 11
such periods in a patient instructed to relax during the test.
Drowsiness is a very interesting state from the electroen-
cephalographer’s viewpoint. It shows particularly marked age-
determined changes; the “hypnagogic” rhythmical 4 to 6/sec
theta activity of late infancy and early childhood represents a
unique feature of that period of life. Later in childhood and, in
many cases, in the declining years of life, the onset of
drowsiness is characterized by greater amounts of slow activity
mingling with the posterior alpha rhythm. An electroen-
cephalographically defined drowsy state does not exist in the
neonate and, in senility, abrupt transitions from wakefulness to
light sleep are again quite common.
In the adult, the onset of drowsiness is characterized by grad-
ual or brisk “alpha dropout” (Fig. 11.8). The alpha waves are
replaced by low-voltage slow activity, mainly in the range of 2 to
7/sec. Very low-voltage 15 to 25/sec activity may be mixed with
the slowing. Thus, the tracing appears to be desynchronized due
to the general voltage decline. This state must be strictly sepa-
rated from periods of enhanced alertness after eye opening or
caused by mental or emotional stress (associated with alpha
blocking). In enhanced alertness, there is also general desyn-
chronization with low-voltage activity, but the slow frequency
component is absent and beta frequency clearly predominates,
mixed with some remnants of alpha. In patients with low-volt-
age fast waking records, the onset of early drowsiness may be
quite poorly defined. Alpha dropout of early drowsiness and
immediate slowing in adulthood is a suspect finding (Fig. 11.9).
Figure 11.8 Alpha dropout with early drowsiness in normal adult.
■ Normal EEG and Sleep: Adults and Elderly 199
Figure 11.9 Early drowsiness in a 56-year-old patient with dizziness
and memory loss. Rhythmical 5 to 7/ sec activity is noted over posterior
regions very shortly after the disappearance of alpha waves. This rhyth-
mical theta activity in earliest drowsiness is not categorically abnormal;
however, incipient cerebrovascular disorder is suspected.
Deepening of drowsiness is associated with enhancement of
slow activity. Trains of 2 to 3/sec and 4 to 7/sec waves reach
medium voltage and may become diffusely predominant. At
this stage, arousing stimuli lead to immediate return of the pos-
terior alpha rhythm, called “paradoxical alpha response” (302).
These stretches of reactivated alpha rhythm are often character-
ized by higher amplitude than the individual’s regular alpha
rhythm. When alpha rhythm is reactivated in deep drowsiness
or in NREM sleep, the maximum is usually frontal rather that
occipital (“short microarousals”).
If a record is dominated by patterns of drowsiness and sleep
and if waking states can be maintained for short periods only,
then there is reason to presume that such a recording is not
within normal limits. In such cases, there is usually a large
amount of diffuse slow (2 to 6/sec) activity preponderant dur-
ing drowsiness and even in early drowsy states. In these individ-
uals, the sequential structure of waking and sleeping is likely to
be disturbed. Cerebrovascular and metabolic problems often
account for such changes. Sleep deprivation must be ruled out
in these cases, but in general healthy sleep-deprived adults are
unlikely to exhibit excessive slowing in states of drowsiness.
A major study of the EEG in drowsiness has been presented by
Santamaria and Chiappa (303). Their work has placed special
emphasis on changes in the alpha rhythm with occurrence of a
centrofrontal alpha and temporal alpha (change in alpha distri-
bution due to early drowsiness) as well as changes in alpha ampli-
tude (diminution as well as increase). These changes of alpha
distribution in early drowsiness do not seem to be the rule. As
Broughton and Hasan (304) have pointed out, only a small por-
tion of their subjects showed anterior diffusion of alpha rhythm
(along with slowing by 0.5 to 1.5/sec) during drowsiness.
200 Part II ■ Normal EEG
Figure 11.10 Vertex wave in deep drowsiness. Enhanced fast activity is
due to medication (chloral hydrate).
In light drowsiness, the P300 response (standard auditory
paradigm) increases in latency and decreases in amplitude “but
the counts of infrequent tones remained correct nevertheless”
(305). This important study shed some light on mental func-
tioning in light drowsiness. On the other hand, EEG coherence
studies show altered inter- and intrahemispheric EEG coher-
ence in light drowsiness (306).
Deep Drowsiness
Vertex Waves
The hallmark of deep drowsiness is the appearance of vertex
waves (Fig. 11.10). These waves indicate an altered state of cere-
bral responsivity; vertex waves are secondary evoked potentials
of several modalities, perhaps mostly auditory, that converge
from their cortical projection areas in regions underlying the
vertex electrode. This region constitutes the posterior portion
of the supplementary motor area of the frontal lobe along the
interhemispheric fissure. The vertex waves represent an evoked
event, like K complexes, discussed later.
The term vertex wave was introduced by Liberson (307). On
effect (308), an earlier term, is no longer used. French authors
have used the term vertex spikes (“pointes au vertex”)
(309,310). The catalog of EEG patterns in the Handbook of
Electroencephalography and Clinical Neurophysiology uses the
term vertex sharp wave (3). Other synonyms are v waves and
vertex sharp transients. Surprisingly, Dutertre (3) does not list
the term biparietal hump (42) as a synonym of vertex waves and
discusses this term among “slow waves.” There is no doubt,
however, that biparietal humps truly denote vertex waves. In the
typical referential montages of Gibbs and Gibbs (42,218),
vertex waves appear most strikingly in parietal leads, because
no midline and rolandic electrodes are placed.
The vertex wave is a compounded potential; a small spike
discharge of positive polarity precedes the large following nega-
tive wave, which is almost always the most prominent feature of
the discharge. Another small positive spiky discharge usually fol-
lows. The complexity of this potential reflects its neurophysio-
logic character as an evoked potential. As a “secondary” evoked
response, it is best compared with the vertex auditory evoked
response that is maximal over the vertex (311), despite the differ-
ent views of Vaughan and Ritter (1970) (see discussion of Storm
van Leeuwen (312). Its principal deflections are found in a
latency range of 50 to 800 msec (313). H. Davis et al. (314) found
a positive deflection P1 at 50 to 60 msec, a negative N1 at about
100 msec, a positive P2 at 170 to 200 msec, and a negative N2 at
about 300 msec. The authors feel that tactile vertex responses
have slightly longer latency and show a more prominent P1.
Vertex waves may occur as an isolated event; their large
amplitude usually towers over the rest of the EEG activity. Quite
frequently, salvos of repetitive vertex waves occur, firing at a
range of about 1/sec or even somewhat faster.
Although the maximum of the vertex waves is almost invari-
ably the vertex itself, their spread may reveal asymmetries that
are fairly common in children. Pathology seldom selectively
depresses the vertex wave, but such cases may occur (315).
Sharp boundaries must be drawn between vertex waves and
rolandic spikes. Truly epileptogenic spikes may occur over the
vertex. In practice, however, the differentiation between abnor-
mal vertex spikes and the physiologic vertex wave is not diffi-
cult, because the configuration of vertex spikes is more slender
and more pointed (also see Ref. 277). Vertex waves may become
small and inconspicuous in aged individuals and are often
poorly demonstrable in such persons.
Positive Occipital Sharp Transients of Sleep (POSTS)
Another important physiologic potential of deep drowsiness
is the POSTS, which become apparent in deep drowsiness and
Figure 11.11 POSTS (positive occipital sharp transients of sleep) in a
34-year-old subject. (From Niedermeyer E, Lentz WJ. Dreaming in non-
REM sleep. Waking and Sleeping. 1976;1:49–59.)
 Chapter 11
persist during light sleep and deep sleep (Fig. 11.11). This dis-
charge was described first by Gibbs and Gibbs (42) as “positive
spike-like waves in occipital areas”; it was found most com-
monly in adolescents and young and middle-aged adults. Its
prevalence declines after the age of 70 years (316).
Similarities to occipital lambda waves were pointed out in
several studies (98,258,317,318), and the term lambdoid activity
of drowsiness and sleep was used. Hess (319) used the term
monophasic theta in view of the predominant theta frequency of
these chiefly rhythmical positive vents, whereas the term occipi-
tal sharp waves of sleep is found in the work of Arfel and Laurette
(320). Kugler and Laub (14) used the term lambda waves during
sleep and subsequently proposed the term rho waves (321,322).
Vignaendra et al. (323) studied the relationship of these poten-
tials to the nocturnal sleep cycle, stressing their abundance in
stage 2 and 3 sleep and scarcity or absence in REM sleep; these
investigations also introduced the aforementioned term positive
occipital sharp transients of sleep (POSTS).
These potentials are found in about 50% to 80% of a healthy
adult population (49% according to Prior and Deacon (317);
79% according to Brenner et al. (324)). It is, therefore, quite dif-
ficult to attribute a special significance to this activity. A much
lower figure (24%) and a striking preponderance of females
were reported by Pristasova and Prochazka (325).
Vignaendra et al. (323) hypothesize that POSTS might rep-
resent a “playback” of information to the visual cortex in order
to reexamine visual material collected during the day. A rela-
tionship of POSTS to dreaming (true dreams rather than hyp-
nagogic imagery) in NREM sleep (stage 2) was suggested by
Niedermeyer and Lentz (326), whereas Dement (327) strongly
emphasizes the lack of dreaming in this stage. According to
Chatrian (245), POSTS are unrelated to oculomotor activity as
well as to visual imagery. In blind or severely amblyopic indi-
viduals, POSTS fail to materialize (324).
LIGHT SLEEP (STAGE 2 SLEEP)
The Transition into Light Sleep (Stage 2 Sleep)
The distinction between deep drowsiness and light sleep is
imprecise and has necessarily been drawn in a somewhat arbi-
trary and artificial manner. In individuals with a well-developed
drowsy state (i.e., in most children, adolescents, and adults),
there is a very smooth transition from drowsiness to sleep. It is
customary to regard the appearance of spindle trains in the 12 to
15/sec range as the signal of sleep onset. Not much can be gained
by the observation of the patient. Even snoring may occur in
what is deep drowsiness from the EEG viewpoint.
In EEG practice, montages ought to be suitable for the
recognition of early sleep patterns. The first runs of spindles
occur over the vertex with considerable spread into parietal and
central leads. These areas should be well covered. A circular
montage covering frontopolar, temporal, and occipital areas
could miss the first minutes of stage 2 sleep. Referential linkages
from the vertex and centroparietal areas to the ipsilateral ear are
quite informative because of the wide interelectrode distance.
■ Normal EEG and Sleep: Adults and Elderly 201
Once sleep has begun, the record shows a mixture of
arousal-related patterns that truly deserve the often overused
term background activity. For the electroencephalographer,
sleep would be a drab and dull affair if the picture were not
enlivened by numerous arousal responses such as spindles, ver-
tex waves, and K complexes. In addition, positive occipital tran-
sients of sleep are extremely common and may reflect some
ongoing visual activities.
In sleep, the electrophysiologic “roar” of the brain, which
accounts for the alpha rhythm and a variety of other slow and
fast activities, mellows and does not muffle event-related pat-
terns to the same degree as in the waking state. Events express-
ing themselves as evoked potentials, such as the vertex wave,
tower over the background activity, whereas the same events
must be extracted by special averaging methods from the elec-
troencephalographic “noise” of the brain in the waking state. In
the more recent view of the “chaos theory,” however, the EEG
noise should not be regarded as an anarchic state (328).
These are but a few words from the vantage point of the EEG
laboratory. The polygraphic approach of nocturnal or 24-hour
polysomnography is quite different and is presented elsewhere.
The polygraphic approach permits further insight into a variety
of autonomically and neurobiochemically governed mecha-
nisms. On the other hand, the polygraphic approach cannot do
full justice to EEG details because of limitations of channels and
data compression due to slow paper speed, so that the clinical
electroencephalographer’s insight into EEG sleep mechanisms
retains its great value.
Background Activity During Light Sleep
Background activity shows considerable interindividual varia-
tion. Slow frequencies ranging from 0.75 to 4/sec are usually pre-
dominant; their voltage is high with a very prominent occipital
peak in small children and gradually falls throughout adoles-
cence and adulthood. The true amount of the slowest frequency
components is not readily assessable because of limitations due
to input capacitance and between-stages capacitance (deter-
mining the time constant and the low-frequency filtering). Very
pronounced slowing, especially in frontal leads, is often due to
perspiration artifacts. In adulthood, the voltage of the slow
activity is moderate and may fall to the extent that the record-
ing must be done with higher sensitivity in order to evaluate the
background amplitudes.
Patients with low-voltage records in the waking state may
also show below-average voltage while asleep. This is particu-
larly true when the low-voltage awake tracing is likely to be
related to organic cerebral problems affecting the synchroniza-
tion of EEG activity, probably in the pontine level (see Ref.
213). Patients with vertebrobasilar artery insufficiency infre-
quently show such low-voltage patterns during both wakeful-
ness and sleep (165,329), and low-voltage patterns are also
common in advanced cases of chronic alcoholism (330). On the
other hand, the low-voltage waking records of tense and overly
anxious patients change to average voltage output in sleep.
Fast frequencies are commonly in the range of 15 to 30/sec,
which lies clearly above the spindle frequency of 12 to 14/sec.
202 Part II ■ Normal EEG

This activity may occur even in unsedated patients with no his-

tory of chronic intake of sedatives. Some individuals show
frequencies in the range of 28 to 40/sec, especially over frontal
areas. Gibbs and Gibbs (42) focused much attention on such
fast patterns (“F1, F2, F3 patterns”) and consider very fast activ-
ity (above 30/sec) an abnormality.
These very fast waves occur in recurrent trains, mainly over
anterior regions. In some individuals, this type of very fast
activity could be drug related, and one wonders if one is deal-
ing with a 2:1 harmonic of the usual sedative-induced fast fre-
quencies (customarily in the 18 to 25/sec range). Relationships
to epileptic changes are not clearly demonstrable (218).
Lorimer (331) found a high prevalence of such fast patterns in
a prison population. Kubicki et al. (262,263) have emphasized
the occurrence of “subvigil beta activity” during short arousal
periods. With the use of the ultrafast frequency range (80 to
1000 Hz), more light will be shed on the function of fast and
ultrafast frequency during sleep.
According to human sleep studies of Marshall et al. (332), a
negative DC shift is associated with the transition from the
waking state to sleep. Luthringer et al. (333) reported a paral-
lelism between slow wave activity and plasma renin levels in
human sleep.

Sleep Spindles
Sleep spindles are also known as sigma activity or sigma waves
(Figs. 11.12 and 11.13). These latter terms were at first interna-
tionally encouraged, for instance by the IFSECN in 1961, Figure 11.13 Light sleep (non-REM stage 2), in a patient aged 40 years.
Well-developed and rather widespread spindle activity, mostly in the
range of 13 to 15/ sec, most prominent in superior frontal, central, and
parietal leads. Also note slow background activity.

but were discouraged in a later glossary from the same federa-

Figure 11.12 Light sleep (non-REM stage 2), in a patient aged 31 years.
Two categories of spindles are shown; spindles of 14 to 15/ sec fre-
quency dominate over centroparietal region, whereas less abundant
spindles in the 12 to 13/ sec range are found over frontal area.
tion (18). In this glossary, spindles are defined as a “group of
rhythmic waves characterized by progressively increasing, then
gradually decreasing amplitude” (see also Ref. 3).
This waxing and waning of amplitudes is exactly what the
term spindle implies: a belly in the middle tapering off to the left
and to the right. Unfortunately, experience teaches us time and
again that this shape of spindle train is not the rule and may be
rather uncommon. As a matter of fact a train of posterior alpha
waves in the waking state may show better spindle shape than
the average sleep spindles. Be this as it may, the fact remains that
the term sleep spindles is in widespread use and should be
retained in order to avoid confusion.
The observation of sleep spindles dates back to the work of
Loomis et al. (334). Gibbs and Gibbs (42) made remarkable
contributions to the description of the wave morphology of
spindles. The first spindle trains show frequencies around
14/sec (from 12.5 to 15.5/sec with a pronounced peak at exactly
14/sec, according to Gibbs and Gibbs). These investigators
described the spatial distribution of spindle trains as primarily
central, but the use of midline electrodes shows a very definite
maximum over the vertex. It is, therefore, worthwhile to include
the vertex region in montages when it is anticipated that the
patient will fall asleep.
 Chapter 11 ■ Normal EEG and Sleep: Adults and Elderly 203

Gibbs and Gibbs (42) separated the 14/sec spindle activity of literature dealing with spindles that, in reality, are trains of
from the 12/sec and the 10/sec spindles. The 12/sec spindle barbiturate-induced beta activity. Bremer (343) suggested that
train appears a little later with deepening sleep although the barbiturate spindles are the animal counterpart of human alpha
patient is still in stage 2. These spindles range, according to rhythm. Barbiturate spindles have been used as animal models
Gibbs and Gibbs (42), from 11 to 13.5/sec with a pronounced for human alpha rhythm (90,91,344). It must not be forgotten
peak at 12/sec. Their spatial distribution is different; there is that this work is based on drug-induced fast activity Steriade et
accentuation over frontal areas, with an impressive maximum al. (103) have strongly refuted the concept of Andersen and
over frontal midline. Gibbs and Gibbs stress bilaterally inde- Andersson (90,91). Based on experimental work in the cat, it is
pendent occurrence of 12/sec over frontal areas, but such asym- now presumed that the reticular thalamic nucleus of the thala-
metries are not always striking. Both types of spindles may be mus generates spindle activity (103,345).
seen simultaneously in stage 2. With deepening sleep (at the In very unusual cases, spindles may be present even in the
transition from stage 2 to 3), an even slower type of spindle waking state, although in attenuated form; this is true for occa-
(around 10/sec) (42) is commonly seen; it is more widespread sional cases of “extreme spindles” (346). Extremely rare are
with a frontal maximum. These spindles are likely to be a fore- cases of typical (unmitigated) spindle activity—along with ver-
runner of rhythmical activity in the 6 to 10/sec range, which is tex waves and K complexes—in the presence of posterior alpha
commonly found in stage 3. This type of activity should be dif- rhythm in a waking individual (347).
ferentiated from sleep spindles (335).
The evolution of sleep spindles in infancy and childhood is K Complexes
presented elsewhere in this volume. Spindles are most impres- K complexes make their appearance in stage 2 sleep and consti-
sive in childhood and adolescence; their voltage tends to tute an impressive response to arousing stimuli (Fig. 11.14).
become smaller throughout adulthood. In old age, spindles of These compounded potentials were first described by Loomis et
low voltage are very common; this is probably not due to old al. (334); the reason for calling them “K complexes” remains
age alone, but due to cerebrovascular disease. Sleep spindles obscure (there have been reports that the naming was made on
share this decline of voltage with vertex and K complexes as age the spur of the moment without attaching any significance to
advances. Sleep spindles have been reported to be more fre- the letter K). Others assume that “K” stands for “knocking” and
quent in women than in men (336); the same type of sex differ- the ensuing arousal activity in the EEG.
ence was also noticed in K complexes. Davis et al. (348) gave an excellent description of the single
The physiologic basis, nature, and significance of spindles components of the K complex. As to the topographical distri-
are beyond the scope of this chapter. Briefly, Hess (319) felt that bution, the K complex shows a maximum over the vertex, but
spindles cannot be explained simply as “slowed beta activity.” there are also K complexes with an indubitable maximum over
Spindles occur independently from faster, mostly 18 to 25/sec,
activity and both patterns may be seen at the same time. Basic
concepts of spindle activity have been discussed by Jankel and
Niedermeyer (337).
The spatial distribution of spindles indicates predominant
frontal lobe involvement. The vertex regions, where spindles
appear first, reflect activity from the most posterior portion of
the supplementary motor region, or frontal lobe. This indicates
a certain relationship to vertex waves (see above) and most K
complexes (see below) recorded from the vertex region. Vertex
waves as well as K complexes are often combined with spindle
trains. There is evidence that vertex waves and K complexes are
arousal responses, whereas the arousal-induced nature of spin-
dles still requires convincing confirmation, despite the encour-
aging findings of Church et al. (338) (also see Ref. 337).
In depth recordings from epileptic patients, sleep spindles
tend to occur in deep frontal regions (from orbital cortex or
from buried gyri of the inferior or middle frontal gyri) while
the patient is drowsy and not yet asleep, according to scalp EEG
evidence (337,339–342). With thalamic depth recordings in
patients with states of chronic pain, independent thalamic
(ventrobasal complex) spindles may appear in states of very
light drowsiness (38). These findings suggest the presence of
multiple generator areas for spindles in humans.
Sleep spindles are probably found in most mammals (see Figure 11.14 Close-up of a K complex, in a patient aged 17 years. Note
Ref. 337). Unfortunately, the term spindles is used quite loosely midline maximum (especially Cz: vertex) and the three principal com-
in experimental neurophysiology. There has been a sizable bulk ponents, the sharp, slow, and fast components.
204 Part II ■ Normal EEG

the frontal midline. Davis et al. had already noted the existence DEEP SLEEP (STAGE 3)
of central and frontal K complexes, but these authors did not
use midline electrodes. Brazier (349) presumed two distinct This stage is recorded much less often than stage 2 under rou-
generators; these were area 6, corresponding with the vertex, tine EEG laboratory conditions. In this stage, the slow back-
and area 9, corresponding with frontal midline. ground activity moves into the foreground and dominates the
As to the wave morphology, the K complex consists of an ini- picture. Delta frequencies in the 0.75 to 3/sec range are partic-
tial sharp component, followed by a slow component that fuses ularly prominent over the anterior regions. Rhythmical activity
with a superimposed fast component. The sharp component is of lesser voltage and in the 5 to 9/sec range is quite common;
biphasic and not seldom multiphasic. Its shape may closely resem- these frequencies probably do not represent a slow spindle
ble that of an isolated vertex wave, but in general the sharp com- equivalent. Sleep spindles in the 10 to 12/sec and even in the 12
ponent of the K complex shows greater complexity and greater to 14/sec ranges are still present but gradually become less
variation from complex to complex. The sharp component was impressive; they seem to be “suffocated” by the luxuriant
thoroughly investigated by Roth and Green (350). The slow com- growth of high-voltage delta activity.
ponent is represented by a large slow wave that may exceed 1000 A close look at the anterior delta activity reveals interspersed
msec in duration; its length and the degree of prominence greatly sharp transients that may be just rudimentary. This indicates a
depend on the filter setting. Shortening of the time constant all certain relationship to K complexes, which seem to be present
but suppresses the slow component. Superimposed on the slow in unending sequences. Nevertheless, more typical isolated K
component are 12 to 14/sec sleep spindles that represent the fast complexes as response to arousing stimuli can also materialize,
component. K complexes may follow each other in long stretches, showing the typical three components and disrupting the gen-
but this is more common in deeper stages of sleep. In the catalog eral slow pattern (Fig. 11.15).
of EEG signals in the Handbook of Electroencephalography and Anterior delta activity with interspersed minor sharp tran-
Clinical Neurophysiology (3), the K complex is described as having sients may form a pattern that has been described as “mitten
only a slow and a fast component; the initiating sharp component pattern.” This pattern has been amply discussed by Gibbs and
is equated with a vertex wave. Gibbs (218), but there are older descriptions by Leemhuis and
The evoked nature of the K complex was clearly demon- Stamps (360), Lyketsos et al. (361), Winfield and Sparer (362),
strated by Niiyama et al. (351) in healthy young adults. These Halász and Kajtor (363), and Gibbs and Gibbs (364). The
authors confirmed N100 and P200 components preceding the authors have categorized the mitten patterns according to the
K complex (313,352) and feel that even an apparently sponta-
neous K complex is, in reality, induced by an afferent stimulus.
The slow component of the K complex has been thought to
be related to a cognitive process or information processing
(353). This is a persuasive concept that is apt to explain the
basic difference between afferent arousing mechanisms of ver-
tex waves and K complexes. The greater spread of the K com-
plex into the frontal region would further bolster this concept.
The K complex is largest in older children and in early adoles-
cence; the sharp component is particularly impressive at that time.
With advancing age, the K complex shows a decline of voltage and
often degenerates into an insignificant slow potential with tiny
superimposed spindle-like waves. A distinction between sponta-
neous and click-evoked K complexes was made by Halász et al.
(354), who investigated the relationship between these two types.
The K complex shows considerable interindividual variability.
Paiva and Rosa (355) distinguish six morphologic types of K com-
plexes, especially on the basis of variations in the negative compo-
nent. These authors also found differences in the configuration of
K complexes with regard to sleep stages 2, 3, and 4. Their study
also confirmed the maximum of K-complex density in stage 2.
According to Weisz et al. (356), the generation of K complexes
is not controlled by the right hemisphere. Unilateral right-sided
control of arousing mechanisms had been proposed by Heilman
and Van den Abell (357) following studies of unilateral spatial
neglect (358). In this condition, K complexes are present, whereas
spindles are absent (356). It might be interesting to note that both Figure 11.15 Deep sleep (non-REM stage 3), in a patient aged 38 years.
spindles and K complexes are absent in fatal thalamic degenera- There is a greater amount of slow activity than in stage 2. Spindles are
tion with insomnia (359), a condition in which the anterior and still abundant but less prominent than in stage 2. K complexes are
the mediodorsal thalamic nuclei degenerate selectively. noted over anterior areas, forming “mitten” patterns.
 Chapter 11
shape and duration of the sharp component, which is called the
“thumb section,” while the slow component would represent
the “hand section” of the mitten.
In a sizable number of healthy persons, there is an “alpha
sleep pattern” that reaches its maximum in stage 3. This pattern
was described first by Scheuler and Stinshoff (335) and
Scheuler et al. (365); it is characterized by rhythmical 7 to
11/sec activity of frontal maximum, mostly intermingled with
slow activity in the delta range. This pattern shows certain peri-
odicities (366). The alpha sleep pattern must be differentiated
from alpha-delta sleep (367,368), a pattern associated with
delta sleep deficit and often noted in patients with fibrositis
(369). By contrast, the alpha sleep pattern has been found to be
associated with a stable sleep organization (370).
VERY DEEP SLEEP (STAGE 4)
Stage 4 is of rather little interest in the clinical EEG laboratory,
although cases of temporal lobe epilepsy may occasionally
withhold their spike activity until stages 3 and 4 are reached;
this, however, is an exception rather than the rule. In this stage,
slow activity in the delta range becomes even more preponder-
ant than in stage 3. Spindles become quite rare but are still
detectable, sometimes only when slow frequencies are filtered
out (327) (Fig. 11.16).
Figure 11.16 Very deep sleep (non-REM stage 4), in a patient aged 41
years. Pronounced generalized 1.5 to 2/ sec activity with frontal maxi-
mum and occasional traces of spindles.
■ Normal EEG and Sleep: Adults and Elderly 205
This stage of very deep sleep seems to have considerable
endocrinologic effect because the release of pituitary growth
hormone is limited to stages 3 and 4 with a peak at stage 4 (371).
Arousal at this stage can be associated with manifestations of
sleep disorders (somnambulism, nocturnal terror, or enuresis)
in predisposed individuals (372), and can produce marked con-
fusion (“sleep drunkenness”) in some patients. On the basis of
coherence analysis, Banquet (373) presumes that a transient par-
tial interhemispheric disconnection occurs in stage 4.
RAPID EYE MOVEMENT (REM) SLEEP
REM sleep should be assessed with polygraphic methods.
Under regular EEG laboratory conditions, the observation of
REM sleep requires a long waiting period, because the first
phase of REM does not appear before 60 to 90 minutes after
sleep onset; sedation of the patient may even lengthen this
period. A prolonged morning or afternoon recording may
induce such REM stages, which are quite common in daytime
naps (374), provided that such naps are long enough.
Whenever the electroencephalographer must search specifi-
cally for REM periods (especially when no regional sleep labora-
tory with facilities for polysomnography is available), stress must
be laid on the use of two channels for electrooculography (two
additional canthus electrodes, connected to the ears), one channel
for cutaneous electromyography (submental region) and a ther-
mocouple or strain gauge for pneumographic documentation.
In patients with a history of narcolepsy and additional
attacks of cataplexy and/or sleep paralysis, REM sleep may
occur at sleep onset. Even with the regular EEG laboratory rou-
tine of short daytime sleep recordings, however, sleep-onset
REM or a REM stage shortly after falling asleep may occur
unexpectedly (Fig. 11.17). Patients with severely disrupted
sleep–waking cycles, in delirium tremens, or with deep-seated
lesions or disturbances involving the brainstem may occasion-
ally show early REM phases.
The REM stage EEG is of low-voltage activity, polyrhythmic,
and similar to stage 1 (light drowsiness); periods of alpha
waves, slightly slower than in wakefulness, or “sawtooth waves”
in the 2 to 6/sec range appear in short bursts over frontal leads
or vertex (375). These sawtooth-like bursts may occur in con-
junction with ocular movements (376). REM sleep also con-
tains a good deal of alpha band activity, which is attenuated
after REM sleep deprivation (377).
The sudden appearance of totally unexpected REM sleep in
a routine EEG laboratory recording must not escape the elec-
troencephalographer, even though the patient has no additional
oculographic leads (also see Ref. 378). The ocular potentials are
mostly quite impressive and almost always give rise to marked
eye movement artifacts in frontopolar and anterior temporal
leads. Once familiarized with the tempo and shape of these dis-
charges, the electroencephalographer will rarely miss REM
states. Of course, the use of polygraphic documentation makes
the task much easier. When an REM phase occurs directly at
sleep onset, the transition from drowsiness to REM is not very
pronounced in the EEG, whereas the change from deep NREM
sleep to REM is a very dramatic one.
206 Part II ■ Normal EEG
Figure 11.17 Unexpected sleep-onset REM episode, recorded without
special ocular leads in a 26-year-old patient with brittle juvenile dia-
betes. Note ocular artifacts and general voltage attenuation with mixed
frequencies.
AROUSAL FROM SLEEP
In adolescents and adults, arousal from sleep is a quick process
with almost immediate change from sleep into a waking pat-
tern, usually with well-developed posterior alpha rhythm. One
single K complex or a sequence of K complexes may mark the
transition. More dramatic rhythmic theta activity during
arousal from sleep can be seen in children and is discussed else-
where in this volume.
REFERENCES
1. Berger H. Über das Elektrenkephalogramm des Menschen. Arch
Psychiatr Nervenkr. 1929;87:527–570.
2. Jasper HH, Andrews HL. Electroencephalography. III. Normal dif-
ferentiation of occipital and precentral regions in man. Arch
Neurol Psychiatry (Chicago). 1938;39:96–115.
3. Dutertre F. Catalogue of the main EEG patterns. In: Remond A, ed.
Handbook of Electroencephalography and Clinical Neurophysiology.
Vol 11A. Amsterdam: Elsevier; 1977:40–79.
4. Başar E. Brain natural frequencies are causal factors for resonances
and induced rhythms. In: Başar E, Bullock TH, eds. Induced
Rhythms in the Brain. Boston: Birkhäuser; 1992:425–467.
5. Bullock TH. Introduction to induced rhythms: a widespread, het-
erogeneous class of oscillations. In: Başar E, Bullock TH, eds.
Induced Rhythms in the Brain. Boston: Birkhauser; 1992:1–26.
6. Eckhorn R, Schanze T, Brosch M, et al. Stimulus-specific synchro-
nizations in cat visual cortex: multiple microelectrode and corre-
lation studies from several cortical areas. In: Başar E, Bullock TH,
eds. Induced Rhythms in the Brain. Boston: Birkhäuser;
1992:47–80.
7. Gray CM, Engel AK, König P, et al. Mechanisms underlying the
generation of neuronal oscillations in cat visual cortex. In: Başar E,
Bullock TH, eds. Induced Rhythms in the Brain. Boston:
Birkhäuser; 1992:29–45.
8. Walter WG. The location of cerebral tumors by electroencephalog-
raphy. Lancet. 1936;2:305–308.
9. Knott JR. The theta rhythm. In: Remond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 6A.
Amsterdam: Elsevier; 1976:69–77.
10. Silbert PL, Radhakrishnan K, Johnson J, et al. The significance of
the phi rhythm. Electroencephalogr Clin Neurophysiol.
1995;95:71–76.
11. Belsh JM, Chokroverty S, Barabas G. Posterior rhythmic slow
activity in EEG after eye closure. Electroencephalogr Clin
Neurophysiol. 1983;56:562–568.
12. Laugier H, Liberson WT. Contribution à l’étude de l’EEG humain.
C R Soc Biol (Paris). 1937;125:13–17.
13. Kugler J. Elektroenzephalographie in Klinik und Praxis. 3rd ed.
Stuttgart: Thieme; 1981.
14. Kugler J, Laub M. “Puppet show” theta rhythm. Electroencephalogr
Clin Neurophysiol. 1971;31:532–533 [Abstract].
15. Magnus O, Van der Holst M. Zeta waves: a special type of slow delta
waves. Electroencephalogr Clin Neurophysiol. 1987;67:140–146.
16. Siepman TAM, Cherian PJ, Visser GH. Zeta waves, an unusual EEG
finding in structural brain lesions: report of two patients. Am
J END Technol. 2004;44:24–29.
17. Cooper R, Winter AL, Crow JJ, et al. Comparison of subcortical,
cortical and scalp activity using indwelling electrodes in man.
Electroencephalogr Clin Neurophysiol. 1965;18:217–228.
18. IFSECN. A glossary of terms commonly used by clinical electroen-
cephalographers. Electroencephalogr Clin Neurophysiol. 1974;37:
538–548.
19. Markand OM. Alpha rhythms. J Clin Neurophysiol. 1990;7:163–189.
20. Petersén I, Eeg-Olofsson O. The development of the electroen-
cephalogram in normal children from the age of 1 through 15
years. Nonparoxysmal activity. Neuropädiatrie. 1971;2:247–304.
21. Townsend RE, Lubin A, Naitoh P. Stabilization of alpha frequency
by sinusoidally modulated light. Electroencephalogr Clin
Neurophysiol. 1975;39:515–518.
22. Mulholland T. The concept of attention in the electroencephalo-
graphic rhythm. In: Evans CR, Mulholland T, eds. Attention in
Neurophysiology: An International Conference. London:
Butterworth; 1969:100–127.
23. Mulholland T, Evans CR. An unexpected artifact in the human
electroencephalogram and the orienting of the eyes. Nature
(Lond). 1965;207:36–37.
24. Fenwick BBC, Walker S. The effect of eye position on the
alpha rhythm. In: Evans CR, Mulholland T, eds. Attention in
Neurophysiology: An International Conference. London: Butterworths;
1969:128–141.
25. Isokawa M, Komisaruk BR. Convergence of finger tremor and EEG
rhythm at the alpha frequency induced by rhythmical photic stim-
ulation. Electroencephalogr Clin Neurophysiol. 1983;55:580–585.
26. Storm van Leeuwen W, Bekkering DH. Some results obtained with
the EEG spectograph. Electroencephalogr Clin Neurophysiol. 1958;
10:563.
27. Cobb WA. The normal adult EEG. In: Hill D, Parr G, eds.
Electroencephalography. New York, NY: Macmillan; 1963:232–249.
 Chapter 11
28. Simonova O, Roth B, Stein J. EEG studies of healthy population—
normal rhythms of resting recording. Act Unv Carol Med (Praha).
1967;13:543–551.
29. Simon O. (Simonova, O.) Das Elektroenzephalogramm. Munich:
Urban & Schwarzenberg; 1977.
30. Brazier MAB, Finesinger JE. Characteristics of the normal elec-
troencephalogram. I. A study of the occipital cortical potentials in
500 normal adults. J Clin Invest. 1944;23:303–311.
31. Wieneke GH, Deinema CHA, Spoelstra P, et al. Normative spectral
data on alpha rhythm in male adults. Electroencephalogr Clin
Neurophysiol. 1980;49:636–645.
32. Kellaway P, Maulsby RL. The Normative Electroencephalographic
Data Reference Library. Final report. Houston, TX: Baylor College
of Medicine; 1966:50–52.
33. Kiloh LG, McComas AJ, Osselton JW. Clinical Electroence-
phalography. 3rd ed. London: Butterworths; 1972.
34. Matousek M, Okawa M, Petersén I. Inter-hemispheric differ-
ences in normals and in psychiatric patients with normal
EEG. Electroencephalogr Clin Neurophysiol. 1981;52:33P–34P
[Abstract].
35. Adrian ED, Matthews BHC. The Berger rhythm, potential changes
from the occipital lobe in man. Brain. 1934;57:345–359.
36. Berger H. Über das Elektrenkephalogramm des Menschen. Zehnte
Mitteilung (10th report). Arch Psychiatr Nervenkr. 1935;103:
444–454.
37. Albe-Fessard D. Electrophysiological techniques used to differ-
entiate thalamic nuclei. In: Remond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 10B.
Amsterdam: Elsevier; 1975:46–58.
38. Gücer G, Niedermeyer E, Long DM. Thalamic recordings in
patients with chronic pain. J Neurol (Berlin). 1978;219:47–61.
39. Lopes da Silva F, Van Lierop THMT, Schrijer CF, et al. Organization
of thalamic and cortical alpha rhythm: spectra and coherences.
Electroencephalogr Clin Neurophysiol. 1973;35:627–640.
40. Berger H. Über das Elektrenkephalogramm des Menschen. Sechste
Mitteilung (6th report). Arch Psychiatr Nervenkr. 1933;99:
555–574.
41. Gloor P. The work of Hans Berger. In: Remond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 1A.
Amsterdam: Elsevier; 1971:11–24.
42. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 1.
Cambridge, MA: Addison-Wesley; 1950.
43. Könönen M, Partanen JV. Blocking of EEG alpha activity during
visual performance in healthy adults. A quantitative study.
Electroencephalography. 1993;87:164–166.
44. Niedermeyer E, Krauss GL, Peyser CE. The electroencephalogram
and mental activation. Clin Electroencephalogr. 1989;20:215–226.
45. Berger H. Über das Elektrenkephalogramm des Menschen. Dritte
Mitteilung (3rd report). Arch Psychiatr Nervenkr. 1931;94:16–60.
46. Berger H. Über das Elektrenkephalogramm des Menschen. Vierte
Mitteilung (4th report). Arch Psychiatr Nervenkr. 1932;97:6–26.
47. Berger H. Über das Elektrenkephalogramm des Menschen.
Siebente Mitteilung (7th report). Arch Psychiatr Nervenkr.
1933;100:301–320.
48. Berger H. Über das Elektrenkephalogramm des Menschen.
Zwölfte Mitteilung (12th report). Arch Psychiatr Nervenkr. 1937;
106:165–187.
49. Adrian ED. The discovery of Berger. In: Remond A, ed. Handbook
of Electroencephalography and Clinical Neurophysiology. Vol 1.
Amsterdam: Elsevier; 1971:5–10.
50. Davis H, Davis PA. Active potentials of the brain in normal per-
sons and in normal states of cerebral activity. Arch Neurol
Psychiatry (Chicago). 1936;36:1214–1224.
■ Normal EEG and Sleep: Adults and Elderly 207
51. Golla F, Hutton EL, Walter WG. The objective study of mental
imagery. I. Physiological concomitance. Appendix on new method
of electroencephalographic analysis. J Ment Sci. 1943;89:216–223.
52. Kuhlo W. The beta rhythms. In: Remond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 6A.
Amsterdam: Elsevier; 1976:29–46.
53. Lennox WG, Gibbs FA, Gibbs EL. The brain wave pattern, an
hereditary trait. Evidence from 74 “normal” pairs of twins. J Hered.
1945;36:233–243.
54. Travis LE, Gottlober AB. Do brain waves have individuality?
Science. 1936;84:532–533.
55. Veldhuizen RJ, Jonlman EJ, Poortvliet DCJ. Sex differences in age
regression parameters of healthy adults—normative data and
practical implications. Electroencephalogr Clin Neurophysiol. 1993;
86:377–384.
56. Lemere F. The significance of individual differences in the Berger
rhythm. Brain. 1936;59:366–375.
57. Saul LJ, Davis H, Davis PA. Correlations between electroen-
cephalograms and the psychological organization of the indi-
vidual. Trans Am Neurol Assoc. 1937;63:167–169.
58. Lindsley DB. Electrical potentials of the brain in children and
adults. J Gen Psychol. 1938;19:285–306.
59. Henry CE, Knott JR. A note on the relationship between personal-
ity and the alpha rhythm of the electroencephalogram. J Exp
Psychol. 1941;28:362–366.
60. Broadhurst A, Glass A. Relationship of personality measures to the
alpha rhythm of the electroencephalogram. Br J Psychiatry.
1969;115:199–204.
61. Nowak SM, Marczynski TJ. Trait anxiety is reflected in EEG alpha
response to stress. Electroencephalogr Clin Neurophysiol. 1981;52:
175–191.
62. Berger H. Über das Elektrenkephalogramm des Menschen.
Vierzehnte Mitteilung (14th report). Arch Psychiatr Nervenkr.
1938;108:407–431.
63. Gastaut H. Correlations between the electroencephalographic and
the psychometric variables (M.M.P.I., Rosenzweig, intelligence
tests). Electroencephalogr Clin Neurophysiol. 1960;12:226–227.
64. Henry CE. Electroencephalograms of normal children. Monogr
Soc Res Child Dev, Nat Res Council. Vol XI. Washington, DC; 1944.
65. Kreezer G. Electrical potentials in the brain in certain types of
mental deficiency. Arch Neurol Psychiatry (Chicago). 1936;36:
1206–1214.
66. Shagass C. An attempt to correlate the occipital alpha frequency of
the electroencephalogram with performance on a mental ability
test. J Exp Psychol. 1946;36:88–92.
67. Knott JR. The alpha rhythm. In: Remond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology. Amsterdam:
Elsevier; 1976:7–29.
68. Gasser T, Von Lucadou-Müller I, Verleger R, et al. Correlating EEG
and IQ: a new look at an old problem using computerized EEG
parameters. Electroencephalogr Clin Neurophysiol. 1983;55:493–504.
69. Klimesch W, Doppelmayr M, Schimke H, et al. Alpha frequency,
reaction time, and the speed of processing information. J Clin
Neurophysiol. 1996;13:511–518.
70. Jasper HH, Cruikshank RM. Visual stimulation and the after-
image as affecting the occipital alpha rhythm. J Gen Psychol.
1937;17:29–48.
71. Fisch BJ, Lemos MS, Hauser WA. Age-related changes in alpha
rhythm frequency and reactivity. Electroencephalogr Clin
Neurophysiol. 1990;75:S45 [Abstract].
72. Frank, G. Circadian periodicity in cerebral electrical output: rela-
tion to body temperature and stage of sleep-wakefulness cycles.
Electroencephalogr Clin Neurophysiol. 1964;17:712 [Abstract].
208 Part II ■ Normal EEG
73. Scheich H. Interval histograms and periodic diurnal changes of
human alpha rhythm. Electroencephalogr Clin Neurophysiol.
1969;26:442 [Abstract].
74. Harding GFA, Thompson CRS. EEG rhythms and the internal
milieu. In: Remond A, ed. Handbook of Electroencephalography and
Clinical Neurophysiology. Vol 6A. Amsterdam: Elsevier; 1976:
176–194.
75. Gundel A. The influence of body temperature on the waking EEG.
Electroencephalogr Clin Neurophysiol. 1984;57:33P [Abstract].
76. Dubois M, Sato S, Lees DE, et al. Electroencephalographic changes
during whole body hyperthermia in humans. Electroencephalogr
Clin Neurophysiol. 1980;50:486–495.
77. Westmoreland B, Klass OW. A distinctive rhythmic EEG discharge
of adults. Electroencephalogr Clin Neurophysiol. 1981;51:186–191.
78. Anand BK, China GS, Singh B. Some aspects of electroencephalo-
graphic studies in yogis. Electroencephalogr Clin Neurophysiol.
1961;13:452–456.
79. Bagchi BK, Wenger MA. Simultaneous EEG and other recordings
during some yogi practices. Electroencephalogr Clin Neurophysiol.
1958;10:193 [Abstract].
80. Gastaut H. Vom Berger-Rhythmus zum Alpha-Kult und zur
Alpha-Kultur. Z EEG-EMG. 1974;5:189–199.
81. Nowlis D, Kamiya J. The control of electroencephalographic alpha
rhythms through auditory feedback and the associated mental
activity. Psychophysiology. 1970;6:476–484.
82. Brown BB. Recognition of aspects of consciousness through asso-
ciations with EEG alpha activity represented by a light signal.
Psychophysiology. 1970;6:442–451.
83. Wallace RK. Physiological effects of transcendental meditation.
Science. 1970;167:1751–1754.
84. Wallace RK, Benson M, Wilson AJ. A wakeful hypometabolic phys-
iologic state. Am J Physiol. 1971;221:795–799.
85. Banquet JP. Comparative study on the EEG spectral analysis dur-
ing sleep and yoga meditation. In: Proceedings of the First European
Congress of Sleep Research. Basel; 1973:389–394.
86. Stigsby B, Rodenberg JC, Moth HB. Electroencephalographic find-
ings during mantra meditation (transcendental meditation). A
controlled qualitative study of experienced meditators. Electro-
encephalogr Clin Neurophysiol. 1981;51:434–442.
87. Dongier M, McCallum WC, Torres F, et al. Psychological and psy-
chophysiological states. In: Remond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 16A.
Amsterdam: Elsevier; 1976:195–256.
88. Scherzer E. Das flache EEG als bioelektrischer Ausdruck der
Erwartungsspannung (psychogene Alphareduktion). Psychiatr
Neurol (Basel). 1966;152:207–212.
89. Bishop GH. The interpretation of cortical potentials. Cold Spring
Harb Symp Quant Biol. 1936;4:305–319.
90. Andersen P, Andersson SA. Physiological mechanisms of the alpha
waves. In: Kellaway P, Petersén I, eds. Clinical Electroencephalography
of Children. New York, NY: Grune & Stratton; 1968:31–48.
91. Andersen P, Andersson SA. Thalamic origin of cortical rhythmic
activity. In: Remond A, ed. Handbook of Electroencephalography and
Clinical Neurophysiology. Vol 2C. Amsterdam: Elsevier; 1974:90–118.
92. Frost J. Physiological bases of normal EEG rhythms. In: Remond
A, ed. Handbook of Electroencephalography and Clinical Neuro-
physiology. Vol 6A. Amsterdam: Elsevier; 1976:150–160.
93. Lopes da Silva F, Van Lierop THMT, Schrijer CF, et al. Essential dif-
ferences between alpha rhythms and barbiturate spindles: spec-
tra and thalamocortical coherences. Electroencephalogr Clin
Neurophysiol. 1973;35:641–645.
94. Watanabe S. Rhythmicity of EEG and stability of alpha rhythm.
Electroencephalogr Clin Neurophysiol. 1981;52:62P [Abstract].
95. Adrian ED, Yamagiwa K. The origin of the Berger rhythm. Brain.
1935;58:323–351.
96. Aird RB, Garoutte B. Studies on the “cerebral pacemaker.”
Neurology (Minneapolis). 1958;8:581–589.
97. Walter DO, Rhodes JM, Brown D, et al. Comprehensive spectral
analysis of human EEG generators in posterior cerebral areas.
Electroencephalogr Clin Neurophysiol. 1966;20:224–237.
98. Perez-Borja C, Chatrian GE, Tyce FA, et al. Electrographic pat-
terns of the occipital lobes in man: a topographic study based on
use of implanted electrodes. Electroencephalogr Clin
Neurophysiol. 1962;14:171–182.
99. Remond A. The importance of topographic data in EEG phe-
nomena and an electric model to reproduce them. In: Walter DO,
Brazier MAB, eds. Advances in EEG Analysis. Amsterdam:
Elsevier; 1968:29–46.
100. Cooper R, Mundy-Castle AC. Spatial and temporal analysis of
the alpha rhythm: a toposcopic analysis. Electroencephalogr Clin
Neurophysiol. 1960;12:153–165.
101. Liske E, Hughes HM, Stowe DE. Cross-correlation of human
alpha activity: normative data. Electroencephalogr Clin
Neurophysiol. 1967;22:429–436.
102. Inouye T, Shinosaki K, Yagasaki A. The direction of spread of
alpha activity over the scalp. Electroencephalogr Clin Neurophysiol.
1983;55:290–300.
103. Steriade M, Gloor P, Llinas RR, et al. Basic mechanisms of cere-
bral rhythmical activities. Electroencephalogr Clin Neurophysiol.
1990;76:481–508.
104. Maddocks JA, Hodge RS, Rex J. Observations on the occurrence
of precentral activities at alpha frequencies. Electroencephalogr
Clin Neurophysiol. 1951;3:370 [Abstract].
105. Schütz E, Müller HW. Über ein neues Zeichen zentralnervöser
Erregbarkeitssteigerung im Elektroenzephalogramm. Klin
Wochenschr. 1951;29:22–23.
106. Gastaut H, Terzian H, Gastaut Y. Étude d’une activité élec-
troencéphalographique méconnue: “Le rhythme rolandique en
arceau.” Mars Méd. 1952;89:296–310.
107. Gastaut H. Étude électrocorticographique de la réativité des
rhythmes rolandiques. Rev Neurol (Paris). 1952;87:176–182.
108. Gastaut H, Dongier M, Courtois G. On the significance of
“wicket rhythms” (“rhythmes en arceau”) in psychosomatic
medicine. Electroencephalogr Clin Neurophysiol. 1954;6:687.
109. Magnus O. The cerebral alpha-rhythm (“rhythme en arceau”).
Electroencephalogr Clin Neurophysiol. 1954;6:349–350.
110. Van der Drift JH, Magnus O. The EEG in cerebral ischemic lesions.
Correlations with clinical and pathological findings. In: Gastaut H,
Meyer JS, eds. Cerebral Anoxia and the Electroencephalogram.
Springfield, IL: Charles C Thomas; 1961:180–196.
111. Kuhlman WN. Functional topography of the human mu rhythm.
Electroencephalogr Clin Neurophysiol. 1978;44:83–93.
112. Pfurtscheller G, Aranibar A. Occipital rhythm activity within the
alpha band during conditioned externally paced movement.
Electroencephalogr Clin Neurophysiol. 1978;45:226–235.
113. Pfurtscheller G, Aranibar A. Änderungen in der spontanen EEG-
Aktivität vor Willkürbewegungen. Neue Wege bei der
Untersuchung der zentralen mu-Aktivität. Z EEG-EMG.
1978;9:18–23.
114. Pfurtscheller G. Central beta rhythms and sensorimotor activities
in man. Electroencephalogr Clin Neurophysiol. 1981;51: 253–264.
115. Pfurtscheller G. Event-related EEG desynchronization.
Electroencephalogr Clin Neurophysiol. 1990;75:S117 [Abstract].
116. Pfurtscheller G. Event-related synchronization (ERS): an electro-
physiological correlate of cortical areas at rest. Electroencephalogr
Clin Neurophysiol. 1992;83:62–69.
 Chapter 11
117. Pfurtscheller G, Klimesch W. Event-related synchronization and
desynchronization of alpha and beta waves in a cognitive task. In:
Başar E, Bullock TH, eds. Induced Rhythms in the Brain. Boston,
MA: Birkhäuser; 1992:117–128.
118. Pfurtscheller G, Neuper C. Simultaneous EEG 10 Hz desynchro-
nization and 40 Hz synchronization during finger movements.
NeuroReport. 1992;3:1057–1060.
119. Pfurtscheller G, Flotzinger D, Neuper C. Differentiation between
finger, toe and tongue movement in man based on 40 Hz EEG.
Electroencephalogr Clin Neurophysiol. 1994;90:456–460.
120. Pfurtscheller G, Lopes da Silva FH. Event-related EEG/MEG syn-
chronization and desynchronization: basic principles. Clin
Neurophysiol. 1999;110:1842–1857.
121. Pfurtscheller G, Lopes da Silva F, eds. Event-related desynchro-
nization. Handbook of Electroencephalography and Clinical
Neurophysiology, revised series. Vol 6. Amsterdam: Elsevier; 1999.
122. Klass D, Bickford RG. Observations on the rolandic arceau
rhythm. Electroencephalogr Clin Neurophysiol. 1957;9:570.
123. Schnell RG, Klass DW. Further observations on the rolandic
arceau rhythm. Electroencephalogr Clin Neurophysiol. 1966;20:95
[Abstract].
124. Beek H. Age and the central rhythm “en arceau.”
Electroencephalogr Clin Neurophysiol. 1958;10:356 [Abstract].
125. Gastaut H, Bacher F, Bert J, et al. Étude des corrélations entre les
variables électroencéphalographiques et psychométriques
(M.M.P.I., Rosenweig). Rev Neurol (Paris). 1959;101:376–384.
126. Gastaut H, Dongier S, Dongier M. Electroencephalographie et
nevroses. Rev Neurol (Paris). 1959;101:435–436.
127. Dongier M, Dongier S. Quelques aspects de l’électroencéphalo-
gramme des névroses. Évol Psychiat. 1958;1:1–18.
128. Picard P, Laboureur P, Navranne P. Examen électroencéphalo-
graphique de personnel navigeant. Méd Aéronaut. 1955;10:
53–58.
129. Simonova O, Roth B, Stein J. An EEG study of healthy subjects.
Normal rhythms of the resting record. Electroencephalogr Clin
Neurophysiol. 1966;20:279 [Abstract].
130. Niedermeyer E, Koshino Y. My-Rhythmus: vorkommen und
klinische Bedeutung. Z EEG-EMG. 1975;6:69–78.
131. Stroganova TA, Orekhova EV, Posikera IN. EEG alpha rhythm in
infants. Clin Neurophysiol. 1999;110:997–1012.
132. Schoppenhorst M, Brauer F, Freund G, et al. The significance of
coherence estimates in determining cerebral alpha and mu activ-
ities. Electroencephalogr Clin Neurophysiol. 1980;48:25–33.
133. Koshino Y, Isaki K. Familial occurrence of the mu rhythm. Clin
Electroencephalogr. 1986;17:44–50.
134. Storm van Leeuwen W, Wieneke G, Spoelstra P, et al. Lack of
bilateral coherence of mu-rhythm. Electroencephalogr Clin
Neurophysiol. 1978;44:140–146.
135. Farnarier C, Mattei JP, Naquet R. The mu rhythm of the vertex
and its reactivity: an infrequent observation. Electroencephalogr
Clin Neurophysiol. 1981;52:55P [Abstract].
136. Salmelin R, Hari R. Spatiotemporal characteristics of sensorimo-
tor neuromagnetic rhythms related to thumb movement.
Neuroscience. 1994;60:537–550.
137. Salmelin R, Hari R. Characterization of spontaneous MEG
rhythms in healthy adults. Electroencephalogr Clin Neurophysiol.
1994;91:237–248.
138. Nashmi R, Mendonça AJ, MacKay WA. EEG rhythms of the sen-
sorimotor region during hand movements. Electroencephalogr
Clin Neurophysiol. 1994;91:456–467.
139. Tannier C, Bentzinger C, Feuerstein J. A case of a lesional unilat-
eral mu rhythm: clinical and EEG evolution. Electroencephalogr
Clin Neurophysiol. 1983;56:21P [Abstract].
■ Normal EEG and Sleep: Adults and Elderly 209
140. Hess R. Postoperative controls. In: Remond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 14C.
Amsterdam: Elsevier; 1975:56–65.
141. Kubicki S, Schoppenhorst M, Sack HJ, et al. Post-traumatischer
mu wave Fokus: der Einfluß von Schlaf, Pharmaka und
motorischer Aktivität. Z EEG-EMG. 1973;4:203–209.
142. Cobb WA, Guiloff RJ, Cast J. Breach rhythm: the EEG related to skull
defects. Electroencephalogr Clin Neurophysiol. 1979;47:251–271.
143. Chatrian GE. Characteristics of unusual EEG patterns: incidence,
significance. Electroencephalogr Clin Neurophysiol. 1964;17:
471–472.
144. Chatrian GE. The mu rhythms. In: Remond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 6A.
Amsterdam: Elsevier; 1976: 46–69.
145. Chatrian GE, Petersen MC, Lazarte JA. The blocking of the
rolandic wicket rhythm and some central changes related to move-
ment. Electroencephalogr Clin Neurophysiol. 1959;11: 497–510.
146. Ciganek L. A contribution to the rolandic “arceau” rhythm.
Electroencephalogr Clin Neurophysiol. 1959;11:185–186.
147. Gastaut H, Naquet R, Gastaut Y. A study of the mu rhythm in
subjects lacking one or more limbs. Electroencephalogr Clin
Neurophysiol. 1965;18:720–721.
148. Brechet R, Lecasble R. Reactivity of mu rhythm to flicker.
Electroencephalogr Clin Neurophysiol. 1965;18:721–722 [Abstract].
149. Koshino Y, Niedermeyer E. Enhancement of rolandic mu rhythm
by pattern vision. Electroencephalogr Clin Neurophysiol. 1975;38:
535–538.
150. Jasper HH, Penfield W. Electrocorticograms in man: effects of
voluntary movement upon the electrical activity of the precentral
gyrus. Arch Psychiatr Nervenkr. 1949;183:163–174.
151. Graf M, Niedermeyer E, Schiemann J, et al. Electrocorticography:
information derived from intra-operative recordings during
seizure surgery. Clin Electroencephalogr. 1984;15:83–91.
152. Van Huffelen A, Poortvliet D, Van der Wulp C. Quantitative elec-
troencephalography in cerebral ischemia. Detection of abnor-
malities in “normal” EEGs. In: Pfurtscheller G, Jonkman J, Lopes
da Silva FH, eds. Brain Ischemia: Quantitative EEG and Imaging
Techniques. Amsterdam: Elsevier; 1984:3–28.
153. Arroyo S, Lesser RP, Gordon B, et al. Functional significance of
the mu rhythm of human cortex: an electrophysiologic study
with subdural electrodes. Electroencephalogr Clin Neurophysiol.
1993;87:76–87.
154. Bostem F, Dongier M, Demaret A, et al. Discussion à propos du
rhythme mu. Rev Neurol (Paris). 1964;111:335–337.
155. Bostem F, Dongier M, Demaret A, et al. Discussion on mu rhythm.
Electroencephalogr Clin Neurophysiol. 1965;18:721 [Abstract].
156. Gastaut H, Broughton R, Regis H, et al. Quelques données nouvelles
à propos due rhythme mu. Rev Neurol (Paris). 1964;111:331–332.
157. Gastaut H, Jus A, Jus C, et al. Étude topographique des réactions
électroencéphalographiques conditionées chez l’homme.
Electroencephalogr Clin Neurophysiol. 1957;9:1–34.
158. Finley WW, Smith HA, Etherton MD. Reduction of seizures and
normalization of EEG in a severe epileptic following sensorimo-
tor biofeedback training. Biol Psychol. 1975;2:189–203.
159. Kuhlman WN. EEG feedback training: enhancement of
somatosensory cortical activity. Electroencephalogr Clin
Neurophysiol. 1978;45:290–294.
160. Kuhlman WN EEG feedback training of epileptic patients: clin-
ical and electroencephalographic analysis. Electroencephalogr
Clin Neurophysiol. 1978;45:699–710.
161. Sterman MB, MacDonald LR, Stone RK. Biofeedback training of
sensorimotor EEG in man and its effect on epilepsy. Epilepsia.
(New York). 1974;15:395–416.
210 Part II ■ Normal EEG
162. Roldan E, Lepicovska V, Dostalek C, et al. Mu-like EEG rhythm
generation in the course of Hatha-yogi exercises. Electro-
encephalogr Clin Neurophysiol. 1981;52:13P [Abstract].
163. Covello A, DeBarros-Ferreira M, Lairy GC. Étude télémétrique
des rhythmes centraux chez l’enfant. Electroencephalogr Clin
Neurophysiol. 1975;38:307–319.
164. Yamada T, Kooi KA. Level of consciousness and the mu rhythm.
Clin Electroencephalogr. 1975;6:80–88.
165. Niedermeyer E. The Generalized Epilepsies. Springfield, IL: Charles
C Thomas; 1972.
166. Niedermeyer E. Focal and generalized seizure discharges in the
electroencephalogram and their response to diazepam. Int Med
Digest. 1972;7:49–61.
167. Isch-Treussard C. Paroxysmal arched rhythms in the child.
Electroencephalogr Clin Neurophysiol. 1972;33:354 [Abstract].
168. Niedermeyer E, Long DM, Hendler NH, et al. Chronic pain and
mu rhythm. Electroencephalogr Clin Neurophysiol. 1982;53:30P
[Abstract].
169. Pfurtscheller G. Rolandic mu rhythm and assessment of cerebral
functions. Am J EEG Technol. 1986;26:19–32.
170. Niedermeyer E. An independent alpha-like rhythm in the tempo-
ral region. Electroencephalogr Clin Neurophysiol. 1990;75:S103
[Abstract].
171. Niedermeyer E. Alpha-like rhythmical activity of the temporal
lobe. Clin Electroencephalogr 1990;21:210–224.
172. Niedermeyer E. Ein unabhängiger alpha-artiger Rhythmus über
dem Temporallappen. Das EEG-Labor. 1990;12:165–173.
173. Niedermeyer E. The “third rhythm”: further observations. Clin
Electroencephalogr. 1991;22:83–96.
174. Niedermeyer E. Alpha rhythms as physiological and abnormal
phenomena. Int J Psychophysiol. 1997;26:31–49.
175. Kendel K, Koufen H. EEG-Veränderungen bei zerebralen
Gefäszinsulten des Hirnstamms. Dtsch Z Nervenheilkd. 1970;197:
42–55.
176. Berger H. Über das Elektrenkephalogramm des Menschen.
Zweite Mitteilung (2nd report). J Psychol Neurol (Leipzig).
1930;40:160–179.
177. Finley K. On the occurrence of rapid frequency potential changes
in the human electroencephalogram. Am J Psychiatry. 1944;101:
194–200.
178. Gallais P, Collomb H, Milletto G, et al. Confrontation entre les don-
nées de l’électroencéphalogramme et des examens psychologiques
chez 522 sujets repartis en trois groupes differents. II. Confrontation
des données de l’électroencéphalogramme et de l’examen psy-
chologique chez 113 jeunes soldats. In: Fischgold H, Gastaut H, eds.
Conditionnement et RŽactivitŽ en ƒlectroencŽphalographie.
Electroencephalogr Clin Neurophysiol Suppl. 1957;6:294–303.
179. Gibbs FA, Gibbs EL, Lennox WG. Electroencephalographic clas-
sification of epileptic patients and control subjects. Arch Neurol
Psychiatry (Chicago). 1943;50:111–128.
180. Mundy-Castle AC. Theta and beta rhythm in the electroen-
cephalogram of normal adults. Electroencephalogr Clin
Neurophysiol. 1951;3:477–486.
181. Mundy-Castle AC. An analysis of central responses to photic
stimulation in normal adults. Electroencephalogr Clin
Neurophysiol. 1953;5:1–22.
182. Obrist WD. The electroencephalogram of normal aged adults.
Electroencephalogr Clin Neurophysiol. 1954;6:235–244.
183. Picard P, Navaranne P, Laboureur P, et al. Confrontations des
données de l’électroencéphalogramme et des examens psy-
chologiques chez 522 sujets repartis en trois groupes differents.
III. Confrontations des données de l’électroencéphalogramme et
de l’examen psychologique chez 309 candidats pilotes a l’aéro-
nautiques. In: Fischgold H, Gastaut H, eds. Conditionnement et
RŽactivitŽ en ƒlectroencŽphalographie. Amsterdam: Elsevier;
1957:304–314.
184. Roger A, Bert J. Étude des corrélations entre les differents vari-
ables EEG. Rev Neurol (Paris). 1959;101:334–360.
185. Vogel F, Fujiya Y. The incidence of some inherited EEG variants
in normal Japanese and German males. Hum Genet.
1969;7:38–42.
186. Kuhlo W. Slow posterior activities. In: Remond A, ed. Handbook
of Electroencephalography and Clinical Neurophysiology. Vol 6A.
Amsterdam: Elsevier; 1976:89–104.
187. Fortuin B, Künkel H. The beta-type EEG in adults.
Electroencephalogr Clin Neurophysiol. 1983;56:67P [Abstract].
188. Kozelak JW, Pedley TA. Beta and mu rhythms. J Clin
Neurophysiol. 1990;7:191–207.
189. Drake ME. Clinical correlates of very fast beta activity in the
EEG. Clin Electroencephalogr. 1984;15:237–241.
190. Vogel F, Götze W. Statistische Betrachtungen über die Beta-
Wellen im EEG des Menschen. Dtsch Z Nervenheilkd.
1962;184:112–136.
191. Vogel F. Untersuchungen zur Genetik der Beta Wellen im EEG
des Menschen. Dtsch Z Nervenheilkd. 1962;184:137–173.
192. Vogel F. Zur genetischen Grundlage fronto-präzentraler Beta
Wellen in EEG des Menschen. Hum Genet. 1966;2:227–237.
193. Vogel F. The genetic basis of the normal human electroen-
cephalogram (EEG). Hum Genet. 1970;10:91–114.
194. Ingvar DH. The concept of “the cerebral hyperfrontality?”
Electroencephalogr Clin Neurophysiol. 1987;67:25P [Abstract].
195. Veilleux M, Westmoreland BF, Sharbrough FW. Fast central mid-
line rhythm. Electroencephalogr Clin Neurophysiol. 1988;69:
91P–92P [Abstract].
196. Remond A, Lesèvre N. Remarques sur l’activité cérébrale des
sujets normaux. La typologie électroencéphalographique dans
ses rapports avec certains caractérs psychologiques. In: Fischgold
H, Gastaut H, eds. Conditionnement et RŽactivitŽ en ƒlec-
troencŽphalographie. Amsterdam: Elsevier; 1957:235–255.
197. Gastaut H. Confrontation entre les données de l’élec-
troencéphalogramme et des examens psychologiques chez 522
sujets repartis en trois groupes differents. V. Conclusions
d’ensemble. In: Fischgold H, Gastaut H, eds. Conditionnement et
RŽactivitŽ en ÉlectroencŽphalographie. Electroencephalogr Clin
Neurophysiol Suppl. 1957;6:321–338.
198. Cohn R. The influence of emotions on the human electroen-
cephalogram. J Nerv Ment Dis. 1946;104:351–357.
199. Giannitrappani D, Kayton L. Schizophrenia and EEG spectral
analysis. Electroencephalogr Clin Neurophysiol. 1974;36:377–386.
200. Gastaut H, Lee MC, Laboureur P. Comparative EEG and psycho-
metric data for 825 French naval pilots and 511 control subjects
of the same age. Aerosp Med. 1960;31:547–552.
201. Matsuura M, Yamamoto K, Fukuzawa H, et al. Age development
and sex differences of various EEG elements in healthy children and
adults. Quantification by a computerized wave form recognition
method. Electroencephalogr Clin Neurophysiol. 1985;60:394–406.
202. Draguhn A, Traub RD, Bibig A, et al. Ripple (around 200-Hz)
oscillations in temporal structures. J Clin Neurophysiol.
2000;17:361–376.
203. Niedermeyer E, Sherman DL. Ultrafast frequencies: a new chal-
lenge for electroencephalography with remarks on ultraslow fre-
quencies. Clin Electroencephalogr. 2001;32:119–121.
204. Niedermeyer E, Sherman DL. Ultrafast EEG frequencies— not to
be neglected in the future. Am J EEG Technol. 2001;41:192–198.
 Chapter 11
205. Schiff SJ, Colella D, Jacyna GM, et al. Brain chirps: spectro-
graphic signatures of epileptic seizures. Electroencephalogr Clin
Neurophysiol. 2000;111:953–958.
206. Curio G. Ain’t no rhythm fast enough: EEG bands beyond beta.
J Clin Neurophysiol. 2000;17:339–340.
207. Curio G. Linking 600-Hz “spike-like” EEG/MEG wavlets (“sigma
bursts”) to cellular substrates: concepts and caveats. J Clin
Neurophysiol. 2000;17:377–396.
208. Traub RD, Jefferys JGR, Whittington MA. Fast Oscillations in
Cortical Circuits. Cambridge, MA: MIT Press; 1999.
209. Ziemann U, Rothwell JC. I-waves in motor cortex. J Clin
Neurophysiol. 2000;17:397–405.
210. Vanhatolo S, Holmes MD, Tallgren P, et al. Very slow EEG responses
lateralize temporal lobe seizures. Neurology. 2003;60:1098–1104.
211. Vanhatolo S, Tallgren P, Anderson S, et al. DC-EEG recording dis-
closes prominent very slow activity patterns in preterm infants.
Clin Neurophysiol. 2002;113:1822–1825.
212. Moruzzi G, Magoun HW. Brain stem reticular formation and
activation of the EEG. Electroencephalogr Clin Neurophysiol.
1949;1:455–473.
213. Magnes J, Moruzzi G, Pompeiano O. EEG-synchronizing struc-
tures in the lower brain stem. In: Wolstenholme GIW, O’Connor
M, eds. The Nature of Sleep. London: Churchill Livingstone;
1961:57–78.
214. Magni F, Moruzzi G, Rossi GF, et al. EEG arousal following inac-
tivation of the lower brain stem by selective injection of barbitu-
rates into the cerebral circulation. Arch Ital Biol. 1959;97:33–46.
215. Lawson AM, Niedermeyer E. Central pontine myelinolysis: clini-
cal and EEG considerations. Am J Electroneurodiagnostic Technol.
2002;42:151–153.
216. Rosas HD, Koroshetz WS, Chen YI, et al. Evidence for more wide-
spread cerebral pathology in early HD. Neurology. 2003;60:
1615–1620.
217. Jasper HH, Fitzpatrick CA, Colomon P. Analogies and opposites
in schizophrenia and epilepsy. Am J Psychiatry. 1939;95:835–851.
218. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 3.
Reading, MA: Addison–Wesley; 1964.
219. Adams A. Studies on flat electroencephalogram in man.
Electroencephalogr Clin Neurophysiol. 1959;11:34–41.
220. Pine I, Pine HM. Clinical analysis of patients with low voltage
EEG. J Nerv Ment Dis. 1953;117:191–198.
221. Vogel F, Götze W. Familienuntersuchungen zur Genetik des nor-
malen Elektroenzephalogramms. Dtsch Z Nervenheilkd. 1959;
178:668–700.
222. Niedermeyer E. Clinical correlates of flat or low voltage records.
Electroencephalogr Clin Neurophysiol. 1963;15:148 [Abstract].
223. Panzani R, Turner M. Étude électroencéphalographique de la
maladie asthmatique. Correlations électroclinques-Hypothèses
pathogéniques. Presse Méd. 1952;3:1826–1828.
224. Lucioni R, Penati G. Sulla frequencza e sul significato in psichia-
tria dei tracciati cosidetti piatti. Riv Neurol. 1966;36:200–208.
225. Chatrian GE. The low voltage EEG. In: Remond A, ed. Handbook
of Electroencephalography and Clinical Neurophysiology. Vol 6A.
Amsterdam: Elsevier; 1976:77–89.
226. Walter WG, Dovey VJ. Electroencephalography in cases of sub-
cortical tumour. J Neurol Neurosurg Psychiatry. 1944;7:57–65.
227. Rugg MD, Dickens AMJ. Dissociation of alpha and theta activity
as a function of verbal and visuospatial tasks. Electroencephalogr
Clin Neurophysiol. 1982;53:201–207.
228. Rinaldi F, Himwich HE. Cholinergic mechanisms involved in
function of mesodiencephalic activating system. Arch Neurol
Psychiatry. 1955;73:396–402.
■ Normal EEG and Sleep: Adults and Elderly 211
229. Riekkinen P, Buzsaki G, Riekkinen P Jr, et al. The cholinergic sys-
tem and EEG slow waves. Electroencephalogr Clin Neurophysiol.
1991;78:89–96.
230. Van Huffelen AC, Magnus O. 4 c/sec vertex spindles. Electro-
encephalogr Clin Neurophysiol. 1973;34:543–546.
231. Daoust-Roy J. A waking 4 Hz vertex rhythm: 4 cps vertex spindles
revisited. Am J EEG Technol. 1989;29:147–163.
232. Arellano AP, Schwab RS. Scalp and basal recording during men-
tal activity. In: Proceedings of the First International Congress of
Psychiatry, Paris, 1950.
233. Brazier MAB, Casby JU. Crosscorrelation and autocorrelation
studies of electroencephalographic potentials. Electroencephalogr
Clin Neurophysiol. 1952;4:201–211.
234. Ishihara T, Yoshii N. Multivariate analytic study of EEG and
mntal activity in juvenile delinquents. Electroencephalogr Clin
Neurophysiol. 1972;33:71–80.
235. Mizuki Y. Frontal midline theta activity during performance of
mental tasks. Electroencephalogr Clin Neurophysiol. 1982;54:25P
[Abstract].
236. Mizuki Y. Frontal lobe: mental functions and EEG. Am J EEG
Technol. 1987;27:91–101.
237. Mizuki Y, Tanaka O, Isozaki H, et al. Periodic appearance of theta
rhythm in the frontal midline during performance of a mental
task. Electroencephalogr Clin Neurophysiol. 1980;49:345–351.
238. Mizuki Y, Takii O, Nishijima H, et al. The relationship between
the appearance of frontal midline theta activity (Fm theta) and
memory function. Electroencephalogr Clin Neurophysiol.
1983;56:56P [Abstract].
239. Takahashi N, Shinomiya S, Mori D, et al. Frontal midline theta
rhythm in young healthy adults. Clin Electroencephalogr.
1997;28:49–54.
240. Gastaut Y. Un signe électroencéphalographique peu connu: les
pointes occipitales survenant pendant l’ouverture des yeux. Rev
Neurol (Paris). 1951;84:640–643.
241. Cobb WA, Pampiglione G. Occipital sharp waves responsive to
visual stimuli. Electroencephalogr Clin Neurophysiol. 1952;4:110–111.
242. Evans CC. Comments on “occipital sharp waves responsive to
visual stimuli.” Electroencephalogr Clin Neurophysiol. 1952;4:111.
243. Evans CC. Spontaneous excitation of the visual cortex and associ-
ation areas—lambda waves. Electroencephalogr Clin Neurophysiol.
1953;5:69–74.
244. Roth M, Green J. The lambda wave as a normal physiological
phenomenon in the human electroencephalogram. Nature
(Lond). 1953;172:864–866.
245. Chatrian GE. The lambda waves. In: Remond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 6A.
Amsterdam: Elsevier; 1976:123–149.
246. Marton M, Szirtes J, Donauer N, et al. Averaged lambda poten-
tials in man and monkey. Electroencephalogr Clin Neurophysiol.
1982;53:14P–15P [Abstract].
247. Green J. Some observations on lambda waves and peripheral stim-
ulation. Electroencephalogr Clin Neurophysiol. 1957;9:691–704.
248. Barlow JS. Some statistical characteristics of electrocortical activ-
ity in relation to visual oculomotor tracking in man. Bol Inst
Estud Med Biol (Mex). 1963;21:497–518.
249. Barlow JS. Evoked responses in relation to visual perception and ocu-
lomotor reaction times in man. Ann NY Acad Sci. 1964;112:432–467.
250. Barlow JS, Ciganek L. Lambda responses in relation to visual
evoked responses in man. Electroencephalogr Clin Neurophysiol.
1969;26:183–192.
251. Remond A, Lesèvre N. Étude de la fonction du regard par
l’analyse des potentiels évoqués visuels (lumière, patterns
212 Part II ■ Normal EEG
complexes et mouvements des yeux). In: Dubois Poulsen A, Lairy
GC, Remond A, eds. La Fonction du Regard. Paris: Colloques de
l’INSERM; 1971:127–175.
252. Remond A, Lesèvre N, Torres F. Étude chronotopographique de
l’activité occipitale moyenne recueilli sur le scalp chez l’homme
en rélation avec le déplacement du regard (complexe lambda).
Rev Neurol (Paris). 1965;113:193–226.
253. Scott DF, Bickford RG. Electrophysiologic studies during scan-
ning and passive eye movements in humans. Science.
1967;155:101–102.
254. Bickford RG, Klass DW. Eye movements and the electroen-
cephalogram. In: Bender MB, ed. The Oculomotor Systems. New
York, NY: Hoeber; 1964:293–302.
255. Scott DF, Groethuysen UC, Bickford RG. Lambda responses in
the human electroencephalogram. Neurology (Minneapolis).
1967;17:770–778.
256. Green J. Lambda waves in the EEG and their relation to some
visual tasks. Proc EPTA. 1954;5:3.
257. Lesèvre N. Étude des reponses moyennes récueillies sur la région
postérieure du scalp chez l’homme au cours de l’exploration
visuelle (“complexe lambda”). Psychol Franç. 1967;12:26–36.
258. Chatrian GE, Bickford RG, Petersen MC, et al. Lambda waves
with depth electrodes in humans. In: Ramey ER, O’Doherty DS,
eds. Electrical Studies on the Unanesthetized Brain. New York, NY:
Hoeber; 1960:291–310.
259. Billings RJ. The origin of the occipital lambda wave in man.
Electroencephalogr Clin Neurophysiol. 1989;72:95–113.
260. Rechtschaffen A, Kales A. A Manual of Standardized
Terminology, Techniques, and Scoring System for Sleep Stages of
Human Subjects. University of California, LA: Brain Information
Service; 1968.
261. Kubicki S, Herrmann WM. The future of computer-assisted
investigation of the polysomnogram: sleep-microstructure. J Clin
Neurophysiol. 1996;13:285–294.
262. Kubicki St, Meyer Ch, Rohmel J. Dies Schlafspindelperiodik. Z
EEG-EMG. 1986;17:55–61.
263. Kubicki St, Höller L, Pastelak-Price C. Subvigil beta activity: a
study of fast EEG patterns in drowsiness. Am J EEG Technol.
1987;27:15–31.
264. Gastaut H, Gomez-Almanzar M, Taury M. Der provozierte
Mittagsschlaf: eine einfache, erfolgreiche Methode zur
Schlafaktivierung bie Epileptikern. Z EEG-EMG. 1983;14:6–11.
265. Gastaut H, Gomez-Almanzar M, Taury M. The enforced nap: a
simple effective method of inducing sleep activation in epileptics.
In: Degen R, Niedermeyer E, eds. Epilepsy, Sleep and Sleep
Deprivation. Amsterdam: Elsevier; 1984:75–83.
266. Niedermeyer E, Höller L. Kurzschlaf im EEG-Eine Fundgrube sonst
übersehener EEG-Abnormitäten. Z EEG-EMG. 1984;15:57–66.
267. Bennett DR. Sleep deprivation and major motor convulsions.
Neurology (Minneapolis). 1963;13:953–958.
268. Bennett DR, Mattson RH, Ziter FA, et al. Sleep deprivation: neu-
rological and electroencephalographic effects. Aerosp Med.
1964;35:888–890.
269. Mattson RH, Pratt KL, Calverley JR. Electroencephalograms of
epileptics following sleep deprivation. Arch Neurol (Chicago).
1965;13:310–315.
270. Oller-Daurella L. La privation de sommeil comme méthode d’ac-
tivation de l’EEG chez l’épileptique. Rev Neurol (Paris).
1966;115:530–535.
271. Pratt KL, Mattson RH, Weikers NJ, et al. EEG activation of
epileptics following sleep deprivation: a prospective study of 114
cases. Electroencephalogr Clin Neurophysiol. 1968;24:11–15.
272. Wittenbecker H, Kubicki S. Statistische Auswertung von 719
Kurzschlafableitungen nach Schlafentzug. Z EEG-EMG. 1976;8:
205–209.
273. Rumpl E, Lorenzi E, Bauer G, et al. Zum diagnostischen Wert des
EEG nach Schlafentzug. Z EEG-EMG. 1977;8:205–209.
274. Deisenhammer E, Klingler D. Das Schlafentzug-EEG bei
Kranken mit gesicherter und fraglicher Epilepsie sowie bei
Kranken ohne epileptische Anfälle. Z EEG-EMG. 1978;9:38–42.
275. Degen R. A study of the diagnostic value of waking and sleep EEGs
after sleep deprivation in epileptic patients on anticonvulsive ther-
apy. Electroencephalogr Clin Neurophysiol. 1980;49: 577–584.
276. Jovanovic UJ. General considerations of sleep and sleep depriva-
tion. In: Degen R, Rodin EA, eds. Epilepsy, Sleep and Sleep
Deprivation. 2nd ed. Amsterdam: Elsevier; 1991:205–215.
277. Kubicki S, Scheuler W, Wittenbecher H. Short-term sleep EEG
recordings after partial sleep deprivation as a routine procedure
to uncover epileptic phenomena: an evaluation of 719 EEG
recordings. In: Degen R, Rodin EA, eds. Epilepsy, Sleep and Sleep
Deprivation. 2nd ed. Amsterdam: Elsevier; 1991:217–230.
278. Klingler D, Tragner H, Deisenhammer E. The nature of the influ-
ence of sleep deprivation on the EEG. In: Degen R, Rodin EA,
eds. Epilepsy, Sleep and Sleep Deprivation. 2nd ed. Amsterdam:
Elsevier; 1991:231–233.
279. Naitoh P, Dement W. Sleep deprivation in humans. In: Remond
A, ed.-in-chief. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 7A. Amsterdam: Elsevier; 1976:146–151.
280. Rodin EA, Luby ED, Gottlieb S Jr. The electroencephalogram
during prolonged experimental sleep deprivation. Electro-
encephalogr Clin Neurophysiol. 1962;14:544–551.
281. Heinemann LG. Der Mehrtägige Schlafentzug in der experi-
mentellen Psychoseforschung: psychopathologie und EEG. Arch
Psychiatr Nervenkr. 1966;208:117–197.
282. Johnson LC. Psychological and physiological changes following
total sleep deprivation. In: Kales A, ed. Sleep: Physiology and
Pathology. Philadelphia: JB Lippincott; 1969:206–220.
283. Jovanovic UJ, Liebaldt GP, Mühl M, et al. Schlafentzug and seine
Begleiterscheinungen. Arch Psychiatr Nervenkr. 1971;124: 183–202.
284. Naitoh P, Kales A, Kollar EJ, et al. Electroencephalographic
changes after prolonged deprivation of sleep. Electroencephalogr
Clin Neurophysiol. 1969;27:2–11.
285. Welch LK, Stevens JB. Clinical value of the electroencephalogram
following sleep deprivation. Aerosp Med. 1971;42:349–351.
286. Klass DW, Fischer-Williams M. Sensory stimulation, sleep and
sleep deprivation. In: Remond A, ed.-in-chief. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 3D.
Amsterdam: Elsevier; 1976:5–73.
287. Aserinsky E, Kleitman N. Regularly occurring episodes of eye
mobility and concomitant phenomena during sleep. Science.
1953;118:273–274.
288. Dahlström A, Fuxe K. Evidence for the existence of monoamino-
containing neurons in the cerebral nervous system. I.
Demonstration of monamines in the cell bodies of brain stem
neuron. Acta Physiol Scand Suppl. 1964;232:1–55.
289. Jouvet M. Biogenic amines and the states of sleep. Science.
1969;163:32–41.
290. Jouvet M. Some monoaminergic mechanisms controlling sleep
and waking. In: Karczmar AG, Eccles JC, eds. Brain and Human
Behavior. New York, NY: Springer; 1972:131–161.
291. Jouvet M, Delorme F. Locus coeruleus et sommeil paradoxal. CR
Soc Biol (Paris). 1965;159:895–899.
292. Ungerstedt U. Stereotaxic mapping of the monoamine pathway
in the rat brain. Acta Physiol Scand Suppl. 1971;367:1–48.
 Chapter 11
293. Dement W, Kleitman N. Cyclic variations in EEG during sleep
and their relation to eye movements, body motility and dream-
ing. Electroencephalogr Clin Neurophysiol. 1957;9:673–690.
294. Erwin CW, Somerville ER, Radtke RA. A review of electroencephalo-
graphic features normal of sleep. Clin Neurophysiol. 1984;1:253–274.
295. Guilleminault C, Baker TL. Sleep and electroencephalography:
points of interest and points of controversy. J Clin Neurophysiol.
1984;1:275–291.
296. Schieber JP, Muzet A, Ferrière PJR. Les phases d’activation tran-
sitoire spontanées au cours du sommeil normal chez l’homme.
Arch Sci Physiol. 1971;25:443–464.
297. Terzano MG, Mancia D, Zacchetti O, et al. The significance of
cyclic EEG changes in Creutzfeldt-Jakob disease: prognostic value
of their course in 9 patients. Ital J Neurol Sci. 1981;3:243–254.
298. Terzano MG, Parrino L. The phenomena of cyclic alternating
pattern. Clin Neurol Neurosurg. 1987;89:164–167.
299. Terzano MG, Parrino L. Functional relationship between micro-
and macrostructure of sleep. In: Terzano MG, Halász P, Declerck
AC, eds. Phasic Events and Dynamic Organization of Sleep. New
York, NY: Raven Press; 1991:101–119.
300. Halász P. Information processing during sleep: introduction. In:
Koella WP, Obal F, Schulz H, et al., eds. Sleep 1986. Stuttgart:
Fischer; 1988:77–78.
301. Evans BM. Periodic activity in cerebral arousal mechanisms—the
relationship to sleep and brain damage. Electroencephalogr Clin
Neurophysiol. 1992;83:130–137.
302. Goldie L, Green JM. Paradoxical blocking and arousal in the
drowsy state. Nature (Lond). 1960;187:952–953.
303. Santamaria J, Chiappa KH. The EEG in Drowsiness. New York:
Demos; 1987.
304. Broughton R, Hasan J. Quantitative topographic electroen-
cephalographic mapping during drowsiness and sleep onset.
J Clin Neurophysiol. 1995;12:372–386.
305. Koshino Y, Nishio M, Murata T, et al. The influence of light
drowsiness on the latency and amplitude of P300. Clin
Electroencephalogr. 1993;24:110–113.
306. Wada Y, Nanbu Y, Koshino Y, et al. Inter-and intrahemispheric
EEG coherence during light drowsiness. Clin Electroencephalogr.
1996;27:24–88.
307. Liberson WT. Functional electroencephalography in mental dis-
orders. Dis Nerv Syst. 1945;5:357–364.
308. Davis PA. Effects of acoustic stimuli on the waking human brain.
J Neurophysiol. 1939;2:494–499.
309. Bancaud J, Bloch V, Paillard J. Contribution E.E.G. à l’étude des
potentiels évoqués chez l’homme au niveau de vertex. Rev Neurol
(Paris). 1953;89:399–418.
310. Gastaut Y. Les pointes négatives évoquées sur le vertex: leur sig-
nification psychophysiologique et pathologique. Rev Neurol
(Paris). 1953;89:382–389.
311. Kooi KA, Tipton AC, Marshall RE. Polarities and field configura-
tion of the vertex components of the human auditory evoked
response: a reinterpretation. Electroencephalogr Clin Neuro-
physiol. 1971;31:166–169.
312. Storm van Leeuwen W. Auditory evoked responses (AERs). In:
Remond A, ed.-in-chief. Handbook of Electroencephalography and
Clinical Neurophysiology. Vol 8A. Amsterdam: Elsevier; 1975:71–85.
313. Weitzman ED, Kremen H. Auditory evoked responses during dif-
ferent stages of sleep in man. Electroencephalogr Clin Neuro-
physiol. 1965;18:65–70.
314. Davis H, Mast T, Yoshie N, et al. The slow response of the human
cortex to auditory stimuli. Electroencephalogr Clin Neurophysiol.
1966;21:105–113.
■ Normal EEG and Sleep: Adults and Elderly 213
315. Niedermeyer E, Pribram HFW. Unilateral suppression of vertex
sharp waves in the sleep electroencephalogram (case report).
Electroencephalogr Clin Neurophysiol. 1966;20:401–404.
316. Wright EA, Gilmore RL. Features of the geriatric EEG: age-
dependent incidence of POSTS. Electroencephalogr Clin
Neurophysiol. 1984;58:49P [Abstract].
317. Prior PF, Deacon PA. Spontaneous sleep in healthy subjects in
long-term serial electroencephalograph readings. Electro-
encephalogr Clin Neurophysiol. 1969;27:422–424.
318. Roth M, Shaw J, Green J. The lambda wave as a normal physio-
logical phenomenon in the human electroencephalogram.
Nature (Lond). 1953;172:864–866.
319. Hess R. The electroencephalogram in sleep. Electroencephalogr
Clin Neurophysiol. 1964;16:44–55.
320. Arfel G, Laurette G. Occipital sharp waves of sleep contrasted
with lambda waves. Clin Electroencephalogr. 1973;4:16–32.
321. Kugler J, Laub M. Lambdoid-Wellen (RHO-Wellen) in Schlaf. In:
Jovanovic UJ, ed. The Nature of Sleep. Stuttgart: Fischer; 1973:
31–35.
322. Kugler J. Lambda waves and rho waves. Electroencephalogr Clin
Neurophysiol. 1976;40:193–194 [Abstract].
323. Vignaendra V, Matthews RL, Chatrian GE. Positive occipital
sharp transients of sleep: relationships to nocturnal sleep cycle in
man. Electroencephalogr Clin Neurophysiol. 1974;37:239–246.
324. Brenner RP, Zauel DW, Carlow TJ. Positive occipital sharp transients
of sleep in the blind. Neurology (Minneapolis). 1978;28:609–612.
325. Pristasova E, Prochazka M. Bioccipital theta rhythmic and positive
occipital sharp waves of sleep. Electroencephalogr Clin Neurophysiol.
1981;52:1P–2P [Abstract].
326. Niedermeyer E, Lentz WJ. Dreaming in non-REM sleep. Waking
Sleeping. 1976;1:49–51.
327. Dement WC. Some Must Watch While Some Must Sleep. San
Francisco: San Francisco Book Co; 1976.
328. Röschke J, Başar, E. The EEG is not a simple noise: strange attrac-
tors in intracranial structures. In: Başar E, ed. Chaos in Brain
Function. Berlin: Springer; 1990:49–62.
329. Niedermeyer E. The electroencephalogram and vertebrobasilar
artery insufficiency. Neurology (Minneapolis). 1963;13:412–422.
330. Engel GL, Romano J. Delirium, a syndrome of cerebral insuffi-
ciency. J Chron Dis. 1959;9:260 (Quoted after Kiloh LG, McComas
AJ, Osselton JW. Clinical Electroencephalography. London:
Butterworth; 1972).
331. Lorimer FM. Further clinical observations on exceedingly fast
activity, a rare electroencephalographic abnormality. Clin
Electroencephalogr. 1970;1:84–91.
332. Marshall L, Molle M, Schreiber H, et al. Scalp recorded direct
current potential shifts associated with the transition to sleep in
man. Electroencephalogr Clin Neurophysiol. 1994;91:346–352.
333. Luthringer R, Branderberger G, Schaltenbrand N, et al. Slow
wave electroencephalic activity parallels renin oscillations during
sleep in humans. Electroencephalogr Clin Neurophysiol. 1995;95:
318–322.
334. Loomis AL, Harvey EN, Hobart GA. Distribution of disturbance-
patterns in the human electroencephalogram with special refer-
ence to sleep. J Neurophysiol. 1938;1:413–430.
335. Scheuler W, Stinshoff D. Das Alpha-Schlafmuster-eine kaum
beachtete EEG-Variante. Z EEG-EMG. 1982;13:34–41.
336. Gaillard J-M, Baudat J, Blois R. The effect of sex depressive state
and minor tranquilizers on K potentials and spindles during
sleep. Clin Neurol Neurosurg. 1987;89:172–173.
337. Jankel WR, Niedermeyer E. Sleep spindles. J Clin Neurophysiol.
1985;2:1–35.
214 Part II ■ Normal EEG
338. Church MW, Johnson LC, Seales DM. Evoked K complexes and
cardiovascular responses to spindle-synchronous and spindle-
asynchronous stimulus clicks during NREM sleep.
Electroencephalogr Clin Neurophysiol. 1978;45:443–453.
339. Caderas M, Niedermeyer E, Uematsu S, et al. Sleep spindles
recorded from deep cerebral structures in man. Clin
Electroencephalogr. 1982;13:216–225.
340. Niedermeyer E, Jankel WR. Falling asleep: depth EEG and ther-
mographic observations. Electroencephalogr Clin Neurophysiol.
1984;58:8P [Abstract].
341. Niedermeyer E, Jankel WR, Uematsu S. Falling asleep: observa-
tions and thoughts. Am J EEG Technol. 1986;26:166–175.
342. Zobniw AM, Yarworth S, Niedermeyer E. Depth electroen-
cephalography. J Electrophysiol Technol. 1975;1:215–240.
343. Bremer F. Cerebral and cerebellar potentials. Physiol Rev. 1958;
38:357–388.
344. Andersen P, Eccles JC. Inhibitory phasing of neuronal discharge.
Nature (Lond). 1962;196:645–647.
345. Steriade M, Deschênes M. Intrathalamic and brainstem-thalamic
networks involved in resting and alert stages. In: Bentivoglio M,
Spreafico R, eds. Cellular Thalamic Mechanisms. Amsterdam:
Elsevier; 1988:37–62.
346. Gibbs EL, Gibbs FA. Extreme spindles: correlation of electroen-
cephalographic sleep pattern with mental retardation. Science.
1962;138:1106–1107.
347. Niedermeyer E, Singer HS, Folstein SE, et al. Hypersomnia with
simultaneous waking and sleep patterns in the electroencephalo-
gram. A case report with neurotransmitter studies. J Neurol
(Berlin). 1979;221:1–13.
348. Davis H, Davis PA, Loomis AL, et al. Electrical reactions of human
brain to auditory stimulation during sleep. J Neurophysiol. 1939;
21:500–514.
349. Brazier MAB. The electrical fields at the surface of the head dur-
ing sleep. Electroencephalogr Clin Neurophysiol. 1949;1:195–204.
350. Roth M, Green J. The form, voltage distribution and physiological
significance of the K complex. Electroencephalogr Clin Neurophysiol.
1956;8:385–402.
351. Niiyama Y, Satoh N, Kutsuzawa O, et al. Electrophysiological evi-
dence suggesting that sensory stimuli of unknown origin induce
spontaneous K complex. Electroencephalogr Clin Neurophysiol.
1996;98:394–400.
352. Ujszaszi J, Halász P. Long latency evoked potential components in
human slow wave sleep. Electroencephalogr Clin Neurophysiol.
1988;69:516–522.
353. Niiyama Y, Fushimi M, Sekine A, et al. K complex evoked in NREM
sleep is accompanied by a slow negative potential related to cogni-
tive process. Electroencephalogr Clin Neurophysiol. 1995;95:27–33.
354. Halász P, Pal I, Rajna P. Relationship between spontaneous and
evoked K complexes. Electroencephalogr Clin Neurophysiol.
1982;53:7P [Abstract].
355. Paiva T, Rosa A. The K complex variability in normal subjects. In:
Terzano MG, Halász P, Declerck AC, eds. Phasic Events and
Organization of Sleep. New York: Raven Press; 1991:167–187.
356. Weisz J, Soroker N, Oksenberg A, et al. Effects of hemi-thalamic
damage on K-complexes evoked by monaural stimuli during
mid-afternoon sleep. Electroencephalogr Clin Neurophysiol.
1995;94:148–150.
357. Heilman KM, Van den Abell T. Right hemispheric dominance for
mediating cerebral activation (quoted after Weisz et al., 1995).
Neuropsychologia. 1979;17(3–4):315–321.
358. Heilman KM, Schwartz HD, Watson RT. Hypoarousal in patients
with neglect syndrome and indifference. Neurology.
1978;28:229–232.
359. Tinuper P, Montagna P, Medori R, et al. The thalamus participates
in the regulation of the sleep-waking cycle. A clinico-pathological
study in fatal familial thalamic degeneration. Electroencephalogr
Clin Neurophysiol. 1998;73:117–123.
360. Leemhuis AJ, Stamps F. Mitten patterns in electroencephalogram
of patients with Parkinsonism. In: Proceedings of the American
EEG Society. Atlantic City, NJ; 1949.
361. Lyketsos G, Belinson L, Gibbs FA. Electroencephalograms of
nonepileptic psychotic patients awake and asleep. Arch Neurol
Psychiatry (Chicago). 1953;69:707–712.
362. Winfield DL, Sparer PJ. The electroencephalogram in paralysis
agitans. Dis Nerv Syst. 1954;15:144–120.
363. Halasz P, Kajtor F. "Mittens"—a new form of electroencephalo-
graphic wave. Ideggyogy Szemle (Hungarian). 1962;15:46–57
(quoted after Gibbs FA, Gibbs EL, 1964).
364. Gibbs FA, Gibbs EL. The mitten pattern: an electroencephalo-
graphic abnormality correlating with psychosis. J Neuro-
psychiatry. 1963;5:6–13.
365. Scheuler W, Stinshoff D, Kubicki St. The alpha sleep pattern.
Differentiation from other sleep patterns and effects of hyp-
notics. Neuropsychobiology. 1983;10:395–399.
366. Scheuler W, Rappelsberger P, Schmatz F, et al. Periodicity analysis
of sleep EEG in the second and minute ranges—example of appli-
cation in different alpha activities in sleep. Electroencephalogr Clin
Neurophysiol. 1990;76:222–234.
367. Hauri P, Hawkins DR. Alpha delta sleep. Electroencephalogr Clin
Neurophysiol. 1973;34:233–237.
368. McNamara ME. Alpha sleep: a mini review and update. Clin
Electroencephalogr. 1993;24:192–193.
369. Moldofsky H, Lue FA. The relationship of alpha and delta EEG
frequencies to pain and mood in “fibrositis” patients treated with
chlorpromazine and I-tryptophan. Electroencephalogr Clin
Neurophysiol. 1980;50:71–80.
370. Scheuler W, Kubicki St, Marquardt J, et al. The alpha-sleep-pat-
tern-quantitative analysis and functional aspects. In: Koella WP,
Obal F, Schulz H, et al., eds. Sleep 1986. Stuttgart: Fischer; 1988.
371. Sassin J, Parkee D, Mace J, et al. Human growth hormone release
relation to slow-wave sleep and sleep-waking cycle. Science.
1969;165:513–515.
372. Gastaut H, Broughton R. A clinical and polygraphic study of
episodic phenomena during sleep. In: Wortis J, ed. Recent
Advances in Biological Psychiatry. Vol VII. London: Plenum Press;
1964:197–221.
373. Banquet JP. Inter- and intrahemispheric relationships of the EEG
activity during sleep in man. Electroencephalogr Clin Neuro-
physiol. 1983;55:51–59.
374. Salzarulo P. Etude é’lectroencéphalographique et polygraphique
du sommeil d’aprés-midi chez le sujet normal. Electroencephalogr
Clin Neurophysiol. 1971;30:399–407.
375. Passouant P. EEG and sleep. Electro-clinical semeiology. In:
Remond A, ed.-in-chief. Handbook of Electroencephalography
and Clinical Neurophysiology. Vol 7A. Amsterdam: Elsevier; 1975:
3–11.
376. Schwartz RA. EEG et mouvements oculaires dans le sommeil de
nuit. Electroencephalogr Clin Neurophysiol. 1962;14:126–128.
377. Roth C, Achermann P, Borbely AA. Alpha activity in the human
REM sleep topography and effect of REM sleep deprivation. Clin
Neurophysiol. 1999;110:632–635.
378. Matsuo F. Recognition of REM sleep in standard EEG.
Electroencephalogr Clin Neurophysiol. 1981;52:490–493.

Activation Methods
TAKEO TAKAHASHI AND KEITH H. CHIAPPA
Various activation procedures—such as hyperventilation, inter-
mittent photic stimulation, or sleep—may be used to enhance
preexisting abnormalities and induce abnormal findings in an
otherwise normal electroencephalogram (EEG). This chapter
discusses common activation methods as well as methods that
are less frequently carried out in EEG laboratories. Sleep-
induced, sleep-deprived, and especially psychotropic drug-
induced effects are not discussed here; they are covered in other
chapters.
HYPERVENTILATION
Precipitation of seizures by hyperventilation (HV) (1) was
known prior to the discovery of the human EEG. Therefore,
EEG activation by HV has been widely used in almost all clin-
ical EEG laboratories since its introduction, and Nims et al. (2)
showed the provocative efficacy of generalized synchronous
paroxysmal discharges and of absence seizures produced by
this test. This method consists of deep and regular respiration
at a rate of about 20/min for a period of 2 to 4 minutes. In
adults, such HV will cause air exchange of 20 to 50 L/min and
a drop in partial pressure of carbon dioxide (PCO2) in the
range of 4 to 7 mm Hg (3).
An understanding of the time course and magnitude of the
systematic changes in blood gases produced by routine HV in
adults is important for clinical interpretations of EEGs.
Achenbach-Ng et al. (4) studied these changes in adults and
children and found that during 3 minutes of HV, the PCO2 fell
a mean of 18 mm Hg and the partial pressure of oxygen (PO2)
rose 7 mm Hg, the former with a near-linear response curve
(see Fig. 12.1). Note that PCO2 reached its nadir 30 seconds
after the end of the 3 minutes of HV. From this point, after the
command to “breathe normally,” PCO2 rose linearly to return
to resting levels in about 5.5 minutes, more slowly for children.
As PCO2 rose during this post-HV period, PO2 fell to 25 mm
Hg below resting values at 5.2 minutes after cessation of HV,
and then rose to resting levels over the following 11.6 minutes
(Fig. 12.1). Any perturbation of ventilation in the post-HV
period (e.g., talking) prolonged these steps. These findings min-
imize the clinical importance of slowing in the EEG that out-
lasts the HV period, and correlate with the “rebuildup” of
post-HV slowing seen in moyamoya disease (see below).
The characteristic EEG response to HV, most prominent in
children, consists of a fluctuating increase of bilaterally syn-
chronous slow activity and slowing of alpha and beta rhythms.
In normal adults, although the slowing is generally not marked,
there are wide differences among individuals.
CHAPTER
12
With regard to the underlying mechanism of HV’s provoca-
tion of slowing and seizure discharges, Gibbs et al. (5) suggested
that diffuse slowing is caused by inadequate compensatory
vasoconstriction of the cerebrum in response to systemic
hypocapnia. According to Gotoh et al. (6), diffuse slowing is
considered to be the direct result of cerebral ischemic anoxia
resulting from hypocapnic cerebral vasoconstriction. Yamatani
et al. (7), measuring cerebral blood flow in the right carotid
artery during HV, reported that decreased PCO2 and cerebral
blood flow were the fundamental factors causing EEG slowing.
Fisch and So (8) summarized the physiologic basis of the EEG
response to HV as the following: alteration of PCO2, rather
than pH or PO2, is the most important factor in producing the
EEG response to HV, whereas the most obvious and dramatic
physiologic effect of HV is decreased cerebral blood flow.
Sherwin (9) attributed the diffuse slowing to synchronous
activity in the nonspecific thalamocortical projecting systems,
which become more active in hypocapnia. Patel and Maulsby
(10) raised the possibility that hypocapnia induced by HV
decreased activity in mesencephalic reticular formation, which
then caused EEG slowing, just as drowsiness and sleep produce
EEG slowing. Similar mechanisms to those described above are
considered to be important for provoking paroxysmal dis-
charges (11); hyperexcitability of neurons may be induced by
respiratory alkalosis (12). However, the change from the pure
delta response to the appearance of intermixed spikes and clear-
cut spike–wave discharges remains obscure. Interestingly, it has
been reported that intravenous administration of diazepam
prevented HV-induced spike patterns, though it failed to atten-
uate the delta response (13).
The magnitude of the HV response depends on a number of
factors (14). First, vigorous exchange of air can enhance activa-
tion effects. For the purpose of routine examination with HV,
however, it is suggested that the rate of breathing be as close as
possible to that of the resting rhythm (15 to 20 breaths/min)
(14). Second, age is an important factor. Slow waves appear much
more abruptly, are more pronounced, and persist longer in chil-
dren than in adults. The degree and abruptness of the response
are related directly to age (15). The most dramatic EEG responses
to HV usually occur between the ages of 8 and 12 years (16).
According to Gibbs et al. (5), in children between 3 and 5 years of
age, 97% of epilepsy patients and 70% of normal subjects showed
diffuse slowing with HV; after 20 years of age, more than 40% of
epilepsy patients showed diffuse slowing, whereas it was seen in
less than 10% of normal subjects. The areas most affected by HV
in children are occipitotemporal regions (17,18). Delta waves
tend to appear initially in the posterior regions and spread
215
216 Part II ■ Normal EEG
Figure 12.1 Time course and magnitude of absolute PCO2
and PO2 changes with 3 minutes of hyperventilation in nine
normal adult subjects. The error bars are 1 standard devia-
tion. A transcutaneous heated membrane technique was
used for the blood gas measurements.
forward in the younger age group, whereas they tend to appear in
the frontal regions and spread backward in the older age group
(19). Third, blood glucose level is also important in determining
the degree of HV response. In adults, a low blood glucose level
(less than 80 mg/dL) tends to enhance appearance of slow waves;
a high level (more than 120 mg/dL) tends to inhibit or prevent
such an effect (20). Delta waves induced by HV can occasionally
be the first indication of pathologic hypoglycemia secondary to
an islet cell tumor in a patient referred for an EEG (19). Fourth,
an erect position as compared to a reclining position enhances
the effect of HV; EEG slowing occurs earlier and with greater
intensity. This is thought to be a result of relative cerebral anoxia
(21). Diffuse slowing induced by HV usually disappears rapidly
after ceasing HV; it may persist for up to 30 seconds in normal
adults, but even further persistence at any age, by itself, is not nec-
essarily an abnormality.
Aside from diffuse slowing, HV may induce diffuse sharp
waves or spike–wave discharges of epileptogenic significance; it
is particularly effective in eliciting bilaterally synchronous
spike–wave discharges in a patient with generalized epilepsies.
Dalby (22) found that patients with absence seizures were more
sensitive to HV than patients with nonabsence seizures, with the
incidence of spike–wave paroxysms being 50% in the former
and 25% in the latter. During HV, not only absence seizures but
also complex partial seizures are induced in epilepsy patients. In
general, latent abnormalities are likely to be activated by HV.
Slow-wave foci associated with localized lesions may also be
aggravated; abnormalities in temporal regions are more prone to
accentuation than those elsewhere (23). Miley and Forster (24)
reported that vigorous HV of longer duration induced abnormal
discharges with or without clinical seizures in 11% of epilepsy
patients with complex partial seizures. Arain et al. (25) have sug-
gested that, because of the high incidence of significant clinical
and EEG events provoked by HV, this activation procedure
should be performed on a daily routine basis in patients in
epilepsy monitoring units.
Unusually prolonged post-HV high-voltage slowing may be
seen in patients with syncopal attacks of various etiologies (26).
Buildup of slow waves after the end of HV (“re-buildup”) may
be a diagnostic finding in children with moyamoya disease (27).
In a study that addressed three cases of moyamoya disease,
Kameyama et al. (28) reported that this EEG re-buildup was
noted about 5 minutes after HV cessation; moreover, correspon-
ding decreases in cerebral blood flow and PO2 were documented
with it. Such a post-HV hypoxia 5 minutes after cessation of
HV was also found in a study of nine normal adult subjects (4)
(Fig 12.1). Note that it is suggested that HV should be avoided
in patients with moyamoya disease (29).
The American EEG Society Guidelines (AEEGS) Committee
(30) proposed the following minimum technical requirements:
HV should be used routinely unless other medically justifiable
reasons (e.g., a recent intracranial hemorrhage, significant car-
diopulmonary disease, sickle cell disease or trait, or patient
inability or unwillingness to cooperate) contraindicate it. It
should be performed for a minimum of 3 minutes. At times,
HV must be performed for a longer period in order to obtain
adequate EEG activation. The guidelines suggest continued
recording for at least 1 minute after cessation of overbreathing,
but the slow time course of the recovery of PCO2 levels dis-
cussed above and the findings in the moyamoya patients indi-
cate that at least 6 minutes of post-HV recording would be
prudent.
Regarding the clinical interpretation of EEG changes seen
during HV, a few simple principles should be kept as guidelines,
and any clinical interpretation that exceeds these guidelines
should have a firm, rational basis for the variance. These guide-
lines are: HV-induced slowing may be of any amplitude, loca-
tion, rhythmicity, and time course and should still be
considered to be within normal limits unless it is significantly
asymmetrical or contains definite spikes.
As with most amplitude asymmetry considerations in clini-
cal EEG interpretations, there are no precise values available for

normal limits in HV-induced slowing, but a 2:1 ratio is defi-
nitely abnormal and any asymmetry greater than 1.5:1 is prob-
ably abnormal. HV-induced slowing can persist for several
minutes after the command to breathe normally, as discussed
above, and this persistence or reappearance is only very rarely
abnormal.
Regarding a clinical interpretative declaration stating that
there are HV-induced spikes in the EEG, a very conservative
approach needs to be employed. As in most other sections of
the EEG, the admixture and interaction of different, nonepilep-
tiform frequencies can sum to produce a waveform with an
epileptiform appearance that is not a true spike with the patho-
physiologic significance of that cortical event. This is especially
true with HV-induced slowing and its superimposed faster fre-
quencies, and even more so in childhood since that age has the
most complex mixture of frequencies and the highest ampli-
tudes of all of those components, including the HV-induced
slowing. In the pediatric age group both paroxysmal drowsy
hypersynchrony and HV-induced bursts of slowing may con-
tain admixed components with a spike-like appearance that are,
despite their superficial appearance, normal variants. This dif-
ferentiation constitutes the major challenge in the interpreta-
tion of pediatric EEGs.
In addition, the presence of altered responsiveness in associ-
ation with high-amplitude rhythmic slowing (HIHARS) with
HV is not abnormal (31). In the latter study, 8 of 12 healthy
nonepileptic children (mean age 9.6 years) exhibited impaired
verbal recall and failed to respond to repeated auditory clicks.
Furthermore, Lum et al. (32) studied 77 episodes of HIHARS
with loss of awareness from 22 children and 107 absence
seizures during HV from another 22 children; eye opening and
eyelid flutter were seen more frequently in absence seizures,
whereas fidgeting, smiling, and yawning occurred more fre-
quently during HIHARS episodes, and arrest of activity, staring,
and oral and manual automatisms were observed in both
groups, so that the EEG is the only reliable means of differenti-
ating the two conditions, that is, epileptic versus nonepileptic.
INTERMITTENT PHOTIC STIMULATION
The effect of intermittent photic stimulation (IPS) on the
human EEG was first studied by Adrian and Matthews (33).
They used sinusoidal IPS derived from a constant light source
in front of which a disc with cutout sectors was rotated. After
this earliest report, similar instruments were used for more than
a decade (34). Walter et al. (35) were the first to report activa-
tion of paroxysmal discharges by IPS with an electronic strobe
light. After this pioneering work, the method of IPS using a
strobe light became popular (36). The strobe lamp is placed at
a distance of 20 to 30 cm in front of the subject’s eyes. IPS is
commonly presented through closed eyelids, but it can be
administered with the eyes open. Details of IPS use in routine
examination vary greatly between EEG laboratories. The fol-
lowing protocol is suggested by Bickford (19): flashes at fre-
quencies of 1, 3, 6, 9, 10, 15, 20, and 30 Hz each are given in
trains of 5-second duration with eyes open and closed in a
room with reduced illumination. It is sometimes advantageous
Chapter 12 ■ Activation Methods 217
to have patients open and close their eyes during the stimula-
tion. In a standardization of screening methods for photosensi-
tivity (37), the following IPS frequencies are used: 1, 2, 3, 4, 6, 8,
10, 12, 14, 16, 18, 20, 60, 50, 40, 30, and 25 Hz (8). The three
main EEG changes induced by IPS are explained below.
With respect to the duration of photic stimulation in the
clinical EEG laboratory, it should be noted that the technologist
must be alert to stop stimulation immediately on identifying a
photoparoxysmal response (PPR) since prolonged stimulation
can trigger a generalized tonic–clonic seizure (GTCS) in a sen-
sitive patient. Also, in these patients, once a PPR has been defi-
nitely identified, further stimulation should not be attempted as
these patients may become more sensitive as stimulation con-
tinues or is repeated, and may launch into a seizure with little
warning. The additional information gained from repeated
stimulation is not worth the risk to the patient.
PHOTIC DRIVING RESPONSE
This is a physiologic response consisting of rhythmic activity
elicited over the posterior regions of the head by IPS frequen-
cies of about 5 to 30 Hz. The term photic driving response (PDR)
should be limited to activity time-locked to the stimulus and of
a frequency identical or harmonically related to the stimulus
frequency (38).
As a rule, PDR is found over posterior regions. In infants,
PDR can be elicited a few hours after birth (39), but PDR
remains relatively small up to about 6 years of age (40). In older
children, PDR becomes much larger, particularly at low fre-
quencies (23). The amplitude of PDR is usually higher in chil-
dren than in adults and again tends to increase in elderly
people. Regardless of age, however, an exaggerated PDR to low
flash frequencies (0.5 to 3 Hz) usually signifies acute or suba-
cute neuronal dysfunction (41); examples include MELAS
(mitochondrial myopathy, encephalopathy, lactic acidosis, and
stroke-like episodes) (8) and the late infantile form of ceroid
lipofuscinosis (42,43).
According to Fisch and So (8), large positive occipital sharp
transients of sleep (POSTS) and lambda waves in response to
scanning a complex pattern are predictive of a prominent PDR.
Destructive cortical lesions may cause unilateral PDR depres-
sion, whereas irritative lesions, such as those associated with
epileptic scars, may produce increased PDR on the side of the
lesion (19). Interpretation suggesting abnormal PDR should be
made carefully because a minor and inconsistent asymmetry of
PDR is infrequently seen even in normal subjects; Coull and
Pedley (44) claim that an asymmetry in amplitude only, in the
absence of other EEG changes, should not be viewed as abnor-
mal. Very low voltage or absence of PDR is of little diagnostic
significance because some normal persons are not responsive to
IPS (41). According to Chiba et al. (45), diffuse PDR could be
observed in 25% of the patients who were receiving hemodial-
ysis treatment.
It has been reported that PDR elicited by 5-Hz IPS with eyes
closed is similar to that elicited by a red (saturated long-
wavelength red) 5-Hz flicker stimulation, whereas PDR elicited
by 5-Hz IPS with eyes open is similar to that elicited by a
218 Part II ■ Normal EEG

5-Hz flickering dot pattern (46). In that case, decreased lumi-
nance, such as 20 candela (cd)/m 2, was used. (This luminosity
is very low compared to that of a stroboscopic IPS, for example,
Grass photostimulator (intensity 8): 3939 cd/m 2 (47).)
Furthermore, using a 5-Hz red flicker stimulation and, to a
greater extent, a 5-Hz flickering dot pattern provided by a visual
stimulator (48), high-amplitude PDR over 50 V was elicited
more frequently. Such an excitatory effect of a dot pattern or
other similar patterns on PDR (49) can be demonstrated easily
by placing the pattern on the strobe light when IPS is given to
the subject with eyes open (50).
Studies of PDR utilizing 5-Hz red flicker and 5-Hz flickering
dot pattern (6 mm in diameter) stimuli by means of visual
stimulators (47,51) yielded the following results (50). High-
amplitude PDR is more frequently elicited using either a red
flicker stimulation or a flickering dot pattern stimulation, irre-
spective of diagnostic group, than a white flicker stimulation
with equal luminance, such as 20 cd/m 2. High-amplitude PDRs
elicited by flickering dot pattern and red flicker stimuli in adult
patients suggest the presence of an occipital lobe disturbance.
Incidence of high-amplitude PDR elicited by a flickering dot
pattern stimulation in adult patients is 2.4 times higher in
female patients than in male patients. A study of PDR in adult
psychiatric outpatients showed that PDR amplitudes increased
significantly with patient age (52). Figure 12.2 depicts EEG
changes in response to 5-Hz flicker stimuli in a young patient
with occipital lobe epilepsy: stimuli using white and red light
elicited 5- and 10-Hz PDRs, whereas those by dot pattern
(6 mm in diameter; 0.5 cycles/degree [c/deg]) elicited 5-Hz
high-amplitude PDRs. Stimuli using dot patterns (0.5 mm:
4.9 c/deg; 1 mm: 2.1 c/deg; 2 mm: 1.5 c/deg; 4 mm: 0.8 c/deg)
also showed a similar finding as that shown in Figure 12.2. As
for high-amplitude PDRs observed in elderly individuals, they
tend to be elicited by stimuli using moderate (2 mm in diame-
ter) to large dot patterns (4 and 6 mm in diameter) as demon-
strated in Figure 12.3. Namely, when dot patterns with lower
spatial frequencies are employed, high-amplitude PDR tends to
be elicited in elderly patients (53), whereas its incidence
increases in younger patients by use of dot patterns with higher
spatial frequencies.
Using pictures such as a Lily, the 5-Hz flickering picture
stimulation also elicited a high-amplitude PDR similar to that
evoked by a flickering dot pattern (54). These PDRs elicited by
visual stimuli not only to the full field, but also to the center,
periphery, and each hemifield (regional visual stimulation)
provide us with useful clinical information (55). Figure 12.4
shows results of 5-Hz fundamental PDRs elicited by regional
visual stimulation in a total of 54 adult psychiatric outpatients,
from whom patients with epilepsy and cerebral vascular disor-
ders were excluded. We made a power spectral analysis on

Figure 12.2 EEG changes in response to 5-Hz white flicker, red flicker, and flickering dot pattern (6 mm in diameter) stim-
uli to the full field (30 30 ) in a 20-year-old woman with occipital lobe epilepsy. Stimulus was provided by use of a
square-type strobe-filter method (92); luminance was maintained at 20 cd/ m 2. Lower tracings show mapping analysis of
photic driving responses (PDRs) elicited by each flicker stimulation.
 Chapter 12 ■ Activation Methods 219

Figure 12.3 EEG changes in response to 5-Hz flickering dot pattern (0.5, 2, and 6 mm in diameter) stimuli to the full-
field (30 30 ) in a 70-year-old woman with benign paroxysmal positional vertigo. Stimulus was provided by use of a
square-type strobe-filter method (92); luminance was maintained by 20 cd/ m 2. Lower tracings show mapping analysis of
photic driving responses (PDRs) elicited by each flickering dot pattern.

occipital PDRs elicited by a 5-Hz flickering dot pattern (6 mm

Figure 12.4 Photic driving response (PDR) elicited by regional visual
stimuli in 54 adult psychiatric outpatients. We used an SLS-5100 visual
stimulator (Nihon Kohden). Power spectral analysis was made on the
record for 5 seconds after each 2-second stimulation. Power spectra are
5-, 10-, 15-, 20-, and 25-Hz PDRs; only that of 5-Hz fundamental PDR,
which was the greatest, is shown in this figure. The root of each power
is expressed for convenience to indicate amplitude. (Modified from
Takahashi T. EEG activation by visual stimuli. In: Sato M, Matsuoka H,
eds. The Newest Clinical Electroencephalography. Tokyo: Asakura
Shoten; 1993:323–344, with permission.)
in diameter: 0.5 c/deg) stimulation to the center (11 11 ),
periphery (11 11 to 57 57 ), full field (57 57 ), and
each hemifield with luminance maintained at 10 cd/m 2. The
average PDR amplitude in response to regional flickering dot-
pattern stimuli in 28 patients aged over 45 years was higher
than the average PDR amplitude in 26 patients aged 20 to 44
years. Amplitudes of PDRs elicited by stimuli to the periphery
and lower hemifields were similar to those elicited by a stimu-
lus to the full field. Regarding PDRs elicited by lateral hemifield
stimuli, contralateral PDRs were slightly higher than ipsilateral
ones, but their difference was not statistically significant.
Particular efficiency in eliciting higher PDR by lower hemifield
stimulation compared with other hemifield stimuli might be a
result of its selective stimulus to the visual cortex, which is
located closely beneath electrodes 01 and 02.
A power spectral analysis of PDRs elicited by the above
visual stimuli may be used as a new tool for evaluation of drugs
influencing the brain (53). Other clinical usefulness of this
EEG examination is reported in the following discussion. In a
patient with amblyopia, a distinct difference of PDRs evoked
by stimulation of 5-Hz flickering dot pattern to each eye was
obtained (56); PDRs evoked by hemifield 5-Hz flickering dot-
pattern stimuli enabled EEG diagnosis of a parietal lobe
metastatic brain tumor in a patient who had suffered from
breast cancer (57).
220 Part II ■ Normal EEG
PHOTOMYOCLONIC RESPONSE
(PHOTOMYOGENIC RESPONSE)
This is a response to IPS characterized by appearance of brief
repetitive muscle spikes over anterior regions of the head. These
often increase gradually in amplitude as stimulation continues
and cease promptly when the stimulus is withdrawn. The
response is associated frequently with eyelid flutter, a vertical
oscillation of the eyelids and eyeballs; sometimes it is associated
with discrete jerking, mostly involving musculature of the face
and head (38).
Principal features of this response were first described by
Gastaut and Rémond (58) and Bickford et al. (59), who intro-
duced the term photomyoclonic response (PMR), which differs
from PPR. Occasionally, a PMR can be seen with a PPR. The
most effective triggering IPS frequency lies between 12 and 18
Hz (60). The PMR tends to appear in conjunction with muscu-
lar tension. It occurs less often in children than in adults.
Gastaut et al. (61) reported that PMR was found in 0.3% of
normal subjects, 3% of patients with epilepsy, 13% of patients
with brainstem lesions, and 17% of patients with psychiatric
disorders. The incidence of PMR in other studies (62–65), how-
ever, was lower than in that of Gastaut et al., ranging from 0.1%
to 0.8%. The PMR is considered to be a nonspecific finding that
is not significant for elimination diagnosis of a seizure disorder
(66). It is enhanced in early stages of alcohol withdrawal in
chronic alcoholics (67) or after sudden withdrawal from barbi-
turates and related sedatives (68). A symptomatic transient
PMR secondary to a metabolic disorder (severe hypocalcemia)
can occur (69).
PHOTOPAROXYSMAL RESPONSE
(PHOTOCONVULSIVE RESPONSE)
This is a response to IPS characterized by spike-and-wave and
multiple spike-and-wave complexes that are bilaterally syn-
chronous, symmetrical, and generalized and that may outlast
the stimulus by a few seconds. There may be associated impair-
ment of consciousness and brisk jerks involving musculature of
the whole body, most predominantly that of the upper extrem-
ities and head (38). Newmark and Penry (66) state that gener-
alized slow activity and posterior spikes are not accepted
universally as a PPR, but the significance of these discharges
increases if they continue after stimulation is discontinued.
Although PPRs in photosensitive epilepsy patients are elicited
by a broad range of IPS frequencies from 1 to 65 Hz (70), the
most epileptogenic frequencies are within the range of 15 to 18
Hz (71). PPR is most frequently induced by 15-Hz IPS with eyes
closed and 20-Hz IPS when the eyes are open (72). When IPS is
given with eyes open, careful attention to eye movement and
position is needed because a directional change from central to
lateral gaze diminishes the effect of IPS in evoking PPR. This
effect is greater than the diminution effect with monocular IPS
stimulation as compared with binocular IPS stimulation in elic-
iting PPR (73). Regarding levels of ambient light, with high-
intensity IPS, the effects of normal ambient lighting may be
negligible (74). During IPS, the PPR may be induced by eye
closure, especially immediately after eye closure (75). Patients
with PPR are less sensitive to IPS when asleep (76,77).
Generalized PPR may be most pronounced in the frontal, cen-
tral, or occipital regions. Hishikawa et al. (76) classified general-
ized PPRs in photosensitive epilepsy patients into PPRs that
appear first in the occipital area and PPRs that occur simultane-
ously over all areas or appear earlier over anterior regions. In the
former group, an augmented visual evoked response could be
obtained. Generalized PPRs in children are usually rhythmic
and higher in amplitude than those seen in adults. Occipital
spikes as a sole response to IPS may not be indicative of epilepsy
(78). Unilateral occipital spikes are rarely induced by IPS. These
patients, as opposed to patients with generalized PPR, often have
a history of a local posterior lesion, mostly traumatic. Very
rarely, frontal spikes can be provoked by IPS (79).
Jeavons (80) found a PPR in 2.8% of the patients referred for
EEG examination, a figure similar to that of Gastaut et al. (61).
The patients, whose seizures are induced by visual stimuli such
as viewing a visual pattern, television (81), or eye closure, are
particularly sensitive to IPS and tend to demonstrate PPR.
Gastaut et al. reported that the PPR is almost entirely confined
to patients with primary subcortical epilepsy and that it occurs
in 40% of cases with absence seizures and in 20% of cases with
GTCSs. According to Stevens (82), PPR occurred in 53% of that
study’s patients with nonfocal seizures, but in only 3% of that
study’s patients with focal seizures. A similar view has also been
expressed by Niedermeyer (11), who states that generalized
PPRs are more highly associated with primary generalized
epilepsy than with incompletely generalized or focal spikes.
However, Guerrini et al. (83) proposed a concept called idio-
pathic photosensitive occipital lobe epilepsy based on detailed
analysis of 10 patients with recurrent episodes of visually
induced occipital seizures; all seizures were stimulus (mostly
television and computer screen) related and began with ele-
mentary visual symptoms, followed in most patients by a slow
clustering of cephalic pain, epigastric discomfort, and vomiting,
with either normal or only mildly impaired responsiveness: all
of them showed PPR of types 1 to 4 (see below). The ictal events
that may accompany PPR are predominantly absence seizures,
GTCSs, and myoclonic jerks, especially of the eyelids or arms
(84). Gambardella et al. (85) reported a 17-year-old girl with
pure photosensitive epilepsy who showed photic-induced
epileptic negative myoclonus (PPR was accompanied by loss of
postural tone in both arms). They claim that negative
myoclonus should be included among the ictal phenomena
accompanying PPR.
Some studies have suggested that presence of PPR may be a
familial trait (73,86,87). Doose et al. (86) reported that PPR
may be regarded as a symptom of susceptibility to convulsions
of the centrencephalic type. In a genetic study of photosensi-
tive epilepsy, Waltz et al. (88) classified PPR into four types:
type 1, spikes with occipital rhythm; type 2, parieto-occipital
spikes with a biphasic slow wave; type 3, parieto-occipital
spikes with a biphasic slow wave and spread to the frontal
region; type 4, generalized spikes and waves or polyspikes and
waves. They found that type 4 occurred more often in both
probands with epilepsy and their siblings than the respective
 Chapter 12 ■ Activation Methods 221

controls, suggesting that coincidence of photosensitivity will
appear as higher if only type 4 is considered to be indicative
of photosensitivity. Photosensitivity is age dependent, being
highest in late childhood and early adolescence; it is more
common in females (86). According to a prospective nation-
wide study made in Great Britain (89), the annual incidence of
cases of epilepsy with type 4 PPR on their first EEG was
approximately 2% of all new cases of epilepsy. When restricted
to the age range of 17 to 19 years, the annual incidence rose to
approximately 10% of all new cases of epilepsy presented in
this age range. Analyzing generalized PPRs found in
128 chronic epilepsy patients, De Graaf et al. (90) reported that
there was a significantly higher occurrence in whites (2.7%, 72
of 2657) as compared with blacks (0.1%, 1 of 848) and subjects
of mixed race (0.9%, 55 of 5958), concluding that genetic
rather than environmental factors influence generalized PPR.
For epileptic seizures triggered by television, photosensitivity
remains the most common single mechanism (47). Since the
first description of space invader epilepsy by Rushton (91),
patients with seizures triggered by electronic screen games
(ESGs) such as video, console, and computer games have been
reported (92,93). A nationwide study made in Great Britain (94)
during 6-month periods identified 118 patients who had a first
seizure while playing ESGs when they were 7 to 19 years old. Of
118 patients, 46 and 25 had definite and probable causal rela-
tionships, respectively; 47 patients had no such relationship.
Within that age group, the annual incidence of first seizure trig-
gered by playing ESGs (71 patients altogether) was estimated to
be 1.5/100,000, representing approximately 3% of all new
patients with epilepsy in this age range. Of 71 patients, 46
showed type 4 PPR, whereas 25 showed types 1 to 3 PPR.
Although photosensitivity is thought to play the most important
role in engendering ESG-induced seizures, other circumstances,
acting either singly or in combination, should be taken into con-
sideration (95). Examples include (i) seizure precipitation by
specific cognitive activities, decision making, hand movements,
etc.; (ii) seizure precipitation by nonspecific emotional factors
relating to the subjects’ engagement in games, such as anxiety or
excitement; (iii) lowering of the seizure threshold by fatigue or
sleep deprivation; and (iv) chance occurrence of a spontaneous
seizure in a person with epilepsy while playing ESGs. Figure 12.5
illustrates generalized paroxysmal discharges elicited by a video
game and generalized PPRs elicited by flickering visual stimuli
in a 13-year-old boy who had an ESG-induced GTCSs.

Figure 12.5 Generalized paroxysmal discharges elicited by a video game (Super Tetris 2 Bombliss) and generalized pho-
toparoxysmal response (PPRs) elicited by flickering stimuli in a 13-year-old photosensitive epilepsy patient. We used a
12-in. television placed 1 m in front of the patient. Full-field (30 30 ) stimuli of a 15-Hz red flicker and a 25-Hz flicker-
ing 1.5 c/ deg dot pattern provided by use of square-type strobe-filter method elicited type 4 PPR. Luminance maintained
at 20 cd/ m 2. This EEG was recorded under drug-free conditions.
222 Part II ■ Normal EEG
Occurrence of a PPR in an otherwise seizure-free individual
is uncommon; these patients tend to have attacks of headache
or psychological problems, such as personality disorder with
hysterical reactions or an anxiety state (60). Incidence of PPR in
nonepileptic, apparently normal subjects is estimated at proba-
bly less than 2% (66).
Reilly and Peters (64) claim that the following criteria are
useful for determining whether or not the patients are epileptic:
prolonged PPR (outlasting the stimulus) suggests probable
epilepsy (90%), whereas self-limited discharge is not diagnostic.
Jayakar and Chiappa (96) studied 35 patients with PPR (1% of
all patients seen in an EEG laboratory in a general hospital) and
found that 77% had a definite history of epilepsy, 9% a ques-
tionable history, and 14% had had no seizures. PPRs were pro-
longed more than 100 msec beyond the stimulus in 11 patients
and self-limited in 24, and the incidence of seizures was not dif-
ferent in the two groups. Furthermore, the PPR prolongation
was related to the time at which the stimulus was stopped after
the appearance of the PPR—shorter durations of stimulation
tended to have longer prolongation.
With respect to the duration of photic stimulation in the clin-
ical EEG laboratory, it should be noted that the technologist must
be alert to stop stimulation immediately on identifying a PPR
since prolonged stimulation can trigger a GTCS in a sensitive
patient. Also, in these patients, once a PPR has been definitely
identified, further stimulation should not be attempted as these
patients may become more sensitive as stimulation continues or
is repeated, and may launch into a seizure with little warning.
Especially in adults without previous seizures and without a
family history of seizures, PPR usually suggests a toxic, meta-
bolic, or drug withdrawal state (41,68,97,98). Symmetrical pos-
terior high-amplitude spikes can be evoked at slow IPS rates in
patients with diffuse encephalopathies, such as progressive
myoclonic epilepsy or Creutzfeldt–Jakob disease (44,99,100).
Spontaneous epileptiform abnormalities (SEAs) are
reported to occur in up to 65% of patients with a PPR. Gilliam
and Chiappa (101) examined seizure classifications and SEAs in
115 consecutive patients who had a PPR. A PPR was the only
epileptiform abnormality in 47 patients (41%), 27 (24%) had
focal SEAs, and 41 (36%) had only generalized SEAs. Seventeen
patients (15%) had partial seizures and 40 (35%) had only gen-
eralized seizures, and seizure classification was strongly associ-
ated with the type of SEA (P 0.0001). Although the PPR is
often presumed to signify primary generalized epilepsy, most
patients with a PPR and focal SEAs have partial seizures. This is
perhaps an indication that the presence of an underlying gener-
alized hyperexcitability, possibly related to a channelopathy (for
example), enhances the likelihood that a focal lesion will gener-
ate an electrophysiologic abnormality.
Several conditions are known to alter PPR. As described
above, PPR is often associated with eye closure and is most likely
to occur immediately after eye closure. This effect may also
result from movement of the eyelids or of the eyes themselves. In
addition, eye closure may produce activation by suddenly elim-
inating the visual pattern and by causing the field of vision to
become red-hued (102,103). When stimuli are presented with
the eyes open, a patterned field of vision (or flickering geomet-
Figure 12.6 A flash lamp and circular-type strobe filter. Luminance
of direct stroboscopic intermittent photic stimulation (IPS) by use
of LS-706A (Nihon Kohden), as shown in the left tra cing, is
3,939 20 cd/ m 2. Inserting a vertical grating pattern (middle tracing)
along the filter holder on the front of the flash lamp as shown in the left
tracing greatly reduces luminosity to 30 cd/ m 2 and spatial frequency of
each pattern filter is 2 c/ deg. Using this filter, the weakest possible stim-
uli of red flicker and pattern flicker ought to be used, taking great care
not to induce clinical seizures in photosensitive individuals; one might,
for example, start stimuli from lower flicker frequencies and terminate
examinations when PPR appear (92).
ric pattern) has been found to be more effective than a homoge-
neous field (or white light flicker) in eliciting PPR (75,104,105).
In addition, saturated long-wavelength red (red) facilitates
induction of PPR (106,107). An EEG activation by use of red
flicker and flickering geometric pattern stimuli, which were
given only in “eyes-open” conditions, revealed a higher rate of
PPR provocation than by ordinary stroboscopic IPS (51). (In the
former method, an SLS-5100 visual stimulator (50) was
employed and luminance maintained at 20 cd/m 2.) Additional
studies (108,109) have disclosed that elicited PPRs were found in
18% of epilepsy patients, suggesting that incidence of photosen-
sitivity in epilepsy patients is 3.6 times higher than that of the
commonly accepted 5% reported by Binnie and Jeavons (110).
Figure 12.6 shows a circular-type strobe filter (93,111), which is
a simple device that is capable of inducing similar activation
effects to the SLS-5100 visual stimulator. Using strobe filters,
visual stimuli (red flicker and flickering geometric pattern, lumi-
nance kept at 30 cd/m 2) are easily presented to patients. Their
stimulus luminance is far less intense than that of a stroboscopic
IPS. In addition to examining photosensitivity with an ordinary
strobe, such low-luminance visual stimuli may be required for
precise examination of photosensitive epilepsy patients.
Chiappa et al. (112) and Hill et al. (113) used functional
magnetic resonance imaging (fMRI) and magnetic resonance
spectroscopy (MRS) to study metabolic and hemodynamic
consequences of photic stimulation–triggered and spontaneous
generalized epileptiform discharges in humans. EEG recordings
were performed in the MR scanner during photic stimulation
in normal subjects and patients with PPR. There were no fMRI-
registered hemodynamic abnormalities found in relation to the
brief PPR seen in the patients. However, irrespective of the
presence of a PPR, these patients showed four unique findings
compared with normal subjects: (i) slightly but significantly
increased lactate levels in the occipital cortex in the resting
state, (ii) an increased area of visual cortex activation with
photic stimulation, (iii) noncontiguous areas of signal attenua-

tion, simultaneous with the occipital cortex stimulus-induced
increased fMRI signal, most prominent in perirolandic regions,
and (iv) a marked decrement (undershoot) of fMRI signal
intensity immediately after the photic stimulation in the occip-
ital cortex and in the region of the posterior cingulate gyrus.
These findings suggest abnormal interictal metabolism and
increased vascular reactivity in the photosensitive patients.
Along with PPR, another term in use is photoepileptiform
response (PER). This response is elicited by photic stimulation.
According to their topography, they are divided into the follow-
ing three categories (8): (i) anterior dominant or generalized
(bisynchronous and approximately symmetrical); (ii) occipital
dominant (bisynchronous and approximately symmetrical);
and (iii) occipital dominant and localized (unilateral or
strongly lateralized). The category 1 PPR is one form of PER
and is most closely associated with epilepsy.
OTHER FORMS OF SENSORY STIMULATION
Visual Stimulation
Various visual stimuli other than IPS may provoke PERs. The sim-
ple maneuver of eye closure induces posterior slow-wave tran-
sients, especially in children (114,115), and may also provoke
paroxysmal discharges (116,117). Paroxysmal discharges similar
to PPR seen in photosensitive epilepsy patients may even be
induced by slow closure of the eyelids (118). In the act of eye clo-
sure carried out in a lighted room, lights off and homogeneous
visual field (ganzfeld) (119–121) are considered to play significant
roles as visual stimuli in producing paroxysmal discharges. In fact,
posterior sharp waves or spikes can be seen occasionally after or at
the onset of IPS; these are called off-response and on-response,
respectively (73,122). Spike–wave discharges associated with
seizures can rarely be induced by a single-flash light stimulus
(102). Paroxysmal discharges may be evoked by red light alone
(93,119,123). Darkness may also induce paroxysmal discharges
(124,125). Similar paroxysmal discharges produced by ganzfeld
and total darkness have been reported by Panayiotopoulos
(126,127), who called it fixation-off sensitivity (FOS); this unpat-
terned vision can be obtained either through 10 spherical lenses
or through underwater goggles covered with semitransparent
paper. He claims that FOS is suspected if EEG abnormalities
appear and persist as long as the eyes remain closed and disappear
when the eyes are opened. In such cases, EEG abnormalities of
FOS can be elicited by elimination of central vision and fixation;
FOS is related to eyes-closed, not eye-closure EEG abnormalities.
According to Panayiotopoulos (126,127), the incidence of FOS-
related EEG abnormalities appears to be as frequent as that of PPR
in children younger than 12 years of age; FOS is probably more
frequently associated with benign childhood epilepsy with occip-
ital paroxysms than with any other epileptic condition.
Pattern, Pattern Flicker, and Red Flicker Stimulation
Geometric patterns often provoke paroxysmal discharges in
epilepsy patients, particularly in photosensitive ones (128–132).
The patterns that are most effective for activation consist of
closely spaced lines or dots with sharply contrasting interfaces
Chapter 12 ■ Activation Methods 223
Figure 12.7 Geometric patterns used for square-type strobe-filter
method. Each pattern was presented 30 cm before the eyes, resulting in
its spatial frequency at 2 c/ deg; pattern contrast is 0.98.
that are arranged geometrically (72). According to Wilkins et al.
(133), to induce paroxysmal discharges, patterns (black and
white stripes of equal width and spacing) must have a spatial
frequency between 1 and 4 c/deg. Line patterns are more epilep-
togenic than dot patterns (Fig. 12.7). Oscillation of a grating
pattern greatly enhanced provocation of paroxysmal dis-
charges; gratings oscillating in a direction orthogonal to the
lines were the most effective, and the optimum oscillation fre-
quency was in the range of 15 to 20 Hz (134).
Figure 12.8 is an example of 3-Hz spike-and-wave complexes
associated with an absence seizure induced by viewing a grating
pattern in a 3.4-year-old girl with severe myoclonic epilepsy in
infancy (135). This patient was very sensitive to vertical, hori-
zontal, and oblique stripes with a wide range of spatial frequen-
cies. She was also sensitive to dot patterns, but slightly less so
than to stripes. Interestingly, this girl repeatedly stared at those
patterns as if she were enjoying the absence seizures. A dot pat-
tern, however, more frequently evokes lambda waves and rhyth-
mic posterior slow activity than a grate pattern (136). Lambda
waves are more frequently evoked in subjects who show good
PDRs, especially to low-frequency IPS. Radhakrishnan et al.
(137) reported pattern-induced negative occipital potentials
(PINOP), which are similar to lambda waves but differ in
response characteristics with surface negativity. Photosensitive
epilepsy patients are also often sensitive to pattern viewing,
with an incidence of 5% (102), 22% (105), or 72% (138).
Paroxysmal discharges induced by viewing usually arise from
the occipital area, suggesting that they may arise primarily in an
epileptogenic visual cortex (103,130,133). Geometric pattern
stimulus occasionally induces fast activity (beta rhythm) (53).
Pattern viewing may also enhance mu rhythm (139).
It has been reported that photosensitive epilepsy patients are
particularly sensitive to full-field visual stimuli of a red flicker
and flickering geometric pattern, when a decreased luminance,
such as 10 to 20 cd/m 2, is used as a stimulating light source (108).
Figure 12.9 illustrates EEG changes in response to stroboscopic
IPS, red flicker, and flickering dot pattern stimuli in a photosen-
sitive epilepsy patient. Particular emphasis ought to be given to
the finding that IPS to eyes open was ineffective, whereas red
224 Part II ■ Normal EEG

Figure 12.8 Generalized 3-Hz spike-and-

wave complex elicited by a horizontal
2.1 c/ deg grating pattern stimulation in a
3.4-year-old girl with severe myoclonic
epilepsy in infancy. Vertical and horizontal
eye movements, electromyogram (EMG) of
the left hand, and electrocardiogram (ECG)
were recorded simultaneously. Downward
and upward arrows indicate the beginning
and the end of a grating pattern stimula-
tion, respectively. In accordance with
occurrence of a generalized paroxysmal
discharge, regular button pressing shown
at the top became absent, suggesting that
she has an absence seizure. (Modified from
Takahashi T, Sasaki M. Pattern sensitivity:
its activation method and clinical EEG find-
ings. Rinsho Nohha (Osaka). 1971;13:
727–732, with permission.)

Figure 12.9 EEG changes in response to stroboscopic IPS, red flicker, and flickering dot pattern stimuli in a 14-year-old
girl with photosensitive epilepsy. By use of Grass PS33-plus photic stimulator (intensity at 8), 15-Hz IPS was given 30 cm
above the eye. Using the strobe-filter method, 15-Hz red flicker and 15-Hz flickering 2 c/ deg dot pattern stimuli to the full
field (25 25 ) were given from 30 cm before the eye; luminance maintained at 20 cd/ m 2.

Figure 12.10 Dot pattern filters used for regional flickering dot pattern
stimuli by use of square-type strobe-filter method. Diameter of a single
dot is 4 mm (0.8 c/ deg), and luminance maintained at 20 cd/ m 2. (See
color insert.)
flicker and flickering dot pattern stimuli elicited generalized
PPRs, suggesting that stroboscopic IPS to eyes open, with lumi-
nance of 3939 cd/m 2, is too strong to elicit PPR in this patient
(140). In addition to hemifield (109,141–144) and quadrant-field
(145) stimuli, visual stimulation to the center and periphery by
use of the above stimuli can also yield useful information to fur-
ther elucidate PPR (109). As an example, Figure 12.10 shows fil-
ters used to provide a flickering dot pattern stimulation to the
center, periphery, and hemifield by the square-type strobe-filter
method (53,93); similar field-size filters are used in red flicker
stimuli. Figure 12.11 demonstrates EEG changes in response to
such regional visual stimuli in the same patient shown in Figure
12.9. A red flicker stimulation to the center promptly elicited gen-
eralized PPR as shown in the left tracing. That to the periphery,
Chapter 12 ■ Activation Methods 225
however, was less effective. Lateral hemifield stimuli elicited gen-
eralized PPR (equal PPR), but upper and lower hemifield ones
were ineffective in this patient. In terms of PPR etiology, equal
PPRs as shown in Figure 12.11 are thought to be related to
genetic factors, whereas an unequal PPR, in which either left or
right hemifield stimulation elicits a PPR, is indicative of acquired
cerebral disturbance, such as head trauma (109).
Regarding provocation of generalized PPR by visual stimuli
to the center and periphery, a red flicker stimulation to the cen-
ter is more effective than that to the periphery, indicating that
macular stimulation plays a central role in it. In eliciting PPR by
flickering dot pattern stimulation to the full field, a dot pattern
with higher spatial frequency, such as 2.1 to 1.5 c/deg, shows the
most significant efficacy; flicker frequencies must be between
20 and 15 Hz in that condition (146). Flickering dot pattern
stimulation to the center too, when a higher frequency dot pat-
tern (2.1 to 4.9 c/deg) is used, elicits generalized PPR similar to
that induced by red flicker stimulation to the center (147),
whereas that to the periphery by use of 0.5 c/deg dot pattern
elicits generalized PPR with apparent local onset from posterior
regions (109). Therefore, photosensitive epilepsy patients who
are sensitive to red flicker and flickering dot pattern stimuli are
thought to have hyperexcitable areas not only in the striate
cortex but also in the parastriate cortices. Local PPR in the
striate cortex elicited by red flicker stimulation would immedi-
ately transmit to the nonspecific diffuse projection system.
This could engender generalized PPR with their simultaneous
occurrence over all areas, as distinguished from those elicited by
flickering dot pattern stimulation. In the latter, initial local PPR
would seem to originate in the striate cortex as well as in the
Figure 12.11 Generalized PPRs elicited
by regional red flicker stimuli in a 14-year-
old girl with photosensitive epilepsy (the
same patient shown in Fig. 12.9).
226 Part II ■ Normal EEG
parastriate cortices. These might produce different generalized
PPR, preceded by spikes over the posterior regions.
Of 18 photosensitive patients sensitive to red flicker stimula-
tion (140), all cases were sensitive to at least one of the flicker-
ing geometric pattern stimuli. Although this finding suggests
that common neuronal pathways are involved in PPR induction
by red flicker and flickering geometric pattern stimuli, two dif-
ferent visual cortical pathways, namely the parvocellular (P)
and the magnocellular (M) systems, are presumed to severally
play a part in the following (93,148): the P system deals with
color, and V4 plays a central role for perception of both shape
and color (149). In human subjects, the region of the lingual
and fusiform gyri has been identified as the color center
(150,151). Regarding reversing patterns, Harding and Jeavons
(47) maintain that they are dealing with transient phenomena
and that perhaps the M system provides critical transmission
that is similar to the mechanism of pattern sensitivity. This
understanding may also hold true for flickering geometric pat-
tern stimulation (53). From findings obtained by regional
visual stimulation technique, however, the following view
appears to be most likely: generalized PPRs elicited by red
flicker and high spatial frequency pattern flicker stimuli to the
center are mediated by the P system, whereas those elicited by
low spatial frequency pattern flicker stimulation to the periph-
ery are mediated by the M system (93).
The above knowledge obtained from EEG diagnosis by low-
luminance visual stimuli provides us important clues for
detailed understanding particularly of seizures induced by
stimulative screen images observed through use of TV, video
systems, computer, and so on (93). In a study that dealt with (i)
30 photosensitive epilepsy patients and (ii) 78 epilepsy patients
with PPR (152), red flicker and pattern flicker stimuli elicited
type 4 PPR more frequently in (i) than in (ii) subjects, suggest-
ing that both the P-ventral and M-dorsal visual pathways are in
a sensitive state in photosensitive epilepsy patients. In contrast,
pattern flicker stimulation alone tended to induce type 4 PPR in
the (ii) subjects, suggesting that only the M-dorsal visual path-
way is in a sensitive state in epilepsy patients with a PPR.
Auditory Stimulation
When compared with visual stimuli, activation of paroxysmal
discharges by auditory stimuli is extremely rare. Being either
transient or continuous, paroxysmal discharges provoked by
auditory stimuli have been found mainly over temporal areas
(153–157). Pure tones have been shown by Arellano et al. (153)
to activate paroxysmal discharges in a psychomotor epilepsy
patient. Prechtl (158) reported that repetitive clicks caused
aggravation of preexisting temporal lobe abnormalities in 37%
of a large group of mixed cases. Stevens (82) found the activat-
ing mechanism in 4.5% of temporal lobe epilepsy patients.
These findings suggest that an epileptic disturbance in the pri-
mary auditory area of superior temporal gyrus (Heschl’s gyrus)
may chiefly be responsible for auditory stimuli that give rise to
such paroxysmal discharges. However, Kiloh et al. (23) showed
that unilateral central sharp waves could be evoked by random
taps. When generalized paroxysmal discharges appear to be
induced by auditory stimulation in drowsiness and sleep, spe-
cial care in interpretation is needed because arousal secondary
to auditory stimuli, not auditory stimulus per se, may play a
role in triggering paroxysmal discharges.
Hearing-induced seizures are precipitated by hearing more
or less specific music (159), voices, and other sounds. According
to Rosenow and Lüders (160), those patients usually have auto-
motor seizures or GTCSs and temporal lobe epilepsy; the
pathogenesis of auditory-induced seizures most probably
involves activation of temporal auditory areas.
Somatosensory Stimulation
When areas of the primary sensorimotor cortex become
hyperexcitable as a result of disease, contralateral afferent stim-
uli may trigger focal spikes and partial seizures. Along with tac-
tile stimulation (161–164), local stimuli by cold wind, rubbing,
vibration, and stretching a muscle appear to rarely provoke
spikes and seizures as reviewed by Takahashi (53). Tapping
anywhere about the upper torso or head produces myoclonic
movements associated with diffuse spike–wave discharges in
some young idiopathic epilepsy patients who are sensitive to
afferent stimuli (41). Focal areas that are sensitive to these
stimuli are the finger, arm, shoulder, chest, face, head, and feet.
In one patient, partial elementary seizures were evoked by
tooth brushing (165). O’Brien et al. (166) reported a 23-year-
old female complex epilepsy patient with seizures precipitated
exclusively by tooth brushing and in whom structural and
functional imaging demonstrated a right posterior frontal
focus. Using tactile stimulation comprising tapping of the pal-
mar tips of the fingers or toes, Langill and Wong (167) investi-
gated 304 patients with rolandic discharges. In this stimulus
condition, right-hand tapping, for instance, would elicit dis-
charges over the left central region, whereas tapping of either
foot would elicit discharges at the vertex. Such tactile-
enhanced discharges constituted 38.2%; they were regarded as
benign, age-related phenomena.
Reviewing published cases of somatosensory reflex epilepsy,
Servit (168) stated that seizures are provoked by stimulation of
the trigeminal somatosensory region in the majority of patients
(over 80%). Exaggerated responses in such patients can be
obtained by electrical stimulation of peripheral nerves. The
amplitude of somatosensory evoked potentials (SEPs) to
median nerve electrical stimulation is markedly enlarged (giant
SEPs) in most patients with cortical myoclonus of various eti-
ology (169–173).
Mima et al. (174) studied proprioception-related SEPs with
passive flexion movement of the middle finger at the proximal
interphalangeal joint in seven cortical myoclonus patients who
had associated giant SEPs. In three of the seven patients, the
proprioception-related SEPs were also enlarged. Therefore, the
authors concluded that hyperexcitability of the sensorimotor
cortex in cortical myoclonus is modality-specific; cortical
excitability is exaggerated to both cutaneous and deep receptor
inputs in some patients, but only to cutaneous input in others.
Figure 12.12 shows an example of unilateral central–parietal
spikes evoked by electrical stimulation of the contralateral
median nerve in a patient with benign childhood epilepsy with
centrotemporal spike.

Figure 12.12 Electrical stimulation of right median nerve evoked
spikes, maximal over left central–parietal areas, in a 6-year-old boy with
benign childhood epilepsy with centrotemporal spike.
Plasmati et al. (175) found increased incidence of giant SEPs
(38%) in 44 benign epilepsy with rolandic spikes (BERS) patients.
According to their further study (176), which investigated 54
BERS patients and 61 age-matched control subjects, short laten-
cies were normal in all BERS patients, whereas the amplitude of
middle latency was increased in a significant percentage of chil-
dren with BERS. Giant responses occurred in 25.9% of BERS
patients and in 4.9% of normal subjects. This phenomenon was
presumed to depend on physiologic, functional cortical hyperex-
citability related to maturation processes. They state that the giant
responses clearly differ from those observed in myoclonic epilep-
sies and are likely to be generated in different neuronal pools in
the sensorimotor cortex. The similar distribution of these giant
middle-latency SEPs with tactile evoked EEG spikes suggests a
common generator for these responses. Electrical stimulation of
the median nerve on routine EEG examination evoked
central–parietal spikes in 3% of patients; all of them were epilepsy
patients, the majority of whom showed PPR as well (156). In a
case reported by Goldie and Green (177), seizures elicited by rub-
bing the face could also be precipitated just by thinking about
rubbing the face or by hypnotic suggestion of cutaneous stroking.
Van Cott et al. (178) reported four adult patients with stimulus-
sensitive seizures and myoclonus following severe hypoxic-
ischemic injury (postanoxic intention myoclonus, also termed
Lance-Adams Syndrome); their EEGs showed bursts of general-
ized spike and polyspike activity following tactile stimulation
associated with clinical seizures. In a comatose patient with
reversible viral encephalitis, the association between stimulus-
induced periodic epileptiform discharges, arousal EEG responses,
and limb jerking was observed (179).
OTHER SENSORY STIMULI
Regarding EEG activation by olfactory stimulation, very few
data exist. Stevens (82) tested 61 epileptics with gross olfactory
stimulation. The stimuli consisted of a strong perfume (Tabu)
and a malodorous H 2S compound. Sixteen out of 61 temporal
lobe epilepsy patients demonstrated exaggerated spiking during
Chapter 12 ■ Activation Methods 227
Figure 12.13 EEG responses to odorous stimuli in a 29-year-old
woman with epilepsy with generalized tonic–clonic seizures (GTCSs)
on awakening.
or immediately after exposure to perfumed air. We also demon-
strated EEG changes in response to olfactory stimuli in an adult
epilepsy patient (79); generalized paroxysmal discharges were
induced by ether and acetic acid. Figure 12.13 shows another
example of odor sensitivity in a patient with GTCSs on awaken-
ing; olfactory stimuli with acetic acid and coffee elicited gener-
alized paroxysmal discharges. In distinction to other modalities
of sensory stimuli, repeated olfactory stimuli tend to quickly
become ineffective. On EEG activation, olfactory stimuli using
vanilla, coffee, and so on, may be desirable because ammonia
and similar pungent substances stimulate the trigeminal nerve.
Very little is known about the influence of vestibular stimu-
lation on the EEG. Behrman and Wyke (180) reported that
bilateral 6-Hz activity arising from the temporal area could be
evoked by caloric stimulation. Focal EEG changes were pro-
duced by rotational stimulation in 9 of 15 epileptics tested
(181); in 4 patients, bilateral 2- to 7-Hz activity arose from the
temporoparietal regions. Karbowski (181) performed caloric
stimulations in 62 epilepsy patients (irrigation for 15 seconds
with 100 mL water at 27 C and at 44 C). In 22 cases, an increase
in, or the appearance of, EEG abnormalities was detected. In
two further cases epileptic seizures were triggered.
Multisensorial and Vegetative Sensory Stimuli
More than 600 patients with hot water epilepsy have been
reported worldwide; more than 90% of them are from southern
India (183). The following factors are presumed to be involved
in its pathophysiology: (i) high temperature, (ii) direct contact
of skin with water, (iii) body part in contact with water, and (iv)
mode of bathing. Generalized epileptiform abnormalities were
found in 15% to 22% of patients (184). Other reports have
described both diffuse and focal abnormalities, predominantly
in the temporal lobes. Ictal EEG abnormalities in 10 patients
include rhythmic left temporo-occipital spikes (183,185).
Seizures induced by eating (eating seizures) are rare.
Rémillard et al. (186,187) proposed two main clinical syn-
228 Part II ■ Normal EEG
dromes of eating seizures, differentiating temporolimbic from
extralimbic, usually suprasylvian, seizure onset on clinical, radi-
ologic, and EEG evidence. In addition, there was a subgroup of
patients activated by certain tastes, such as sweets (186). They
argued that the occurrence of eating seizures may be related to
the excitation of a critical mass of epileptogenic cortex by vari-
ous afferent stimuli, citing roles of the following: taste, mastica-
tion, gastric distension, somatosensory or proprioceptive input,
and subcortical structures (187).
Seizures of “abdominal epilepsy” have reportedly been
induced by vegetative sensory stimuli, such as enemas and
injection of Prostigmin (116); such EEGs are characterized by
appearance of generalized spike–wave complexes, occasionally
associated with temporal and hemispheric foci.
Unexpected Startling Stimuli
One may rarely encounter patients with startle epilepsy
(163,164,188). Startle-induced seizures are focal or generalized
tonic or atonic seizures, not always associated with loss of con-
sciousness (189). Unexpected startling, mostly loud noises, and
other unexpected somatosensory stimuli, such as sudden move-
ments, are common precipitating factors; very rarely, visual
stimuli are capable of startling. Those EEGs are usually accom-
panied by a temporary desynchronization followed by a gener-
alized or focal seizure pattern (116). The interictal awake EEG
shows frontal, central, or multifocal epileptiform discharges in
the majority of patients (190–192). Based on studies by animal
models and clinical observations (189,193), startle epilepsy has
been classified as “proprioceptive epilepsy” (194). From a clini-
cal-EEG study of 22 children with startle-provoked epilepti-
form seizures (SPESs), considering the high diversity of SPES
patients, Tibussek et al. (195) infer that a common underlying
pathophysiologic mechanism is unlikely.
Transcranial Magnetic Stimulation
Hufnagel et al. (196) used the method of transcranial magnetic
stimulation (TMS) to induce paroxysmal discharge; positive
findings were obtained in 12 of 13 patients. Schuler et al. (197)
statistically compared activation effects of paroxysmal discharge
with those by TMS and HV in 10 patients with drug-resistant
partial epilepsy. TMS caused an activation of the focal paroxys-
mal discharges only in three patients, whereas HV produced it in
six cases; TMS even caused significant reduction of paroxysmal
discharge in two patients. These results led them to conclude that
TMS was not better than HV. To contrast to low-frequency ( 1
Hz) TMS studies, Tassinari et al. (198) reported results obtained
by high-frequency ( 1 Hz) TMS (repetitive TMS [rTMS]) in
more than 60 epilepsy patients; rTMS-induced seizures were
found in 2 of 10 patients with progressive myoclonic epilepsy and
in 1 of 4 patients with epilepsia partialis continua.
OTHER FORMS OF ACTIVATION
Eye Movements
Blinking is known to evoke spike–wave discharges (199,200).
Nadkarni et al. (201) reported that a 6-year-old boy’s blinking
triggered central midtemporal spikes (CMTSs) occurring reflex-
ively, following the blinks by 100 to 200 msec. These blink-
triggered CMTSs appeared almost exclusively while the patient
was awake; it was inferred that the blinking while awake was
associated with spikes resulting from abnormal and excessive
excitatory interconnections between the precentral frontal areas
and the epileptogenic zone of the rolandic cortex. In one patient,
seizures were triggered by blinking when beginning to speak
(202). The phi rhythm (referring to posterior rhythmic slow
waves occurring after eye closure) is an unusual age-related phe-
nomenon seen mostly in young patients. According to Silbert
et al. (203), who defined it as a minimum of three consecutive
monomorphic posterior delta waves occurring within 2 seconds
of eye closure on at least two occasions, the phi rhythm has no
significance for diagnosis of structural cerebral lesions or
epilepsy. In epilepsy patients with this rhythm, however, the
seizure disorder is more likely to be generalized than partial. In
epilepsy patients, spike–wave discharges can also be induced not
only by closing eyes (116,117,199,204–206), but also by opening
of the eyes (114) and performing conjugate eye movements
(103,114,207). However, closing eyes is the most common effec-
tive technique (114).
Paroxysmal discharges can also be provoked by voluntary
eye closure and even by opening the eyes in darkness (103,119,
199,200,204), indicating that spike–wave discharges may be
evoked without the influence of visual stimuli. Generalized and
focal paroxysmal discharges induced by such eye movements
are usually maximal over the frontal areas (103). Sensory affer-
ents arising from contraction of the orbicularis oculi muscle
(200) and proprioceptive impulses from the ocular muscles
(207) are thought to play a role in the mechanism of provoca-
tion of seizure discharges by closing eyes and by conjugate eye
movements. In such epilepsy patients, similar paroxysmal dis-
charges to those described above may also be induced by ocular
pressure (205) or passive lid closure in darkness (208). In an
epilepsy patient, right occipital spikes and complex partial
seizures could be induced by convergence (209).
The above findings are most often seen in photosensitive
epilepsy patients (66). Figure 12.14 is an example of generalized
spike–wave discharges with an anterior maximum induced by
eye closure in darkness in comparison with that induced by a
red flicker stimulation in a photosensitive epilepsy patient.
Figure 12.15 shows another example of generalized spike–wave
discharges induced by eye closure in darkness in a photosensi-
tive epilepsy patient; similar spike–wave discharges were also
induced by passive lid closure and ocular pressure in darkness.
Limb and Trunk Movements
Just as the rolandic mu rhythm is blocked by contralateral move-
ments, suppression of central spikes is also commonly observed
with contralateral motor activities, such as clenching of the fist
(13). Alternatively, active or passive movements of the limbs may
provoke focal motor seizures and focal paroxysmal discharges in
a rare reflex epilepsy patient (210). Forster (119) reported a
patient who had clonic seizures of the right hand; seizures
invariably occurred with the use of the right hand, such as with
writing or finger tapping, even occurring when the patient
imagined movement of the hand. Rarely, seizure occurred with

use of the left hand. These seizures were associated with focal
spikes over the left motor areas.
Movement after various periods of rest may also induce
seizures, including tonic seizure and paroxysmal choreoatheto-
sis (119,211,212). In such patients, in addition to startle or sud-
denness, the motor movement itself and the proprioceptive
feedback secondary to the movements are thought to be
involved in precipitating seizures (119,194).
Neuropsychological EEG Activation
Reflex epilepsies are divided into a simple and a complex group
according to the triggering mechanism. The former comprises
seizure precipitation by simple sensory stimuli and by move-
ments, whereas the latter involves complex mental processes,
such as reading (213–215), and a variety of other cognitive
tasks. Regarding seizures elicited by cognitive tasks, the term
praxis-induced seizures (216–218) has recently become popular.
Chapter 12 ■ Activation Methods 229
Figure 12.14 Generalized paroxysmal discharges induced
by a 15-Hz red flicker stimulation and eye closure in a dark
room in a 24-year-old woman with photosensitive epilepsy.
A red flicker stimulus was provided by a visual stimulator
(48); luminance maintained at 20 cd/ m 2.
All descriptions cited below have only recently been grouped
together as praxis-induced seizures (214): epilepsia arith-
metices, drawing-induced seizures, chess and card epilepsy,
seizures during card games and draughts, reflex epilepsy evoked
by decision making or by specific psychic activity, seizures
evoked by playing with Rubik’s cube, seizures induced by think-
ing, and reflex epilepsy with response to games of chance, cal-
culations, and spatial decisions.
By use of a method of neuropsychological EEG activation
(NPA), Matsuoka et al. (219) reported results obtained from
480 Japanese patients with different types of epilepsy. NPA tasks
consist of reading, speaking, writing, written arithmetic calcu-
lation, mental arithmetic calculation, and spatial construction.
They provoked epileptic discharges in 38 patients (7.9%) and
were accompanied by myoclonic seizures in 15 patients,
absence seizures in 8, and simple partial seizures in 1. Among
the cognitive tasks, mental activities mainly associated with the
Figure 12.15 Generalized paroxysmal discharges
induced by eye closure, passive lid closure, and ocular
pressure in an 18-year-old man with photosensitive
epilepsy. Simultaneous recordings of horizontal and
vertical eye movements are shown (bottom).
230 Part II ■ Normal EEG
use of the hands, that is, writing (68.4%), written calculation
(55.3%), and spatial construction (63.2%), provoked the most
discharges, followed by mental calculation (7.9%) and reading
(5.3%). As for precipitating events, action programming–type
activities were thought to be the most crucial in 32 out of the
38 patients (84.2%), followed by thinking-type activities in
4 patients (10.5%). Regarding classification of epilepsies,
seizure-precipitating mental activities were almost exclusively
(in 36 out of the 38 patients) related to idiopathic generalized
epilepsies (IGEs), and were rarely (in only 2 out of the 38
patients) related to temporal lobe epilepsy. In IGE patients,
provocative effects of NPA were related to myoclonic seizures
rather than absence or GTCSs. These results suggest that NPA is
a useful tool for examining the relationship between cognitive
function and epileptic seizures.
Shown in Figure 12.16 are paroxysmal discharges induced by
written arithmetic in a juvenile myoclonic epilepsy (JME)
patient; the following new question (right tracing) provoked
more intensified paroxysmal discharges than those by the first
one (left tracing). The right tracing of Figure 12.17 illustrates
bilateral paroxysmal discharges elicited by the Uchida–Kraepelin
test (continuous one-digit addition) in a JME patient who had
recurrent myoclonic seizures during video games; he was non-
photosensitive (220). Those shown on the right appear to be
similar to generalized paroxysmal discharges evoked by a video
game (the left tracing), suggesting that seizures induced by video
games are not caused by visual stimuli, but are instead produced
by combined factors of hand movements and calculation, which
would be an example of praxis.
During mental work requiring concentration of attention,
some subjects show theta-activity bursts in the frontal midline
(221). However, magnetoencephalography analysis (222) dis-
closed that frontal midline theta-burst activities elicited by sim-
ilar tasks in fact distribute widely in the frontal cortices of both
cerebral hemispheres. Therefore, the authors claim that an
expression of frontal mental theta waves is more appropriate
Figure 12.16 Generalized paroxysmal
discharges induced by written arithmetic in
a 25-year-old man with juvenile myoclonic
epilepsy. The left paroxysmal discharge in
the left tracing appeared 10 seconds after
calculation was started. As shown in the
right tracing, another question elicited
intensified paroxysmal discharges that
appeared 2 seconds after a calculation was
started; simultaneous video-EEG monitor-
ing revealed that myoclonic seizures of
both arms were associated with it.
than that of “frontal midline theta” being recorded maximally at
the frontal midline (223).
Emotional Changes
It has long been thought that certain emotional states may be
influential either directly or indirectly in precipitating seizures
(224,225). Stevens (226) used emotional activation of the EEG
in 30 epilepsy patients, 8 with GTCSs, and 22 with complex par-
tial seizures. Stimuli consisted of a stressful interview that
elicited strong feelings of anger, sorrow, embarrassment, mirth,
and pleasure. Two thirds of GTCS patients showed no abnormal
changes, while three quarters of complex partial seizure
patients had abnormal findings. Abnormal changes not
observed in previous tracings were seen in nearly half of the
complex partial seizure group, but in only one patient in the
GTCS group. Similar emotional activation of EEG patterns has
been reported by Small et al. (227).
Hypoglycemia
Provocation of convulsive seizures and spike discharges by
hypoglycemia, especially during insulin shock treatment of
schizophrenia (228), has been well documented (229);
spike–wave discharges are usually observed when the patient
has “muscle twitches.” Convulsions and EEG abnormalities
also occur in patients with spontaneous episodes of hyperin-
sulinism secondary to islet cell tumors of the pancreas; the
same finding may be observed in rare patients with nonpan-
creatic tumors associated with hypoglycemia. Hypoglycemia
of nonpancreatic origin (e.g., ketogenic diet) occasionally
facilitates generalized paroxysmal activity (11). Therefore,
EEG activation by fasting may be informative in generalized
epilepsy patients if no paroxysmal discharges could be found
on routine EEG examination. Attempts have also been made
to induce paroxysmal activity through hypoglycemia pro-
duced by tolbutamide (230,231). Hypoglycemic activation
may also by a useful technique for predicting the presence of
 Chapter 12 ■ Activation Methods 231

Figure 12.17 Paroxysmal discharges elicited by video game and Uchida–Kraepelin test in a 16-year-old boy with JME.
The left tracing shows generalized paroxysmal discharges associated with myoclonic seizures of upper extremities; they
occurred 3 minutes after the game (Puyo Puyo) started. Uchida–Kraepelin test at 2 and 3 minutes elicited bilateral parox-
ysmal discharges associated with myoclonic seizures of both hands.

pathology in patients considered for anterior temporal lobec- operative electrocorticographic and depth electrode recording
tomy (232). (239). These methods, however, have not been widely used.

Antiepileptic Drug Withdrawal Procedure for

PHARMACOLOGIC ACTIVATION
Drugs
Alpha-chloralose or alpha-chloralose-scopolamine has occa-
sionally been used for EEG activation in epilepsy patients with
normal routine EEG (233,234). The adult patient was given a
preparation consisting of scopolamine hydrobromide 0.5 mg
and alpha chloralose 500 mg, in capsule form, by mouth (233).
Of 30 epilepsy patients, 22 showed activation efficacies.
Intravenous tripelennamine (Pyribenzamine) is also known to
activate focal or diffuse paroxysmal discharges (235). In addi-
tion, it has been reported that intravenous administration of
some antiparkinsonism agents, such as 1 mg of scopolamine
hydrobromide (236), 10 to 20 mg of procyclidine hydrochloride
(237), or 10 mg of methixene hydrochloride (238), is capable of
provoking focal paroxysmal discharges, especially in temporal
lobe epilepsy patients. Intravenous alfentanil hydrochloride,
which is a short-duration opioid analgesic, has been demon-
strated to be effective in inducing temporal lobe spikes for intra-
Long-Term Monitoring
Before epilepsy surgery, antiepileptic drug (AED) withdrawal in
the video-EEG monitoring environment is currently being car-
ried out for the purpose of increasing the probability of inter-
ictal and ictal abnormalities occurring during the record. The
most important risk of AED withdrawal is the development of
status epilepticus (240,241). Such complications of AED with-
drawal, however, can be prevented if patients are under the care
of experienced staff (8).
Metrazol and Megimide
Pentylenetetrazol (Metrazol) has been used therapeutically to
induce convulsions in schizophrenia patients (242). Ziskind et
al. (243) demonstrated the value of Metrazol in triggering
paroxysmal discharges. A series of similar studies followed:
Kaufman et al. (244), Ziskind and Bercel (245), Cure et al. (246),
Jasper and Courtois (247), and Bancaud (248). The effectiveness
of a given amount of Metrazol depends to a large extent on the
232 Part II ■ Normal EEG
rapidity of its injection as well as the total amount in relation to
body weight (247). These authors recommended dilution of a
10% solution of Metrazol in normal saline so that 1 cc of the
mixture contains 1 mg/kg of body weight. One milliliter is then
given thereafter every 30 seconds.
Among epilepsy patients, those with bilaterally synchronous
EEG disturbances were more sensitive than those with focal
cortical seizures (246). On the other hand, Ajmone-Marsan and
Ralston (249) showed activation of various types of focal
seizures by this method. They reported that the epileptogenic
supplementary motor area appears to be very sensitive to
Metrazol stimulation.
Gastaut (250) described an activation method combined
with Metrazol and IPS (see Ref. (53)). He pointed out that this
test differentiates between epilepsies involving a subcortical
diencephalic mechanism and those of cortical origin. A low
threshold to IPS was found in idiopathic (absence seizure group
in particular) and myoclonic epilepsies. Normal or elevated
thresholds were found in the focal cortical group. Although this
method was more sensitive than activation by Metrazol alone, a
high incidence of low-threshold responses (5 mg/kg or less) in
a number of other conditions such as schizophrenia and hyste-
ria, and even normal subjects, has been observed.
Bemegride (Megimide) was first investigated as a barbiturate
antagonist (251). Megimide was then found to have a convul-
sant action similar to that of Metrazol (252,253).
Despite numerous studies, especially between 1946 and 1970
(254), neither of the above tests is presently considered to be a
satisfactory method for distinguishing between potential
epileptic and nonepileptic individuals. Furthermore, the risk of
failure in seizure control with resection of a focus that was
localized by chemically (Megimide) induced seizures is two
times higher than that with resection of a spontaneous seizure
focus (255). Therefore, drugs are no longer used to induce
seizures for localization in epilepsy surgery (8).
Re-enactment of the Triggering Situation
When a fixed set of ambient circumstances is associated with
convulsive or nonconvulsive paroxysmal attacks, re-enactment
of the situation, if possible, is desirable for reaching a diagnosis.
Fariello et al. (256) reported that re-enactment determined the
diagnosis in 30 of 32 patients with paroxysmal disorders. The
authors claim that an appropriately executed re-enactment
should entail polygraphic recording of at least EEG, electrocar-
diogram, and respiration.
CONCLUDING REMARKS
The greater the number of different activation methods used in
EEG evaluation of an epilepsy patient, the greater the chance of
obtaining abnormal findings. From a practical point of view,
however, desirable activations are those methods that can be
carried out easily, in a short time, with affordable equipment,
and without undesirable side effects for patients. Special meth-
ods of activation are needed in epilepsy patients for whom
seizures tend to appear under certain conditions while awake.
Lastly, considering the methods while awake, a report by Labate
et al. (257) provides us important clues for future EEG studies:
performing standard awake EEG in patients with JME, they
obtained a significantly greater rate of detection of generalized
epileptiform abnormalities in the morning than in an after-
noon session.
ACKNOWLEDGMENTS
The authors wish to thank Professor Y. Tsukahara for his discus-
sions of the manuscript. They are also very grateful to Mr. B.A.
Fast for manuscript preparation assistance.
REFERENCES
1. Foerster O. Hyperventilationsepilepsie. Zbl Ges Neurol Psychiatr.
1924;38:289–293.
2. Nims LF, Gibbs EL, Lennox WG, et al. Adjustment of acid–base
balance of patients with petit mal epilepsy to overventilation. Arch
Neurol Psychiatry. 1940;43:262–269.
3. Morrice JKW. Slow wave production in the EEG, with reference
to hyperpnoea, carbon dioxide and autonomic balance.
Electroencephalogr Clin Neurophysiol. 1956;8:49–72.
4. Achenbach-Ng J, Siao TCP, Marvroudakis N, et al. Effects of rou-
tine hyperventilation on PCO2 and PO2 in normal subjects: impli-
cations for EEG interpretations. J Clin Neurophysiol. 1994;11:
220–225.
5. Gibbs FA, Gibbs EL, Lennox WG. Electroencephalographic response
to overventilation and its relation to age. J Pediatr. 1943;23:497–505.
6. Gotoh F, Meyer JS, Takagi Y. Cerebral effects of hyperventilation in
man. Arch Neurol. 1965;12:410–423.
7. Yamatani M, Konishi T, Murakami M, et al. Hyperventilation activa-
tion on EEG recording in childhood. Epilepsia. 1994;35:1199–1203.
8. Fisch BC, So EL. Activation methods. In: Ebersole JS, Pedley TA,
eds. Current Practice of Clinical Electroencephalography. 3rd ed.
Philadelphia, PA: Lippincott Williams & Wilkins; 2003:246–270.
9. Sherwin I. Hyperventilation: mode of action and application in
electroencephalography. Am J EEG Technol. 1984;24:201–211.
10. Patel VM, Maulsby RL. How hyperventilation alters the EEG: a
review of controversial viewpoints emphasizing neurophysiologi-
cal mechanisms. J Clin Neurophysiol. 1987;4:101–120.
11. Niedermeyer E. The Generalized Epilepsies. Springfield, IL: Charles
C. Thomas; 1972.
12. Esquivel E, Chaussai M, Plouin P, et al. Physical exercise and
voluntary hyperventilation in childhood absence epilepsy.
Electroencephalogr Clin Neurophysiol. 1991;79:127–132.
13. Niedermeyer E. Focal and generalized seizure discharges in the
electroencephalogram and their response to intravenous diazepam.
Int Med Dig. 1972;7:49–61.
14. Bostem F. Hyperventilation. In: Remond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 3D.
Amsterdam: Elsevier; 1976:74–88.
15. Ziegler DK, Hassanein RS, Dick AR. Effect of age and depth of
hyperventilation on a quantitative electroencephalographic
response. Clin Electroencephalogr. 1975;6:184–190.
16. Petersen I, Eeg-Olofsson O. The development of the electroen-
cephalogram in normal children from age 1 through 15 years—
nonparoxysmal activity. Neuropaediatrie. 1971:2:247–304.
17. Kellaway P. An orderly approach to visual analysis: the parameters
of the normal EEG in adults and children. In: Klass DW, Daly DD,
eds. Current Practice of Clinical Electroencephalography. New York:
Raven Press; 1979:69–147.

18. Petersen I, Eeg-Olofsson O. The development of the electroen-
cephalogram in normal children from the age of 1 through 15 years:
non-paroxysmal activity. Neuropadiatrie. 1971;2:247–304.
19. Bickford RG. Activation procedures and special electrodes. In:
Klass DW, Daly DD, eds. Current Practice of Clinical Electro-
encephalography. New York: Raven Press; 1979:269–305.
20. Davis H, Wallace WM. Factors affecting changes produced in elec-
troencephalograms by standardized hyperventilation. Arch Neurol
Psychiatry. 1942;47:606–625.
21. Billinger TW, Frank GS. Effects of posture on EEG slowing during
hyperventilation. Am J EEG Technol. 1969;9:22–27.
22. Dalby MA. Epilepsy and 3 per second spike and wave rhythms.
Acta Neurol Scand. 1969;40(suppl):1–183.
23. Kiloh LG, McComas AJ, Osselton JW. Clinical Electro-
encephalography. 3rd ed. London: Butterworths; 1972.
24. Miley CE, Forster FM. Activation of partial complex seizures by
hyperventilation. Arch Neurol. 1977;34:371–373.
25. Arain AM, Arbogast PG, Abou-Khalil BW. Utility of daily super-
vised hyperventilation during long-term video-EEG monitoring.
J Clin Neurophysiol. 2009;26(1):17–20.
26. Engel J Jr. A practical guide for routine EEG studies in epilepsy.
J Clin Neurophysiol. 1984;1:109–142.
27. Kodama N, Aoki Y, Hiraga H, et al. Electroencephalographic findings
in children with moyamoya disease. Arch Neurol. 1979;36:16–19.
28. Kameyama M, Shirane R, Tsurumi Y, et al. Evaluation of cerebral
blood flow and metabolism in childhood moyamoya disease: an
investigation into “re-build-up” on EEG by positron CT. Child
Nerv Syst. 1986;2:130–133.
29. Diamini N, Goyal S, Jarosz J, et al. Paroxysmal episodes, “re-build up”
phenomenon and moyamoya disease. Epileptic Disord. 2009;11(4):
324–328.
30. AEEGS Guidelines Committee. Guideline one: minimum techni-
cal requirements for performing clinical electroencephalography.
J Clin Neurophysiol. 1994;11:4–5.
31. Epstein MA, Duchowny M, Jayakar P, et al. Altered responsive-
ness during hyperventilation-induced EEG slowing: a non-
epileptic phenomenon in normal children. Epilepsia. 1994;35(6):
1204–1207.
32. Lum LM, Connolly MB, Farrell K, et al. Hyperventilation-induced
high-amplitude rhythmic slowing with altered awareness: a video-
EEG comparison with absence seizures. Epilepsia. 2002;43(1):
1372–1378.
33. Adrian ED, Matthews BHC. The Berger rhythm: potential changes
from the occipital lobes in man. Brain. 1934;57:355–385.
34. Bickford RG, Daly D, Keith HM. Convulsive effects of light stimu-
lation in children. Am J Dis Child. 1953;86:170–183.
35. Walter WG, Dovey VJ, Shipton H. Analysis of the electrical
response of the human cortex to photic stimulation. Nature. 1946;
158:540–541.
36. Kasteleijn-Nolst Trenité DGA. Reflex seizures induced by intermit-
tent light stimulation. In: Zifkin BG, Andermann F, Beaumanoir A,
et al., eds. Reflex Epilepsies and Reflex Seizures: Advances in
Neurology. Vol 75. Philadelphia, PA: Lippincott-Raven; 1998:99–121.
37. Kasteleijn-Nolst Trenité DGA, Binnie CD, Harding GFA, et al. Photic
stimulation: standardization of screening methods. Epilepsia.
1999;40(suppl 4):75–79.
38. Chatrian GE, Bergamini L, Dondey M, et al. A glossary of terms
most commonly used by clinical electroencephalographers.
Electroencephalogr Clin Neurophysiol. 1974;37:538–548.
39. Ellingson RJ. Cortical electrical responses to visual stimulation in
the human infant. Electroencephalogr Clin Neurophysiol. 1960;12:
633–677.
Chapter 12 ■ Activation Methods 233
40. Walter VJ, Walter WG. The effect of physical stimuli on the EEG.
Electroencephalogr Clin Neurophysiol. 1951;2(suppl):60–68.
41. Kooi KA, Tucker PR, Marshall RE. Fundamentals of Electro-
encephalography. Hagerstown, MD: Harper & Row; 1978.
42. Pampiglione G, Harden A. So-called neuronal ceroid lipofusci-
nosis: neurophysiological studies in 60 children. J Neurol
Neurosurg Psychiatry. 1977;40:323–330.
43. Pinsard N, Livet MO, Saint-Jean M. A case of cerebral lipidosis
with an atypical presentation. Rev Electroencephalogr Neurophysiol
Clin. 1978;8:175–179.
44. Coull BM, Pedley TA. Intermittent photic stimulation: clinical use-
fulness of non-convulsive responses. Electroencephalogr Clin
Neurophysiol. 1978;44:353–363.
45. Chiba T, Kitawaki M, Aoki Y, et al. Hemodialysis and electroen-
cephalogram: through a long term observation. Rinsho Noha
(Osaka). 1982;24:195–201.
46. Takahashi T, Tsukahara Y. Influence of red light and pattern on
photic driving. Tohoku J Exp Med. 1979;127:45–52.
47. Harding GFA, Jeavons PM. Photosensitive Epilepsy. New ed.
London: Mac Keith Press; 1994.
48. Tsukahara Y, Takahashi T. Visual stimulator for EEG activation.
Electroencephalogr Clin Neurophysiol. 1973;35:333–335.
49. Tsukahara Y, Takahashi T. Pattern-evoked high-amplitude photic
driving in epileptic patients. In: Ito M, ed. Integrative Control
Functions of the Brain. Vol 2. Tokyo: Kodansha Scientific; 1979:
90–92.
50. Takahashi T. Activation methods. In: Niedermeyer E, Lopes da
Silva F, eds. Electroencephalography: Basic Principles, Clinical
Applications and Related Fields. 3rd ed. Baltimore, MD: Williams &
Wilkins; 1993:241–262.
51. Takahashi T, Tsukahara Y, Kaneda S. EEG activation by use of stro-
boscope and visual stimulator SLS-5100. Tohoku J Exp Med. 1980:
130:403–409.
52. Takahashi T, Kataoka K, Tsukahara Y. Power spectral analysis of
photic driving elicited by flickering dot pattern and red flicker
stimuli in adult psychiatric outpatients—with special reference to
age and gender. Tohoku J Exp Med. 1988;156:165–173.
53. Takahashi T. Activation methods. In: Niedermeyer E, Lopes da
Silva F, eds. Electroencephalography: Basic Principles, Clinical
Applications and Related Fields. 4th ed. Baltimore, MD: Williams &
Wilkins; 1999;261–284.
54. Takahashi T. Photic Driving Response elicited by Low-Luminance
Visual Stimulation: An Atlas. Sendai, Japan: Tohoku University
Press; 2005.
55. Takahashi T. EEG activation by visual stimuli. In: Sato M,
Matsuoka H, eds. The Newest Clinical Electroencephalography.
Tokyo: Asakura Shoten; 1993:323–344.
56. Takahashi T, Fukushi SM, Suzuki T. Diagnosis of amblyopia on the
basis of the photic drivings evoked by flickering dot pattern. J
Sendai City Hosp. 1981;2:21–26.
57. Takahashi T, Tomioka H. Photic driving evoked by hemi-field
flickering dot pattern stimulation in a patient with brain tumor.
Electroencephalogr Clin Neurophysiol. 1985;61:381–384.
58. Gastaut H, Rémond A. L’activation de l’électroencéphalogramme
dans les affections cérébrales non épileptogènes (vers une neuro-
physiolgie clinique). Rev Neurol. 1949;81:594–598.
59. Bickford RG, Sem-Jacobsen GW, White PT, et al. Some observa-
tions on the mechanism of photic and photometrazol activation.
Electroencephalogr Clin Neurophysiol. 1952;4:275–282.
60. Niedermeyer E, Fineyre F, Riley T, et al. Myoclonus and the elec-
troencephalogram: a review. Clin Electroencephalogr. 1979;10:
75–95.
234 Part II ■ Normal EEG
61. Gastaut H, Trevisan C, Naquet R. Diagnostic value of electroen-
cephalographic abnormalities provoked by intermittent photic
stimulation. Electroencephalogr Clin Neurophysiol. 1958;10:194–195
[abstract].
62. Kooi K, Thomas MH, Mortenson FN. Photoconvulsive and photo-
myoclonic responses in adults: an appraisal of their clinical signif-
icance. Neurology. 1960;10:1051–1058.
63. Melsen S. The value of photic stimulation in the diagnosis of
epilepsy. J Nerv Ment Dis. 1959;128:508–519.
64. Reilly EW, Peters JF. Relationship of some varieties of electroen-
cephalographic photosensitivity to clinical convulsive disorders.
Neurology. 1973;23:1040–1057.
65. Small JG. Photoconvulsive and photomyoclonic responses in psychi-
atric patients. Electroencephalogr Clin Neurophysiol. 1971;2:78–88.
66. Newmark ME, Penry JK. Photosensitivity and Epilepsy: A Review.
New York: Raven Press; 1979.
67. Victor M. The role of alcohol in the production of seizures. In:
Niedermeyer E, ed. Modern Problems of Pharmacopsychiatry—
Epilepsy. Vol 4. Basel: Karger; 1970:185–199.
68. Wikler A, Essig CF. Withdrawal seizures following chronic intoxi-
cation with barbiturates and other sedative drugs. In: Niedermeyer
E, ed. Modern Problems of Pharmacopsychiatry— Epilepsy. Vol 4.
Basel: Karger; 1970:170–184.
69. Bax-aan de Stegge BM, van Putten MJAM. A patient with a transient
photomyogenic response. Clin Neurophysiol. 2010;121:118–120.
70. Harding GFA, Harding PF. Televised material and photosensitive
epilepsy. Epilepsia. 1999;40(suppl 4):65–69.
71. Topalkara K, Alarcon G, Binnie CD. Effects of flash frequency and
repetition of intermittent photic stimulation on photoparoxysmal
responses. Seizure. 1998;7:249–255.
72. Klass DW, Fischer-Williams M. Sensory stimulation, sleep and sleep
deprivation. In: Remond A, ed. Handbook of Electroencephalography
and Clinical Neurophysiology. Vol 3D. Amsterdam: Elsevier;
1976:5–73.
73. Jeavons PM, Harding GFA. Photosensitive Epilepsy: A Review of the
Literature and a Study of 460 Patients. London: Heinemann; 1975.
74. Van Egmond P, Binnie, CD, Veldhuizen R. The effect of back-
ground illumination on sensitivity to intermittent photic stimula-
tion. Electroencephalogr Clin. Neurophysiol. 1980;48:599–601.
75. Panayiotopoulos CP. Effectiveness of photic stimulation on various
eye-states in photosensitive epilepsy. J Neurol Sci. 1974;23:
165–173.
76. Hishikawa Y, Yamamoto J, Furuya E, et al. Photosensitive epilepsy:
relationships between the visual evoked responses and the epilep-
tiform discharges induced by intermittent photic stimulation.
Electroencephalogr Clin Neurophysiol. 1967;23:320–334.
77. Sato S, Dreifuss FE, Penry JK. Photic sensitivity of children with
absence seizures in slow wave sleep. Electroencephalogr Clin
Neurophysiol. 1975;39:479–489.
78. Maheshwari MC, Jeavons PM. The clinical significance of occipi-
tal spikes as a sole response to intermittent photic stimulation.
Electroencephalogr Clin Neurophysiol. 1975;39:93–95.
79. Takahashi T. Activation methods. In: Niedermeyer E, Lopes da
Silva F, eds. Electroencephalography: Basic Principles, Clinical
Applications and Related Fields. Baltimore, MD/Munich: Urban &
Schwarzenberg; 1982:179–195.
80. Jeavons PM. The use of photic stimulation in clinical electroen-
cephalography. Proc Electrophysiol Technol Assoc. 1969;16:225–240.
81. Wilkins AJ, Darby CE, Binnie CD, et al. Television epilepsy: the role
of pattern. Electroencephalogr Clin Neurophysiol. 1979;47:163–171.
82. Stevens JR. Control and peripheral factors in epileptic discharge:
clinical studies. Arch Neurol. 1962;7:330–338.
83. Guerrini R, Dravet C, Genton P, et al. Idiopathic photosensitive
occipital lobe epilepsy. Epilepsia. 1995;36:883–891.
84. Kasteleijn-Nolst Trenité DGA, Binnie CD, Meinardi H.
Photosensitive patients: symptoms and signs during intermittent
photic stimulation and their relation to seizures in daily life.
J Neurol Neurosurg Psychiatry. 1987;560:1436–1549.
85. Gambardella A, Aguglia U, Oliveri RL, et al. Photic-induced
epileptic negative myoclonus: a case report. Epilepsia. 1996;37:
492–494.
86. Doose H, Gerken H, Hien-Volpel KF, et al. Genetics of photosen-
sitive epilepsy. Neuropaediatrie. 1969;1:56–73.
87. Takahashi T. Photoconvulsive response and disposition. Folia
Psychiatr Neurol Jpn. 1976;30:349–356.
88. Waltz S, Christen H-J, Doose H. The different patterns of pho-
toparoxysmal response—a genetic study. Electroencephalogr Clin
Neurophysiol. 1992;83:138–145.
89. Quirk JA, Fish DR, Smith SJM, et al. Incidence of photosensitive
epilepsy: a prospective national study. Electroencephalogr Clin
Neurophysiol. 1995;95:260–267.
90. De Graaf AS, Lombard CJ, Classen DA. Influence of ethnic and
geographic factors on the classic photoparoxysmal response in
the electroencephalogram of epilepsy patients. Epilepsia. 1995;36:
219–223.
91. Rushton DN. Space invader epilepsy. Lancet. 1981;1:501.
92. Kasteleijn-Nolst Trenité DGA, Martins da Silva A, Ricci S, et al.
Video games are exciting: a European study of video game-
induced seizures and epilepsy. Epileptic Disord. 2002;4:121–128.
93. Takahashi T. Photosensitive Epilepsy: EEG Diagnosis by Low-
Luminance Visual Stimuli and Preventive Measures. Tokyo: Igaku-
Shoin Publication Service; 2002.
94. Quirk JA, Fish DR, Smith SJM, et al. First seizures associated with
playing electronic screen games: a community-based study in
Great Britain. Ann Neurol. 1995;37:733–737.
95. Binnie CD, Harding GFA, Richens A, et al. Video games and
epileptic seizures—a consensus statement. Seizure. 1994;3:
245–246.
96. Jayakar P, Chiappa KH. Clinical correlations of photoparoxysmal
responses. Electroencephalogr Clin Neurophysiol. 1990;75:251–254.
97. Solomon S, Fine D. The precipitation of seizures by photic stim-
ulation in a patient with hypoparathyroidism. J Nerv Ment Dis.
1960;130:253–260.
98. Victor M, Brausch C. The role of abstinence in the genesis of
alcoholic epilepsy. Epilepsia. 1967;8:1–20.
99. Gastaut H. Introduction to the study of organic generalized
epilepsies. In: Gastaut H, Jasper H, Bancaud J, et al., eds. The
Physiopathogenesis of the Epilepsies. Springfield, IL: Charles C.
Thomas; 1969:147–157.
100. Lee RG, Blair RDG. Evolution of EEG and visual evoked response
changes in Jakob–Creutzfeldt disease. Electroencephalogr Clin
Neurophysiol. 1973;35:133–142.
101. Gilliam FG, Chiappa KH. Significance of spontaneous epilepti-
form abnormalities associated with a photoparoxysmal response.
Neurology. 1995;45(3):453–455.
102. Bickford RG, Klass DW. Sensory precipitation and reflex mecha-
nisms. In: Jasper HH, Ward AA, Pope A, eds. Basic Mechanisms of
the Epilepsies. Boston, MA: Little, Brown and Company;
1969:543–564.
103. Takahashi T, Tsukahara Y. Generalized paroxysmal discharges
induced by visual stimuli and eye movements. Tohoku J Exp Med.
1975;115:1–10.
104. Jeavons PM, Harding GFA, Panayiotopoulos CP, et al. The
effect of geometric patterns combined with intermittent photic

stimulation in photosensitive epilepsy. Electroencephalogr Clin
Neurophysiol. 1972;33:221–224.
105. Takahashi T, Tsukahara Y. Photoconvulsive response induced by
use of “visual stimulator”. Tohoku J Exp Med. 1980;130:273–281.
106. Takahashi T, Tsukahara Y. Influence of color on the photoconvulsive
response. Electroencephalogr Clin Neurophysiol. 1976;41:124–136.
107. Takahashi T, Tsukahara Y. Usefulness of blue sunglasses in pho-
tosensitive epilepsy. Epilepsia. 1992;33:517–521.
108. Takahashi T. Techniques of intermittent photic stimulation and
paroxysmal responses. Am J EEG Technol. 1989;29:205–218.
109. Takahashi T. Pathophysiological mechanisms of photosensitivity
in IGEs. In: Malafosse A, Genton P, Hirsch E, et al., eds. Idiopathic
Generalized Epilepsies: Clinical, Experimental and Genetic Aspects.
London: John Libbey; 1994:305–315.
110. Binnie CD, Jeavons PM. Photosensitive epilepsies. In: Roger J,
Bureau M, Dravet C, et al., eds. Epileptic Syndromes in Infancy,
Childhood and Adolescence. 2nd ed. London: John Libbey;
1992:299–305.
111. Takahashi T. EEG diagnosis of photosensitive seizures and its
preventive measures. Jpn J Clin Neurophysiol. 2000;28:236–245.
112. Chiappa KH, Hill RA, Huang-Hellinger F, et al. Photosensitive
epilepsy studied by functional magnetic resonance imaging and
magnetic resonance spectroscopy. Epilepsia. 1999;40(suppl 4):3–7.
113. Hill RA, Chiappa KH, Huang-Hellinger F, et al.. Hemodynamic
and metabolic aspects of photosensitive epilepsy revealed by
functional magnetic resonance imaging and magnetic resonance
spectroscopy. Epilepsia. 1999;40:912–920.
114. Takahashi T. EEG activation by movement of the eyelids and
eyes. Rinsho Noha (Osaka). 1976;18:334–344.
115. Westmoreland BF, Sharbrough FW. Posterior slow wave tran-
sients associated with eye blinks in children. Am J EEG Technol.
1975;15:14–19.
116. Gastaut H, Tassinari CA. Triggering mechanisms in epilepsy: the
electro-clinical point of view. Epilepsia. 1966;7:85–138.
117. Green JB. Seizures on closing the eyes. Neurology. 1968;18:391–396.
118. Darby CE, De Korte RA, Binnie CD, et al. The self-induction of
epileptic seizures by eye closure. Epilepsia. 1980;21:31–42.
119. Forster FM. Reflex Epilepsy, Behavioral Therapy and Conditional
Reflexes. Springfield, IL: Charles C. Thomas; 1977.
120. Gumnit RJ, Niedermeyer E, Spreen O. Seizure activity uniquely
inhibited by patterned vision. Arch Neurol. 1965;13:363–368.
121. Takahashi T. Influence on EEGs of gazing at a fixation point and
of ganzfeld. Brain Nerve (Tokyo). 1976;28:95–103.
122. Colon EJ, Vingerhoets HM, Notermans SLH, et al. Off-response:
its clinical incidence in very young children. Electroencephalogr
Clin Neurophysiol. 1979;46:601–604.
123. Takahashi T, Tsukahara Y. EEG activation by red color. Igaku No
Ayumi (Tokyo). 1972;83:25–26.
124. Lugaresi E, Cirignotta F, Montagna P. Occipital lobe epilepsy with
scotosensitive seizures: the role of central vision. Epilepsia.
1984;25:115–120.
125. Panayiotopoulos CP. Conversion of photosensitive to scotosensi-
tive epilepsy: report of a case. Neurology. 1979;29:1550–1554.
126. Panayiotopoulos CP. With comments and contributions from
Binnie, C.D., and Takahashi, T. Fixation-off-sensitive epilepsies:
clinical and EEG characteristics. In: Wolf P, ed. Epileptic Seizures
and Syndromes. London: John Libbey; 1994:55–66.
127. Panayiotopoulos CP. Fixation-off sensitive epilepsies: clinical and
EEG characteristics. In: Zifkin BG, Andermann F, Beaumanoir
A, et al., eds. Reflex Epilepsies and Reflex Seizures: Advances in
Neurology. Vol 75. Philadelphia, PA: Lippincott-Raven; 1998:
139–157.
Chapter 12 ■ Activation Methods 235
128. Bickford RG, Klass DW. Stimulus factors in the mechanism of tel-
evision induced seizures. Trans Am Neurol Assoc. 1962;87:176–178.
129. Binnie CD, Wilkins AJ. Visually induced seizures not caused by
flicker (intermittent light stimulation). In: Zifkin BG,
Andermann F, Beaumanoir A, et al., eds. Reflex Epilepsies and
Reflex Seizures: Advances in Neurology. Vol 75. Philadelphia, PA:
Lippincott-Raven; 1998:123–138.
130. Chatrian GE, Lettich E, Miller LH, et al. Pattern-sensitive
epilepsy. 1. An electrographic study of its mechanisms. Epilepsia.
1970;11:125–149.
131. Darby CE, Wilkins AJ, Binnie CD, et al. Routine testing for pat-
tern sensitivity. J Electrophysiol Technol. 1980;6:202–210.
132. Klass DW. Pattern activation of seizures. In: Lüders HO,
Noachtar S, eds. Epileptic Seizures: Pathophysiology and Clinical
Semiology. New York: Churchill Livingstone; 2000:598–608.
133. Wilkins AJ, Binnie CD, Darby CE. Visually-induced seizures. Prog
Neurobiol. 1980;15:85–117.
134. Binnie CD, Findlay J, Wilkins AJ. Mechanisms of epileptogenesis
in photosensitive epilepsy implied by the effects of moving pat-
terns. Electroencephalogr Clin Neurophysiol. 1985;61:1–6.
135. Takahashi T, Sasaki M. Pattern sensitivity: its activation method
and clinical EEG findings. Rinsho Nohha (Osaka). 1971;13:
727–732.
136. Takahashi T. Study of visual epilepsy. 4. Pattern activation and
clinical EEG findings. Clin Neurol (Tokyo). 1974;14:479–488.
137. Radhakrishnan K, Sunku AJ, Donat JF, et al. Pattern-induced
negative occipital potentials (PINOP). J Clin Neurophysiol. 2007;
24:277–280.
138. Stefansson SB, Darby CE, Wilkins AJ, et al. Television epilepsy
and pattern sensitivity. Br Med J. 1977;2:88–90.
139. Koshino Y, Niedermeyer E. Enhancement of rolandic mu rhythm
by pattern vision. Electroencephalogr Clin Neurophysiol. 1975;38:
535–538.
140. Takahashi T, Nakasato N, Yokoyama H, et al. Low-luminance
visual stimuli compared with stroboscopic IPS in eliciting PPR in
photosensitive patients. Epilepsia. 1999;40(suppl 4):44–49.
141. Soso MJ, Lettich E, Belgum JH. Pattern-sensitive epilepsy. II.
Effects of pattern orientation and hemifield stimulation.
Epilepsia. 1980;21:313–323.
142. Takahashi T. Lateral hemifield flickering pattern stimulation in a
patient with pattern-sensitive epilepsy. Epilepsia. 1983;24:548–556.
143. Takahashi T. Hemifield red flicker stimulation in a patient with
pattern-sensitive epilepsy. Epilepsia. 1984;25:223–228.
144. Wilkins AJ, Binnie CD, Darby CE. Interhemispheric differences
in photosensitive epilepsy. I. Pattern sensitivity thresholds.
Electroencephalogr Clin Neurophysiol. 1981;52:461–468.
145. Takahashi T. Hemi- and quadrant-field visual stimuli in EEG
diagnosis of photosensitive epilepsy. In: Kimura J, Shibasaki H,
eds. Recent Advances in Clinical Neurophysiology. Amsterdam:
Elsevier; 1996:223–228.
146. Takahashi T, Tsukahara Y. Photoparoxysmal response elicited by
flickering dot pattern stimulation and its optimal spatial fre-
quency of provocation. Electroencephalogr Clin Neurophysiol.
1998;106:40–43.
147. Takahashi T, Tsukahara Y. Photoparoxysmal response elicited by
flickering dot pattern stimulation to the center and periphery.
Clin Neurophysiol. 2000;111:1968–1973.
148. Takahashi T, Tsukahara Y. Photoparoxysmal response-evoked by
regional visual stimuli in a photosensitive epilepsy patient. In:
Koga Y, Nagata K, Hirata K, eds. Pan-Pacific Conference on Brain
Topography. Amsterdam: Elsevier; 1997:507–511.
149. Zeki S. A Vision of the Brain. Oxford: Blackwell Scientific; 1993.
236 Part II ■ Normal EEG
150. Allison T, Begleiter A, McCarthy G, et al. Electrophysiological
studies of color processing in human visual cortex. Electro-
encephalogr Clin Neurophysiol. 1993;88:343–355.
151. Lueck CJ, Zeki S, Friston KJ, et al. The colour centre in the cere-
bral cortex of man. Nature. 1989;340:386–389.
152. Takahashi T. EEG Atlas of Photosensitive Epilepsy Studied Using
Low-Luminance Visual Stimulation: An Atlas. Sendai, Japan:
Tohoku University Press; 2008.
153. Arellano AP, Schwab RS, Casby JU. Sonic activation.
Electroencephalogr Clin Neurophysiol. 1950;2:215–217.
154. Gastaut HJ, Benoit PH, Vigouroux M, et al. Potentiels évoqés par
des stimuli auditifs sur la région temporale de certains épilep-
tiques. Electroencephalogr Clin Neurophysiol. 1954;6:557–564.
155. Peet RM, Daly DD, Bickoford RG. Some clinical and electroen-
cephalographic responses to sound stimulation in epileptic
patients. Trans Am Neurol Assoc. 1952;7:215–217.
156. Takahashi T. EEG responses to various stimuli. In: Shimazono Y,
Kitamura K, Otomo E, eds. Atlas of Clinical EEG. Tokyo: Bunkodo;
1974:157–193.
157. Vigouroux M, Benoit P, Gastaut H. Evoked potentials by auditory
stimuli in the temporal region in an epileptic patient.
Electroencephalogr Clin Neurophysiol. 1954;6:163–164.
158. Prechtl HFR. Provocation of electroencephalographic changes in
the temporal region by intermittent acoustic stimuli.
Electroencephalogr Clin Neurophysiol. 1959;11:511–519.
159. Wieser HG, Hungerbueler H, Siegel AM, et al. Musicogenic
epilepsy: review of the literature and case report with ictal single
photon emission computed tomography. Epilepsia. 1997;38:
200–207.
160. Rosenow F, Lüders HO. Hearing-induced seizures. In: Lüders
HO, Noachtar S, eds. Epileptic Seizures: Pathophysiology and
Clinical Semiology. Philadelphia, PA: Churchill Livingstone;
2000:580–584.
161. De Marco P, Negrin P. Parietal focal spikes evoked by contralat-
eral tactile somatotopic stimulation in four non-epileptic sub-
jects. Electroencephalogr Clin Neurophysiol. 1973;34:308–312.
162. De Marco P, Tassinari CA. Extreme somatosensory evoked
potential (ESEP): an EEG sign forecasting the possible occur-
rence of seizures in children. Epilepsia. 1981;22:569–575.
163. Deonna T. Reflex seizures with somatosensory precipitation:
clinical and electroencephalographic patterns and differential
diagnosis, with emphasis on reflex myoclonic epilepsy of infancy.
In: Zifkin BG, Andermann F, Beaumanoir A, et al., eds. Reflex
Epilepsies and Reflex Seizures: Advances in Neurology. Vol 75.
Philadelphia, PA: Lippincott-Raven; 1998:193–206.
164. Vignal J-P, Biraben A, Chauvel PY, et al. Reflex partial seizures of
sensorimotor cortex (including cortical reflex myoclonus and
startle epilepsy). In: Zifkin BG, Andermann F, Beaumanoir A,
et al., eds. Reflex Epilepsies and Reflex Seizures: Advances in
Neurology. Vol 75. Philadelphia, PA: Lippincott-Raven; 1998:
207–226.
165. Holmes GL, Blair S, Eisenberg E, et al. Tooth-brushing-induced
epilepsy. Epilepsia. 1982;23:657–661.
166. O’Brien TJ, Hogan RE, Sedal L, et al. Tooth-brushing epilepsy: a
report of a case with structural and functional imaging and elec-
trophysiology demonstrating a right frontal focus. Epilepsia.
1996;37:694–697.
167. Langill L, Wong PKW. Tactile-evoked rolandic discharges: a
benign finding? Epilepsia. 2003;44:221–227.
168. Servit Z. The application of reflex theory in the interpretation of
the clinical picture, genesis and treatment of epilepsy. Epilepsia.
1962;3:209–228.
169. Berkovic SF, So NK, Andermann F. Progressive myoclonus
epilepsies: clinical and neurophysiological diagnosis. J Clin
Neurophysiol. 1991;8:261–274.
170. Dawson GD. Investigations on a patient subject to myoclonic
seizures after sensory stimulation. J Neurol Neurosurg Psychiatry.
1947;10:141–162.
171. Halliday AM. The electrophysiological study of myoclonus in
man. Brain. 1967;90:241–284.
172. Shibasaki H, Yamashita Y, Neshige R, et al. Pathogenesis of giant
somatosensory evoked potentials in progressive myoclonic
epilepsy. Brain. 1985;108:225–240.
173. Sutton GG, Mayer RF. Focal reflex myoclonus. J Neurol Neurosurg
Psychiatry. 1974;37:207–217.
174. Mima T, Terada K, Ikeda A, et al. Afferent mechanism of cortical
myoclonus studied by proprioception-related SEPs. Electro-
encephalogr Clin Neurophysiol. 1997;104:51–59.
175. Plasmati R, Michelucci R, Forti A, et al. The neurophysiological
features of benign partial epilepsy with rolandic spikes. In: Degan
R, Dreifuss FE, eds. Benign Localized and Generalized Epilepsies of
Early Childhood [Epilepsy Res. suppl. 6]. Amsterdam: Elsevier;
1992:45–48.
176. Plasmati R, Pastorelli F, Tassinari CA. Giant midlatency somatosen-
sory evoked potentials in benign epilepsy with rolandic spikes. In:
Lüders HO, Noachtar S, eds. Epileptic Seizures: Pathophysiology and
Clinical Semiology. New York: Churchill Livingstone; 2000:62–68.
177. Goldie L, Green JN. A study of the psychological factors in a case
of sensory reflex epilepsy. Brain. 1959;82:505–524.
178. Van Cott AC, Blatt I, Brenner RP. Stimulus-sensitive seizures in
postanoxic coma. Epilepsia. 1996;37:868–874.
179. Koutroumanidis M, Tsatsou K, Bonakis A, et al. Stimulus-
induced bilateral central periodic discharges, cortical myoclonus
and arousal responses in mild reversible coma. Clin Neurophysiol.
2008;119:2459–2464.
180. Behrman S, Wyke BD. Vestibulogenic seizures: a consideration of
vertiginous seizures, with particular reference to convulsions
produced by stimulation of labyrinthine receptors. Brain.
1958;81:529–541.
181. Munter M, Goetze W, Krokowski G. Telemetrische EEG.
Untersuchungen ratatorischer Vestibulatisreizung. Dtsch Z
Nervenheilkd. 1964;186:137–148.
182. Karbowski K. Epileptic seizures induced by vestibular and audi-
tory stimuli. In: Beaumanoir A, Gastaut H, Naquet R, eds. Reflex
Seizures and Reflex Epilepsies. Geneva: Editions Médicine &
Hygiène; 1989:255–260.
183. Acharya JN, Kotagal P. Activation of seizures by hot water. In:
Lüders HO, Noachtar S, eds. Epileptic Seizures: Pathophysiology
and Clinical Semiology. New York: Churchill Livingstone;
2000:615–627.
184. Mani KS, Mani AJ, Ramesh CK. Hot-water epilepsy—a peculiar
type of reflex epilepsy: clinical and EEG features in 108 cases.
Trans Am Neurol Assoc. 1975;99:224–226.
185. Lisovoski F, Prier S, Koskas P, et al. Hot-water epilepsy in an adult:
ictal EEG, MRI and SPECT features. Seizure. 1992;1:203–206.
186. Rémillard GM, Andermann F, Zifkin BG, et al. Eating epilepsy: a
study of ten surgically treated patients suggests the presence of
two separate syndromes. In: Beaumanoir A, Gastaut H, Naquet
R, eds. Reflex Seizures and Reflex Epilepsies. Geneva: Editions
Médecine & Hygiène; 1989:289–300.
187. Rémillard GM, Zifkin BG, Andermann F. Seizures induced by
eating. In: Zifkin BG, Andermann F, Beaumanoir A, et al., eds.
Reflex Epilepsies and Reflex Seizures: Advances in Neurology.
Vol 75. Philadelphia, PA: Lippincott-Raven; 1998:227–240.

188. Alajouanine T, Gastaut H. La syncinésie-sursaut et l’épilep-
sitesursaut è déclanchement sensorial ou sensitif inopiné. Les
faits anatomocliniques (15 observations). Rev Neurol. 1955;93:
29–41.
189. Rosenow F, Lüders HO. Startle-induced seizures. In: Lüders HO,
Noachtar S, eds. Epileptic Seizures: Pathophysiology and Clinical
Semiology. Philadelphia, PA: Churchill Livingstone; 2000:585–592.
190. Aguglia U, Tinuper P, Gastaut H. Startle-induced epileptic
seizures. Epilepsia. 1984;25:712–720.
191. Manford MR, Fish DR, Shorvon SD. Startle provoked epileptic
seizures: features in 19 patients. J Neurol Neurosurg Psychiatry.
1996;61:151–156.
192. Saenz-Lope E, Herranz FJ, Masdeu JC. Startle epilepsy: a clinical
study. Ann Neurol. 1984;16:78–81.
193. Bancaud J, Talairach J, Lamarche M, et al. Hypothèses neuro-
physiopathologiques sur l’épilepsie-sursault chez l’homme. Rev
Neurol. 1975;131:559–571.
194. Loiseau P, Duché B. Seizures induced by movement. In:
Beaumanoir A, Gastaut H, Naquet R, eds. Reflex Seizures and
Reflex Epilepsies. Geneva: Editions Médicine & Hygiène; 1989:
109–114.
195. Tibussek D, Wohlrab G, Boltshauser E, et al. Proven startle-
provoked epileptic seizures in childhood: semiologic and electro-
physiologic variability. Epilepsia. 2006;47:1050–1058.
196. Hufnagel A, Elger CE, Durwen HF, et al. Activation of the epilep-
tic focus by transcranial magnetic stimulation of the human
brain. Ann Neurol. 1990;27:49–60.
197. Schuler P, Claus D, Stefan H. Hyperventilation and transcranial
magnetic stimulation: two methods of activation of epileptiform
EEG activity in comparison. J Clin Neurophysiol. 1993;10:111–115.
198. Tassinari CA, Cincotta M, Zaccara G, et al. Transcranial magnetic
stimulation and epilepsy. Clin Neurophysiol. 2003;114:777–798.
199. Green JB. Self-induced seizures. Arch Neurol. 1966;15:579–586.
200. Vignaendra V, Ghee LH, Lee LC, et al. Epileptic discharges trig-
gered by blinking and eye closure. Electroencephalogr Clin
Neurophysiol. 1976;40:491–498.
201. Nadkarni MA, Postolache V, Gold A, et al. Central mid-temporal
spikes triggered by blinking. Electroencephalogr Clin Neurophysiol.
1994;90:36–39.
202. Terzano MG, Parrino L, Manzoni GC, et al. Seizures triggered by
blinking when beginning to speak. Arch Neurol. 1983;40:103–106.
203. Silbert PL, Radhakrishnan K, Johnson J, et al. The significance of
the phi rhythm. Electroencephalogr Clin Neurophysiol. 1995;95:
71–76.
204. Lewis JA. Eye closure as a motor trigger for seizures. Neurology.
1972;22:1145–1150.
205. Robinson LJ. Induction of seizures by closing of the eyes, or by
ocular pressure, in a patient with epilepsy. J Nerv Ment Dis.
1939;90:333–336.
206. Tieber E. Anfallmuster bei Augenschluss. Bericht über drei Fälle
in einer Familie. Neuropaediatrie. 1972;3:305–312.
207. Shanzer S, Aprial R, Atkin A. Seizures induced by eye deviation.
Arch Neurol. 1965;13:621–626.
208. Toma S, Mano T, Shiozawa Z. Seizure discharges induced by eye
closure. Clin Neurol (Tokyo). 1975;15:340–346.
209. Vignaendra V, Lim CL. Epileptic discharges triggered by eye con-
vergence. Neurology. 1978;28:589–591.
210. Arseni C, Stoica I, Serbanescu T. Electro-clinical investigations on
the role of proprioceptive stimuli in the onset and arrest of con-
vulsive epileptic paroxysms. Epilepsia. 1967;8:162–170.
211. Hudgins RL, Corbin KB. An uncommon seizure disorder: famil-
ial paroxysmal choreoathetosis. Brain. 1966;89:199–204.
Chapter 12 ■ Activation Methods 237
212. Lishman WA, Symonds CP, Whitty CW, et al. Seizures induced by
movement. Brain. 1962;85:93–108.
213. Radhakrishnan K, Silbert PL, Klass DW. Reading epilepsy: an
appraisal of 20 patients diagnosed at the Mayo Clinic, Rochester,
Minnesota, between 1949 and 1989, and delineation of the
epileptic syndrome. Brain. 1995;118:75–89.
214. Wolf P. Activation of seizures by reading and praxis. In: Lüders
HO, Noachtar S, eds. Epileptic Seizures: Pathophysiology and
Clinical Semiology. New York: Churchill Livingstone; 2000:
609–614.
215. Mayer T, Wolf P. Perioral reflex myoclonia in reading epilepsy
and juvenile myoclonic epilepsy. In: Wolf P, Inoue Y, Jifkin B, eds.
Reflex Epilepsies: Progress in Understanding. Esher, UK: John
Libbey Eurotext; 2004:57–70.
216. Goossens LAZ, Andermann F, Andermann E, et al. Reflex seizures
induced by calculation, card or board games, and spatial tasks: a
review of 25 patients and delineation of the epileptic syndrome.
Neurology. 1990;40:1171–1176.
217. Inoue Y, Seino M, Kubota H, et al. Epilepsy with praxis-induced
seizures. In: Wolf P, ed. Epileptic Seizures and Syndromes. London:
John Libbey; 1994:81–92.
218. Inoue Y, Zifkin B. Praxis induction and thinking induction: one
or two mechanisms? A controversy. In: Wolf P, Inoue Y, Jifkin B,
eds. Reflex Epilepsies: Progress in Understanding. Esher, UK: John
Libbey Eurotext; 2004:41–55.
219. Matsuoka H, Takahashi T, Sasaki M, et al. Neuropsychological
EEG activation in patients with epilepsy. Brain. 2000;123:
318–330.
220. Takahashi Y, Shigematsu H, Kubota H, et al. Nonphotosensitive
video game-induced partial seizures. Epilepsia. 1995;36:837–841.
221. Ishihara T, Yoshii N. Multivariate analytic study of EEG and men-
tal activity in juvenile delinquents. Electroencephalogr Clin
Neurophysiol. 1972;33:71–80.
222. Sasaki K, Tsujimoto T, Nishikawa S, et al. Frontal mental theta
wave recorded simultaneously with magnetoencephalography
and electroencephalography. Neurosci Res. 1996;26:79–81.
223. Yamaguchi Y. Frontal midline theta activity. In: Yamaguchi N,
Fujisawa K, eds. Recent Advances in EEG and EMG Data
Processing. Amsterdam: Elsevier; 1981:391–396.
224. Guey J, Bureau M, Dravet C, et al. A study of the rhythm of petit
mal absences in children in relation to prevailing situations. The
use of EEG telemetry during psychological examinations, school
exercises and periods of inactivity. Epilepsia. 1969;10:441–451.
225. Temkin NR, Davis GR. Stress as a risk factor for seizures among
adults with epilepsy. Epilepsia. 1984;25:450–456.
226. Stevens JR. Emotional activation of the electroencephalogram in
patients with convulsive disorders. J Nerv Ment Dis. 1959;128:
339–351.
227. Small JG, Stevens JR, Milstein V. Electroclinical correlates of
emotional activation. J Nerv Ment Dis. 1964;138:146–155.
228. Sakel M. Neue Behandlugsmethode der Schizophrenie. Wien:
Moritz Perthes; 1935.
229. Poire R. Hypoglycemic epilepsy: clinical, electrographic and bio-
logical study during induced hypoglycemia in man. In: Gastaut
H, Jasper HH, Bancaud J, et al., eds. The Physiopathogenesis of the
Epilepsies. Springfield, IL: Charles C. Thomas; 1969:75–110.
230. Green JB. The activation of electroencephalographic abnormali-
ties by tolbutamide-induced hypoglycemia. Neurology. 1963;13:
192–200.
231. Mosowich A. The role of oral hyperglycaemiants as activators in
epilepsy. In: Chorobski J, ed. Neurological Problems. Oxford:
Pergamon; 1967:117–130.
238 Part II ■ Normal EEG
232. Sperling MR. Hypoglycemic activation of focal abnormalities in
the EEG of patients considered for temporal lobectomy.
Electroencephalogr Clin Neurophysiol. 1984;58:506–512.
233. Bercel NA. Experience with a combination of scopolamine and
alpha chloralose (S.A.C.) in activating normal EEG of epileptics.
Electroencephalogr Clin Neurophysiol. 1953;5:297–304.
234. Verdeaux G, Verdeaux J, Marty R. L’activation des élec-
troencéphalogrammes par le chloralose. Electroencephalogr Clin
Neurophysiol. 1954;6:19–28.
235. King G, Weeks SD. Pyribenzamine activation of the electroen-
cephalogram. Electroencephalogr Clin Neurophysiol. 1965;18:
503–507.
236. Vas CJ, Exley KA, Parsonage MJ. An appraisal of hyoscine as an
EEG activation agent. Electroencephalogr Clin Neurophysiol. 1967;
22:373–377.
237. Vas CJ, Exley KA, Parsonage MJ. Activation of the EEG by procy-
clidine hydrochloride in temporal lobe epilepsy. Epilepsia.
1967;8:241–251.
238. Vas CJ, Exley KA, Parsonage MJ. Methixene hydrochloride as an
EEG activating agent. Epilepsia. 1967;8:252–259.
239. Cascino GD, So EL, Sharbrough FW, et al. Alfentanil-induced
epileptiform activity in patients with partial epilepsy. J Clin
Neurophysiol. 1993;10:520–525.
240. Fagan K, Lee S. Prolonged confusion following convulsion due to
generalized nonconvulsive status epilepticus. Neurology. 1990;40:
1689–1694.
241. Simon R. Physiologic consequences of status epilepticus.
Epilepsia. 1985;26(suppl 1):S58–S66.
242. von Meduna I. Versuch über die biologische Beeinflussung des
Ablaufs der Schizophrenie. I. Campher und Cardiazol Krämpfe.
Zbl Ges Neurol Psychiatr. 1935;152:235–262.
243. Ziskind E, Sjaardema H, Bercel NA. Minimal electroencephalo-
graphic response to Metrazol as a method for measuring the con-
vulsive threshold for use in human beings. Science. 1946;104:
462–463.
244. Kaufman IC, Marshall C, Walker AE. Metrazol activated elec-
troencephalography. Res Publ Assoc Nerv Ment Dis. 1946;26:
476–486.
245. Ziskind E, Bercel NA. Preconvulsive paroxysmal electroen-
cephalographic changes after Metrazol injection. Res Publ Assoc
Res Nerv Ment Dis. 1946;26:487–501.
246. Cure C, Rasmussen T, Jasper HH. Activation of seizures and elec-
troencephalographic disturbances in epileptic and in control sub-
jects with “Metrazol”. Arch Neurol Psychiatry. 1948;59:691–717.
247. Jasper H, Courtois G. A practical method for uniform activation
with intravenous Metrazol. Electroencephalogr Clin Neurophysiol.
1953;5:443–444.
248. Bancaud J. EEG activation by Metrazol and Megimide in the diag-
nosis of epilepsy. In: Remond A, ed. Handbook of Electro-
encephalography and Clinical Neurophysiology. Amsterdam:
Elsevier; 1976:105–120.
249. Ajmone-Marsan C, Ralston BL. The Epileptic Seizure: Its
Functional Morphology and Diagnostic Significance. Springfield,
IL: Charles C. Thomas; 1957.
250. Gastaut H. Combined photic and Metrazol activation of the
brain. Electroencephalogr Clin Neurophysiol. 1950;2:249–261.
251. Shulman A, Shaw F, Cass N, et al. A new treatment of barbiturates
intoxication. Br Med J. 1955;1:1238–1244.
252. Delay J, Schuller E, Drossopoulo G, et al. Un nouvel activant des
électroencéphalogrammes: l’imide de l’acide éthyl-méthyl glu-
tarique (NP 13), ou Mégimide. Rev Neurol. 1956;94:315–318.
253. Drossopoulo G, Gastaut H, Verdeaux G, et al. Comparison of
EEG “activation” by pentamethylenetetrazol (Metrazol) and
bemegride (Megimide). Electroencephalogr Clin Neurophysiol.
1956;8:710–711.
254. Ajmone-Marsan C. Pentylenetetrazol: historical notes and com-
ments on its electroencephalographic activation properties. In:
Lüders HO, Noachtar S, eds. Epileptic Seizures: Pathophysiology
and Clinical Semiology. New York: Churchill Livingstone;
2000:563–569.
255. Wieser HG, Bancaud J, Talairach J, et al. Comparative value of
spontaneous and chemically and electrically induced seizures in
establishing the lateralization of temporal lobe seizures.
Epilepsia. 1979;20:47–59.
256. Fariello RG, Booker HE, Chun RWM, et al. Reenactment of the
triggering situation for the diagnosis of epilepsy. Neurology. 1983;
33:878–884.
257. Labate A, Ambrisio R, Gambardella A, et al. Usefulness of a
morning routine EEG recording in patients with juvenile
myoclonic epilepsy. Epilepsy Res. 2007;77:17–21.

Artifacts of Recording
BARBARA DWORETZKY, SUSAN HERMAN, AND WILLIAM O. TATUM IV
R ecording electric activity from the brain is subject
to extracerebral interference or “artifact” that is present
in essentially every electroencephalogram (EEG) (1).
Artifacts may be derived from a variety of sources. They can
obscure the cortically generated signal to a variable degree and
render it impossible to interpret or lead to misinterpretation of
the EEG as abnormal and possibly “epileptic” in origin. Failure
to recognize such potentials as artifact is one of the leading rea-
sons for inappropriate use of antiepileptic medications, and a
common reason for patient referral to an epilepsy center (2,3).
The ideal EEG reflects purely cerebral activity and negligible
extracerebral electric activity. It is performed by a qualified EEG
technologist who ideally creates an artifact-free recording
within the confines of a technically adequate recording on a
cooperative patient in a quiet and controlled environment. This
ideal EEG is pragmatically unachievable but the principles
behind such a recording form the basis for recognizing the
“contaminants” or “noise” that exist in the locations where EEG
is performed. EEG interpretation is an acquired visual art, and
the identification of common artifacts is often a matter of
“automatic” recognition by the reader.
To minimize confusion generated by artifact, approved
guidelines for electrode placement are essential (4,5). With the
ubiquitous use of digital EEG synchronized with video, newly
discovered rhythmic artifacts are being identified on a regular
basis in special care units (SCUs). Offline reformatting of the
EEG is not only possible, but also desirable, placing the respon-
sibility of recognition, identification, and elimination of EEG
artifact on the physician interpreting the EEG as well as the
technologist. The burgeoning field of epilepsy and ICU moni-
toring with video correlation has led to easy confirmation of
artifact sources and has produced many “art(ifact) collectors.”
Some artifacts are extremely common and desirable for routine
recording purposes. For example, eye movements can help
define the normal stages of sleep. The foundation for recording
principles in EEG may be subdivided into artifacts that are
physiologic and those that are nonphysiologic (Table 13.1).
Physiologic artifacts include those generated by the patient’s
biologic activity with electric potentials that are generated near
the head (muscles, eyes, heart), or by patient movement. These
may require camouflaging or ignoring the extracerebral source
since elimination is not always possible. Nonphysiologic arti-
facts are generated anywhere in the EEG recording system from
the electrode–scalp interface to the instrument itself or from
the environment, including from devices near or within the
patient. While many biologic artifacts show characteristic
CHAPTER
13
waveform morphologies and electric fields and therefore can be
fairly easily identified, nonphysiologic artifacts show a wide
variety of morphologies that can either distort or obscure nor-
mal EEG activity. In the most severe cases, artifact may render
the recording uninterpretable. When artifacts result from
equipment malfunction, the EEG machine should be immedi-
ately tested to ensure there are no electric safety hazards. In the
digital age of EEG recording, it is easier to know with certainty
the generators of EEG contaminants by examining the patient
relative to the recording, and subsequently successfully elimi-
nating the artifact that is identified (proof of principle).
Illustrations of artifacts are provided to enhance the principles
in this chapter.
GENERAL PRINCIPLES
The recognition of EEG artifact is an acquired skill. It relies on
maintaining a low threshold to expect artifact in every epoch of
the recording. For certain ubiquitous artifacts such as eye
movements, noticing diminished amplitudes, slower frequen-
cies, or changes in the direction of the dipoles, for example, may
suggest a state change (i.e., drowsiness, rapid eye movement
[REM] sleep). A marked asymmetry in eye movements may
suggest asymmetric electrode placement, enucleation (pros-
thetic eye), or a frontal skull defect (6).
There are some general principles that can be applied when
confronting possible EEG artifact. Artifact caused by loose elec-
trodes can be distinguished from normal activity by its some-
times characteristic shape, positive charge, restriction to a single
electrode, and apparent superimposition on normal cerebral
Tabl e 1 3 . 1
EEG Artifact
Physiologic Nonphysiologic
Eye movement Electrode and lead
Electromyographic Instrumental
Glossokinetic Environmental
Electrocardiographic Internal electric stimulators
Sweat
Patient movement
239
240 Part II ■ Normal EEG
activity. Because the artifact is usually limited to a single elec-
trode, it will be seen only in channels containing that electrode
and will not show a “field” to neighboring electrodes.
Reformatting into various montages, particularly referential
montages, may help to identify a discharge as artifactual.
Because some electrode artifacts may be difficult to recognize
and may mimic abnormal patterns, any electrode showing high
impedance or frequent pops should be reapplied or replaced. In
some cases, high impedance or loose electrodes show severe
artifact with minor patient movements. These electrodes
should be reapplied as well. Confirming that electrodes are
placed and labeled correctly in the proper headbox jack should
be accomplished when expected waveforms appear in the
wrong channel.
Neurosurgical procedures and certain developmental dis-
orders disrupt normal cortical architecture and can yield pos-
itive discharges or ones with unusual fields (7) and thus
mimic electrode artifact. Skull asymmetries and edema
require documentation to properly reflect the differing ampli-
tudes seen on the recording. Breach rhythms that emerge fol-
lowing craniotomy may become “spikier” with
superimposition of myogenic potentials (Fig. 13.1). Rhythmic
or periodic activity on an EEG recording should be matched
with known rhythmic physiologic (i.e., cardiac, respirations)
or nonphysiologic activity (i.e., drips, pumps, machines) to
uncover the source. If the waveforms appear in more than one
nonadjacent brain region at once (nonanatomic pattern), the
activity is likely to be due to artifact. Additionally, if the wave-
forms appear in all leads and are unusual in appearance, the
Figure 13.1 Myogenic potentials (arrows) superimposed on a breach rhythm in the left parasagittal chain can
appear “spikier.”
reference electrode, ground, or preamplifier should be
checked for integrity.
The elimination of artifact from the EEG once it is recog-
nized involves identifying the source. Troubleshooting artifact
involves checking the electrodes (including the ground and
impedances), the apparatus, the patient (if still connected),
and the ambient activity. Artifact, most apparent in a loose or
dry electrode, is easily eliminated by regelling or replacing the
electrode during the study. Eliminating the artifact by chang-
ing a lead or turning off an external device can confirm the
source. The technologist should annotate the recording when
activity generated by the patient or the external environment
is interfering with the quality of the study. EEGs that run con-
tinuously over days such as epilepsy monitoring, ambulatory
EEG, or ICU monitoring studies may contain large amounts
of artifact. This may not allow proper artifact elimination as
the technologist is not present for the majority of the record-
ing. However, even when a technologist is on site, trou-
bleshooting is not always simple or successful. It may be
necessary to camouflage rather than eliminate artifact.
Artifacts can mimic nearly any abnormal pattern (8), so it is
critical that every attempt to discover the origin of the artifact
be undertaken.
ARTIFACTS OF PHYSIOLOGIC ORIGIN
Physiologic sources of artifacts, apparent on every EEG record-
ing, are generated from the biologic properties of the patient
and the many sources within the body that can act as electric

dipoles. The most common and important of these artifacts are
produced by the eyes, the muscles of the head and face, and the
heart. Other important physiologic generators of artifact
include the tongue or glossokinetic artifact, respiration, and
potentials generated in the skin by sweat glands. In addition, the
patient’s own movements, especially if rhythmic, may be a
source of confusion on an EEG. Placing an extra electrode on or
near a moving body part or utilizing video allows the reader to
confirm the nonbrain origin.
Eye Movement Artifact
Eye movement artifacts are seen in virtually every awake and
conscious individual during routine EEG. Most eyeball artifact
is easy to recognize by its anterior location and its bilateral,
synchronized appearance. When the eyes move, the electric
dipole between the positively charged cornea and the nega-
tively charged retina also moves, yielding a large potential
depending on the direction and rate of the movement. In gen-
eral, it is symmetric as long as lead placement is measured well,
the patient is not missing an eyeball (Fig. 13.2), both retinas
are functional, and the frontal bones are intact (6). Eye move-
ment monitors placed above and below the eye using one to
Figure 13.2 Lateral eye movements are noted asymmetrically only from the right due to an enucleated left eyeball
(prosthetic eye).
Chapter 13 ■ Artifacts of Recording 241
two channels help identify potentials that are in phase with
cerebral, and out of phase with extracerebral, ocular sources
(9). Eye movement artifact is important for the determination
of the normal waking background as eye opening blocks the
alpha rhythm and eye closure produces alpha. The artifact is
produced by the Bell’s phenomenon as the eyeballs roll upward
with the closure of the eyes, producing an upward movement
of the positive dipole created by the cornea. Eye movements
can be quite rapid with fluttering eyelids occurring at rates up
to 13 Hz and appear abnormal (Fig. 13.3) especially if they are
asymmetric (Fig. 13.4)! They can also be rather slow in the
delta range mimicking abnormal patterns such as frontal inter-
mittent rhythmic delta activity (FIRDA; Fig. 13.5). With hori-
zontal eye movements, the artifact is generated over the lateral
eye leads, F7 and F8 (Fig. 13.6). It can be quite striking as in the
case of lateral gaze nystagmus, generally noted more from the
side of the fast phase (10). The characteristic deflection that
includes a rapid rise and more gradual fall with the corrective
movement is quite easy to recognize. The side of the positive
phase reversal of the rapid rise indicates the direction of the
fast component of the nystagmus. With leftward eye move-
ment, the F7 electrode has positive reversals at the same time
242 Part II ■ Normal EEG

Figure 13.3 Eye blinks can be so rapid that they appear abnormal. Note the bifrontal symmetric 9-Hz activity that
represents eye fluttering, elicited with eye closure (arrows).

Figure 13.4 Unilateral eye fluttering is noted in a patient with a detached retina on the right.
 Chapter 13 ■ Artifacts of Recording 243

Figure 13.5 Slower appearing eye blinks (2.5 to 3 Hz) intermixed with EMG appear similar to 3-Hz spike-and-wave
activity. Eye leads show out-of-phase activity (arrow), confirming origin from eye movements and distinguishing
from epileptiform activity and frontal intermittent rhythmic delta activity.

Figure 13.6 Rapid lateral eye movements demonstrated on an awake patient (gray bar). Note the eye leads LUC
and RLC, showing the out-of-phase activity. The arrow points to a lateral rectus spike.
244 Part II ■ Normal EEG
Figure 13.7 Left-beating nystagmus is identified by the F7 electrode (longitudinal bipolar montage, top two chan-
nels of the third electrode chain) demonstrating positive reversals at the same time that the F8 electrode reveals neg-
ative reversals (gray bar). Most of the eye movements are preceded by lateral rectus spikes.
that the F8 electrode reveals negative reversals (Fig. 13.7).
Additionally, eye movements are crucial to correctly identify
the stages of sleep such as the slow roving subdelta in the lat-
eral eye leads representing drowsy eyeballs during stage Ia and
REM artifact also seen in F7 and F8 lateral eye leads during
REM sleep (Fig. 13.8).
Electroretinogram (ERG)
During photic stimulation, light from the flash stimulus may
produce artifact in the frontopolar leads (Fp1/Fp2). This arti-
fact can be mistaken for photic driving due to the entrainment
with the stimuli, but is not located occipitally. The source may
be the retina (ERG) or from a nonphysiologic source such as a
frontopolar electrode with high impedance creating a photo-
cell. Covering the eye with a towel will block the input to the
retina (ERG) and this should not be confused with the photo-
electric effect (11).
Muscle Artifact
Muscle is another commonly observed artifact in the EEG. It is
high-frequency activity noted to be very “spiky” but too fast to
be an epileptic discharge. It is observed more while the patient
is awake, and may obscure critical portions of the recording,
for example, during hypermotor seizures. The frontalis and
temporalis muscles are principal sites of myogenic artifact.
Frontalis muscle becomes most involved with forced eye clo-
sure, and photic stimulation (10). Temporalis muscles become
active with jaw clenching, chewing, or bruxism appearing as
bursts of fast activity (Fig. 13.9). Technologists can ask the
patient to open the mouth that relaxes the jaw and diminishes
this artifact. Frontalis muscle contraction during periocular
movement may elicit sustained or individual motor unit
potentials that can appear like “railroad tracks” (Fig. 13.10).
Lateral rectus muscle spikes representing motor unit potentials
recorded from the lateral orbit may simulate small interictal
discharges (i.e., F7/F8) (Fig. 13.11). During intermittent photic
stimulation, eye flutter artifact at frequencies of 6 Hz may
mimic generalized spike-and-wave when coupled with myo-
genic potentials and create a “pseudo-photoparoxysmal
response” (Fig. 13.12) when muscle spikes are time-locked to
the flash frequency.
Glossokinetic Artifact
The tongue is a very large muscle centrally located inside the
head. Similar to the eye, the tongue creates a bioelectric dipole
with the tip of the tongue negative relative to the root (11).
During oropharyngeal motion, a direct current (DC) potential
is produced that is often diffusely seen with frontal and tempo-
ral predominance (11). Cephalad–caudad motions may be
produced by tongue movements involuntarily while speaking
or when swallowing (Fig. 13.13) producing delta frequency
waves that may mimic FIRDA distinguishable by the
superimposed muscle artifact. Similar artifact may be
produced when patients are asked to say “lilt” or “ta ta ta,”
identifying similar patterns as artifact more conclusively
(Fig. 13.14). Horizontal tongue movements (i.e., tongue in
 Chapter 13 ■ Artifacts of Recording 245

Figure 13.8 This longitudinal bipolar montage demonstrates slow roving eye movements in the lateral eye leads
(F7 and F8) signifying drowsiness. Eye monitors are LUC (left upper canthus) and RLC (right lower canthus). The
alpha rhythm attenuates (vertical arrow and line), and slow lateral eye movements appear. There is also EKG arti-
fact (arrowheads) and lateral rectus spikes (horizontal arrow).

Figure 13.9 Bursts of myogenic artifact in an alternating pattern are consistent with bruxism.
246 Part II ■ Normal EEG

Figure 13.10 The left frontopolar electrode (Fp1) (arrows) demonstrates a motor unit firing that appears like “rail-
road tracks.”

Figure 13.11 Lateral rectus spikes (arrows) are noted bilaterally in the frontopolar leads. Lateral eye movements
show out-of-phase deflections at F7 and F8 (gray bar).
 Chapter 13 ■ Artifacts of Recording 247

Figure 13.12 Photomyogenic artifact is noted in the frontal (not occipital) eye leads during intermittent photic
stimulation (stimulus artifact is not displayed). This should not be mistaken for photoparoxysmal response.

Figure 13.13 Diffuse anterior predominant delta during waking state (eye blinks, muscle) corresponds to drinking
(glossokinetic artifact).
248 Part II ■ Normal EEG
Figure 13.14 Bifrontal 2- to 4-Hz square-shaped waveforms representing lingual (glossokinetic) artifact. Note that
these are in phase in eye channels, so cannot represent eye movements.
cheek) can create unilateral artifact on the EEG that may be
confusing. Validation is possible through application of tongue
movement monitors with electrodes placed above and below
the mouth over the cheek and chin (10). Using a bipolar mon-
tage, positive phase reversals are evident with tongue move-
ment in a similar fashion to those recorded during eye
movements.
Electrocardiogram (EKG) Artifact
While many artifacts “contaminate” the EEG, some (i.e., the EKG)
are essential for interpreting physiologic functions that may occur
during the recording session. The heart is another key source of
EEG artifact and is variably present on recordings depending on
the montage used and the size of the patient. Referential montages
often accentuate EKG artifact (Fig. 13.15), especially using ipsilat-
eral ear reference with the larger intra-electrode distances. Linked
ears montage can reduce the artifact (Fig. 13.16). Overweight
patients with short stocky necks as well as babies may be predis-
posed to EKG artifact because the dipole is situated closer to the
recording electrodes and better able to transmit the current (10).
This may be reduced by altering head position. Recording the
EKG during routine EEG is essential to enable recognition of the
cardiac–cerebral relationship. Channel I of the EKG is approxi-
mated with electrodes linking the left and right chest using bipo-
lar recording to identify the QRS. EKG interference may be noted
as a small spike at 1/sec that in isolation mimics an interictal spike
or in succession and on a low-amplitude record mimics periodic
lateralized or generalized periodic epileptiform discharges
(PLEDs or GPEDs). Bipolar montages may reveal apparent inter-
ictal epileptiform discharges (IEDs) that are “focal” diphasic
waveforms predominately in the temporal derivations of the left
hemisphere due to the vector created by the electric conduction of
the heart. Opposite polarities of the R-wave from the EKG may be
seen at the left (negative) and right (positive) ear electrodes. EKG
can be readily identifiable as the source because of the regularity
and time-locked waves synchronous with the EKG. Importantly,
when an electrode is positioned over an artery, the blood pulsing
through the vessel moves it regularly to produce a periodic and
smooth waveform time-locked to the cardiac rhythm appearing
approximately 200 msec after the QRS complex (Fig. 13.17)
(10–12). Pulse artifact can be eliminated by moving the electrode
off the artery.
Sweat
Scalp perspiration will also produce artifact by creating
unwanted electric connections between electrodes. Perspiration
artifact appears as very low–frequency ( 0.5 Hz) low-amplitude
undulating waves. Changes in the DC electrode potential from
perspiration may result in an unstable baseline (sweat sway)
and crossing of tracings in adjacent channels (Fig. 13.18).
Perspiration artifacts are usually seen in multiple adjacent
channels or over the entire scalp and poses a risk for bridging
between electrodes (see also the section “Electrode and Input
Lead Artifacts”). They can be reduced somewhat by increasing
the low-frequency filter, and by drying the scalp and reapplying
electrodes, or by cooling the patient with a fan.
Patient Movements
Patient movements cause the leads or electrodes to move, and
thus can provide a large source of “physiologic” artifact on

Figure 13.15 Bipolar recording shows prominent EKG artifact time-locked to the QRS complex, maximal in ampli-
tude over both temporal–occipital regions (arrows).
recordings. This is especially true for the awake and ambulatory
patient (13), but is also quite notable in those who are agitated
and confused. Synchronous video has been inordinately helpful
in distinguishing the varieties of patient movements when a
technologist is not present for the recording. Regular appearing
patient movements such as tremor may manifest at different
frequencies and amplitudes such as the 6-Hz tremor of
Parkinson disease (Fig. 13.19) or scratching near an electrode.
During bipolar recording, extra adjacent electrodes placed on
the suspected limb are helpful to confirm tremor artifact but
they may be readily apparent in the EKG lead. Sudden or rapid
patient movements such as shivering (Fig. 13.20) can mimic the
paroxysmal quality of an electrographic seizure, but usually do
not show progression. Additional artifact that leads to regular
movement of the head or body from large pulses through the
aorta or ballistocardiographic artifact gives rise to delayed
pulses that can appear more generalized but still linked to the
EKG (14). During electrocerebral inactivity recordings at sensi-
tivities of 2 V/mm, ballistocardiographic artifact and pulse
artifact may be more prominent with small scalp arteries caus-
ing repetitive movements of electrodes. Stabilizing the head
with towels under the neck often eliminates this artifact.
Artifact due to respiration is also generated by the movement of
the body or the head (Fig. 13.21). This can be seen during
hyperventilation over posterior head regions. A technologist
can observe the breathing and annotate the recording to indi-
cate where inhalation and exhalation occur (11). In the epilepsy
monitoring unit (EMU), the technologist is away from the
patient but continuous video and the ability to transmit and
Chapter 13 ■ Artifacts of Recording 249
display the time-locked behavior with EEG usually leads to
accurate interpretation. Technical guidelines for recording in
the EMU have been developed by the American Clinical
Neurophysiology Society (15). For further discussion, see
Chapter 37. Long-term monitoring provides a greater likeli-
hood of deterioration of the patient–electrode interface with
drying out of the electrolytic gel or nonadherence of the elec-
trodes to promote artifact that may override interpretation dur-
ing behavioral events. In the EMU, approximately 20% to 25%
of patients with episodes will be manifesting psychogenic
nonepileptic attacks (PNEA) (16). These are described in more
detail in Chapter 33. The importance of recognizing behavioral
sources in addition to historic information is crucial to recog-
nizing nonepileptic events especially if there is “pseudo-evolu-
tion” simulating an ictal recording (Fig. 13.22) (4).
ARTIFACT OF NONPHYSIOLOGIC ORIGIN
Electrode and Input Lead Artifacts
Electrode artifacts are one of the most common types of
nonphysiologic artifact. Any unusual waveform that is
restricted to a single electrode should be considered to be arti-
factual (17,18). The scalp–electrode interface and its impor-
tance in accurate recording of cerebral electric activity are
described in Chapters 5, 6, and 8. Disruptions of this interface
result in several types of artifact. Poor electrode contact or
“loose” electrodes are a frequent cause of artifact. The charac-
teristic electrode “pop” occurs when an electrode moves and the
250 Part II ■ Normal EEG
Figure 13.16 A: EKG artifact is accentuated when using an ipsilateral ear reference. B: EKG artifact is substantially
diminished by linking the ears.
electric double layer is disturbed, creating a DC potential simi-
lar to discharging a capacitor (19). These electrode movements
may be minute and not well seen by visual observation. Pops
are sudden positive (or less commonly negative) discharges
and usually show an initial high-voltage very steep deflection
followed by an exponential decay caused by the amplifier’s low-
frequency filter. The shape of the “pop” often resembles that of
the calibration waveform, with “mirror images” of the discharge
seen in bipolar channels with the electrode in common
(Fig. 13.23). Pops may be single or repetitive. When an elec-
trode shows repeated pops, the impedance should be measured
and the electrode reapplied. Recording from an additional elec-
trode placed close to the electrode with suspected artifact may
also help to determine whether the activity is artifactual.
Similar pops can be seen with drying of electrode gel or paste,
poor connection of the electrode to the electrode wire, broken
lead wires, or faulty connection between electrode pin and the
pin jack. If reapplication of the electrode does not fix the arti-
fact, the electrode should be examined for corrosion at these
contact points.
In some cases, poor or fluctuating contact between elec-
trode and scalp results in continuous low-amplitude rhythmic
or semirhythmic slow activity, simulating focal slowing or
seizures (Fig. 13.24) (6). Again, this can be distinguished from
pathologic brain activity by restriction to a single electrode
contact. Similar artifacts can be caused by drying of electrolyte
 Chapter 13 ■ Artifacts of Recording 251

Figure 13.17 Pulse artifact is noted at O1 time-locked to the EKG but occurring 250 msec after the QRS.

Figure 13.18 Subdelta activity is noted with crossing baselines consistent with “sweat sway.”
252 Part II ■ Normal EEG

Figure 13.19 Tremor artifact (6 Hz) picked up over the left temporal chain (and EKG lead) due to motor unit fir-
ing and rhythmic movement artifact.

Figure 13.20 Shiver and body jerks causing occipital “pops” appear as part of a nonepileptic psychogenic event.
 Chapter 13 ■ Artifacts of Recording 253

Figure 13.21 Respiration artifact occurring every 3 seconds, subtle on the left side of the page, and enhanced on
the right side by increasing the gain. Artifact is caused by rocking head movements on posterior electrodes.

Figure 13.22 Patient using a pumice stone to sand the calluses of her left foot. Subtle back and forth motion of
the head/ body renders this artifact nearly indistinguishable from seizure without the help of video.
254 Part II ■ Normal EEG

Figure 13.23 Electrode pop is noted over the P4 electrode, most prominent in the area of skull breach, which
evolves into faster, almost ictal-appearing discharges.

Figure 13.24 Electrode pop imitating a very localized seizure. The Fp1 electrode is not well placed or has air-dried
paste. Rhythmic 5-Hz activity remains quite focal but then slows to 4 Hz almost “evolving.”

between electrode and scalp or by movement of the electrode
leads. Changing impedances from movement of the electrode
can cause rhythmic artifact (see also the section “Artifacts of
Physiologic Origin”: respiration, hyperventilation, tremor,
pulse artifact). Since EEG instruments use differential ampli-
fiers, each amplifier channel receives input from two elec-
trodes, and the impedances must match reasonably well.
Guidelines for minimum standards suggest that individual
electrodes must have impedances of less than 5000 (5).
Mismatched impedance between electrodes will cause the EEG
amplifiers to work improperly, causing distortion of the EEG
signal in channels containing those electrodes (20). High
impedance electrodes can cause falsely low or high amplitudes
and loss of low-frequency components of EEG. In addition,
channels with mismatched electrode impedances are more sus-
ceptible to environmental artifacts, particularly 60-Hz artifact
(21). Digital EEG machines utilize a single machine or refer-
ence electrode, and all other montages are reformatted based
on this referential recording. The impedance of the machine
reference electrode is therefore of critical importance, and
problems can result in technically limited studies that interfere
with interpretation.
Ground lead recordings occur when a very high imped-
ance or disconnected electrode exceeds the input impedance
of the recording system (22). In this case, the ground elec-
trode is substituted for the poor contact electrode, sometimes
leading to bizarre waveforms (Fig. 13.25). Salt bridges occur
when electrode paste or gel is smeared between two electrode
locations. EEG activity recorded from each electrode there-
Figure 13.25 Ground lead recording. Very high impedance electrode at O1 causes ground electrode (at Fpz) to
become input 2 of channels including O1. Therefore, T5-O1 effectively becomes T5-Fpz, resulting in “upside down”
eye blinks (arrows) in this channel.
Chapter 13 ■ Artifacts of Recording 255
fore becomes the sum of the activity from the two electrodes,
leading to spuriously low-amplitude recordings. Sometimes
activity in the channel containing the salt bridge can appear
flat or even isoelectric. This type of artifact is more common
when using closely spaced electrodes, as in the International
10-10 modified system or when using the full 10-20 place-
ment in neonates.
Bimetal potentials can occur when dissimilar metals make
contact, either at the electrode–wire interface or between the
electrode pin and jack box receptacle. Dental fillings with dis-
similar metals may also cause spiky artifacts when the metals
touch during chewing and speaking.
Finally, high impedance electrodes can cause photovoltaic or
photoelectric artifact during intermittent photic stimulation.
Each flash causes a photochemical reaction in the electrode.
This is seen as brief spike-like transients in the high impedance
electrodes simultaneously with the flash (Fig. 13.26).
Photoelectric artifact can be confirmed by its ability to be
blocked by covering the involved electrode.
Proper electrode application and maintenance is the best
way to reduce electrode artifact. Electrodes should be cleaned
well and inspected regularly to look for signs of corrosion
between the electrode and conducting wire, damaged insula-
tion, or broken lead wires. Faulty electrodes should be dis-
carded. If silver chloride electrodes are used, they should be
rechlorided on a regular schedule. Good electrode contact
should be ensured by adequate preparation of the skin to
remove surface oils. Electrode impedance should be measured
and verified to be less than 5000 (5). Extra electrode paste
256 Part II ■ Normal EEG
Figure 13.26 The T4 electrode acts as a “photocell” detecting the 9-Hz photic stimulus artifact. There is also
increased 60-Hz artifact in this high impedance channel when the photic stimulator is active.
between electrodes should be removed. Covering the electrode
with collodion or EC2 paste prevents small movements of the
electrode and slows drying of the electrode paste. Electrode
wires should be bunched together to prevent movement and
excessive tension on electrodes. Occasionally, the leads are
placed into the incorrect jack inside the headbox and can yield
unusual fields (Fig. 13.27).
Instrumental Artifacts
Malfunction of almost any part of the EEG machine can
cause artifact. Improved design and construction of EEG
machines, as well as the nearly universal use of digital tech-
nology, has reduced the likelihood of equipment malfunc-
tion, but this still occurs frequently enough that
electroencephalographers (EEGer) and EEG technologists
must have a thorough understanding of the most common
technical artifacts.
Modern amplifiers are small, low-power, single-chip multi-
channel devices with solid-state integrated circuits. Amplifier
“noise” is caused by thermal agitation of electrons in the
amplifier circuits. Noise contains components at all frequen-
cies and therefore increases with the bandwidth of the ampli-
fier. Noise levels should be less than 2 V measured peak to
peak (5). This noise is typically only problematic when
recording at sensitivities of 1 to 2 V/mm, such as used for
electrocerebral inactivity recordings (22). A malfunctioning
amplifier component (capacitor, resistor, transistor) can pro-
duce larger amplitude noise, which is added to and distorts
cerebral activity.
In analog EEG machines, each amplifier had multiple
switchable components that could malfunction, causing
excessive noise, abnormal transients, or loss of signal (22).
Electrode selection switches were used to determine which
two electrodes were input to each amplifier. Switch malfunc-
tions and incorrect electrode selection were common sources
of artifact. Similar errors could occur in sensitivity, low-
frequency filter, and high-frequency filter selectors. In digital
EEG machines, amplifiers are hardwired, with each channel
made up of a hardwired input 1 and machine reference as
input 2, reducing the likelihood of switch problems. EEG is
acquired with filters wide open and digital filters are applied
without altering the original signal. Software errors in sensi-
tivity, filters, or montages, however, can still lead to spurious
displays.
Failure of components common to all channels often causes
lack of power or 60-Hz artifact. Excessive 60-Hz artifact can be
caused by a faulty or absent ground connection to the patient,
or defects in the power supply or other parts of the instrument
(Fig. 13.28A). Portable equipment is more susceptible to
mechanical wear and tear, such as vibration or jolting of
machines. Loose wiring or loosening of circuit board connec-
tions is a common source of artifact, causing loss of signal or
intermittent failures.
Digital artifacts include aliasing (inadequate sampling rate),
errors in analog to digital conversion, skew errors, multiplexing
artifacts, and blocking (23). These are discussed in detail in
Chapter 7. Aliasing is caused by sampling at a rate less
than twice the highest frequency present in the signal, and can

Figure 13.27 A: Classic left temporal discharges appear as bizarre bilateral sharp waves with multiple phase rever-
sals and temporal regions show bizarre alpha and theta activity. This is due to lead T3 connected into T4 jack in the
jack box (and vice versa). B: After correction of lead error.
introduce spurious slower frequencies into the EEG.
Multiplexing artifacts occur when the samples from multiple
amplifiers are switched. Blocking can occur when the amplifier
is saturated, producing a squared-off waveform at the limits of
amplifier resolution.
Finally, incorrect equipment display can cause artifacts
(11). In analog machines, damage to or misalignment of pen
recorders can result in apparent asynchrony of simultaneous
discharges, while displacement from zero baseline leads to dis-
torted morphology of waveforms. Incorrect pressure at the pen
tip can also distort waveforms: if pressure of the pen on paper
is too great (overdamping), the pen drags and acts as a high-
frequency filter, while underdamping results in pen overshoot.
Malfunction of paper drive mechanisms can lead to variations
in chart speed, causing waveforms to appear longer or shorter
than actual brain activity. In digital displays, using monitors
with inadequate display resolution can cause aliasing, intro-
ducing spurious slow frequencies into the display that resolve
when a screen resolution is adequate to display all recorded
samples (24).
Environmental Artifacts
A wide variety of artifacts arise from objects and electric devices
around the patient. The vast array of morphologies and origins
can make identification of these artifacts difficult. Typically,
artifactual waveforms from external devices show very different
morphologies from normal cerebral activity, and may often be
identified as artifactual even if the exact source cannot be
found. Simultaneous video recording may also help to identify
Chapter 13 ■ Artifacts of Recording 257
some environmental artifacts, such as movement of people and
objects around the patient (25).
Mains Power Supply (Alternating Current Artifact)
The most common source of electric interference is the alter-
nating mains power supply, at 60 Hz in the United States and
50 Hz in Europe. Electric interference from mains power sup-
plies can be either electrostatic or electromagnetic. In a prop-
erly designed modern neurophysiology lab, power lines are
adequately grounded and severe electric interference is rarely
problematic. If necessary, interference can be reduced by shield-
ing power cables (with outer metal braid connected to ground),
by using a shielded room for recordings, or by moving the
patient farther from the source of interference in the room (22).
Environmental artifacts are more common in areas outside the
EEG lab (“hostile environments”), particularly intensive care
units and operating rooms, where other medical equipment can
interfere with EEG recordings.
Sixty-hertz alternating current can cause interference via
electrostatic effects. Power lines will couple to other nearby
conductors (including the patient, technologist, electrode
wires, metal bed frames, etc.) and induce a small alternating
potential of opposite polarity to the mains potential (22). While
small, this potential can be very large in comparison to the volt-
age of EEG activity. Connecting the patient to ground (earth)
will reduce this potential. Artifact can also be reduced keeping
all electrode wires and cables short, bundling electrode wires as
a group, and unplugging unnecessary equipment. A notch filter
can be used to remove excessive 60-Hz artifact; however, it will
258 Part II ■ Normal EEG

Figure 13.28 A: Sixty-hertz artifact (electrostatic artifact) obscuring most of the record because of the lack of a ground
electrode. B: More typical appearance of 60-Hz artifact in a high impedance O1 electrode. C: Same segment of EEG as
B with 60-Hz notch filter on.

remove signals of interest around 60 Hz and so should not be
used routinely (Fig. 13.28B and C). Other electric equipment
near the patient can cause electromagnetic interference from
current flowing through cables or from transformers and
motors. The intensity of the magnetic flux is proportional to
current. Common hospital and EEG lab sources of interference
include photic stimulators, portable air pumps, monitoring
devices, lights and lamps, fans, air conditioning, pneumatic
boots, heating and cooling blankets, electric beds, intravenous
infusion pumps, feeding delivery systems, elevators, computers,

Figure 13.29 Twenty-hertz artifact (a harmonic of 60 Hz) caused by a cooling blanket obscures nearly all of
the EEG.
telephones, diathermy, and x-ray equipment. Several of these
are discussed in more detail below. The electric field generated
by the other equipment is inductively coupled to the electrode
wires, usually producing an out-of-phase harmonic of 60-Hz
activity (Fig. 13.29). The harmonics will not be eliminated by
the notch filter. Electromagnetic artifact can be reduced by (i)
ensuring that all equipment has a shielded metal case connected
to ground, (ii) plugging all equipment into the same connec-
tion point, and (iii) ensuring that power cables do not run in
parallel and are remote from electrode wires and cables.
Sixty-hertz artifact is present in small amounts in all EEG
recordings, but is usually of low amplitude because amplifier
common mode rejection removes most of the artifact. When
there are high or mismatched impedance electrodes, the ability
of the amplifier to cancel out noise is reduced, and 60-Hz arti-
fact will appear in channels including the high impedance elec-
trodes (26,27). This artifact is often a handy “detector” for high
impedance electrodes during EEG recordings. These electrodes
should be reapplied as soon as possible. If 60-Hz artifact is seen
in all channels, this usually indicates a problem with the
ground electrode or the ground of the EEG machine, and
should be investigated promptly to ensure there is no electric
safety hazard (6).
Electrostatic Artifacts
Electrostatic forces may induce currents in electrode wires or
cables. Movement of the electrode wires themselves will cause
focal or widespread slow waves at the frequency of the
moving wires. Using short electrode wires, bundling the wires
in plastic tubing, or wrapping them with gauze will decrease
this type of artifact. Movement of the input cable between
electrode jack box and the amplifier can also cause capacitive
Chapter 13 ■ Artifacts of Recording 259
coupling or transfer of energy between adjacent cables. This is
usually readily identified as artifact because of widespread
distribution and unusual electric fields. Rhythmic artifacts
can be induced by movement of electrode wires during pat-
ting (burping or soothing infants) or chest percussion therapy
(Fig. 13.30) (28). Similar rhythmic or repetitive artifact may
be produced by electrode wire or patient movement from ven-
tilators, percussive anti-bedsore beds (Fig. 13.31), intra-
arterial balloon pumps, and oscillatory ventilators. The shape
of the artifact depends on the frequency of activation of the
device, and is usually a slow wave or a mixture of frequencies
superimposed on a slow wave, and is typically stereotyped or
monomorphic. Percussive beds or oscillatory ventilators cause
artifact in the theta or alpha frequency range, usually best seen
in posterior head electrodes or electrodes in contact with the
bed. An extra lead can be connected to the ventilator tubing to
reveal the source.
Movement of people or objects around the patient may
induce electrostatic artifacts, even if the wires themselves do not
move. Examples can be seen with movements of nurses and EEG
technologists around the bed, movement of clothing or bedding
(especially synthetic fabrics), swaying of tubing near the patient,
or movement of water in tubes near the electrode wires or input
cable (22). These waveforms typically are high voltage, with
complex morphologies and nonphysiologic fields. A common
source of electrostatic artifact is from movement of water in res-
piratory tubing in intubated patients. Water movement results
in high-voltage slow waves with superimposed fast components
(Fig. 13.32), and usually occurs with a widespread field over the
frontal regions or the side of the head nearest to the respiratory
tubing (29). Similarly, electrostatic charges on drops of fluid in
intravenous drips may cause spike-like potentials coinciding
260 Part II ■ Normal EEG

Figure 13.30 Rhythmic appearing waves are noted over the right side from patting the chest while the right side
of the head is on the bed.

with the drops of the infusion (Fig. 13.33) (30). Piezoelectricity
is an electric charge generated by the mechanical deformation of
polymeric materials. Artifact from electric infusers may be
caused by this static electricity generation between polyvinyl
chloride tubing and the pump roller head (31). Unusual artifacts
may also be produced by continuous veno-venous hemofiltra-
tion and dialysis machines (32,33). These artifacts arise from a
combination of static and piezoelectric currents in pumps rotat-
ing between 50 and 600 rpm, causing currents to flow into the
patient (and then to electrodes) via fluid in the tubing.

Figure 13.31 Rhythmic artifact generated from an anti-bedsore percussion bed.

 Chapter 13 ■ Artifacts of Recording 261

Figure 13.32 Ventilator artifact noted periodically with superimposed faster waves from water vibrating inside the
tubing.

Figure 13.33 Electrostatic charges generate spike-like artifact from intravenous drips (arrows).
262 Part II ■ Normal EEG
Figure 13.34 Rhythmic activity
over the right temporal region looks
like a partial seizure until the dialy-
sis machine is turned off demon-
strating machine artifact.
The technologist should try to minimize movements
around the patient and note any changes in the EEG caused by
such movements. In addition, the technologist should include
notations about equipment connected to the patient, with an
attempt to correlate unusual EEG transients with equipment
function (e.g., respirator function or IV drops). Briefly turn-
ing off a machine during the EEG can confirm the origin of
the rhythmic artifact such as from a dialysis machine
(Fig. 13.34). Electrodes or transducers near the presumed site
of artifact generation may aid in identification of artifact
sources.
Electromagnetic Artifacts
High-frequency interference from nearby television and radio
stations is rare but difficult to eliminate. Artifact can be seen in
all channels. Hospital paging systems and walkie-talkies using
radio frequency carriers may also cause artifact. Noncellular
telephones may cause ringing artifacts, related to the voltages
that develop in the telephone lines as the telephone rings
(Fig. 13.35) (34). Portable and cell phones can produce artifact
even when not being used (35). The effect of cell phones
depends on distance, operating frequency, and communication
technology. Cell phones may cause repetitive sharp and slow-
wave complexes with a nonphysiologic distribution. These can
be widespread but may be of higher voltage in electrodes near
the cell phone. Many of these are eliminated by the EEG
machine’s antialiasing high-frequency filter but may cause
bizarre intermittent artifact. Microwave ovens near the EEG
lab may produce radio frequency interference. Televisions and
computer screens can cause high-frequency interference at the
screen refresh rate when they are in close proximity to
the patient. In the operating room, the electrocautery devices
generate high-voltage high-frequency signals that travel
through the body to the recording electrodes. The EEG
machine should be turned off during use of coagulation.
However, high-frequency cautery in adjacent rooms may also
produce interference (36).
Any equipment with a motor may produce artifact due to
the switching magnetic field within the motor. This activity can
be constant or intermittent. Infusion motor artifact is generated
by automatic programmable intravenous infusion pumps (30).
This typically is characterized by very brief spiky transients,
which are stereotyped and repeat at regular intervals (37). Slow
components may also be present.
Line isolation monitors, commonly present in the intensive
care unit and operating room, scan electric outlets for leakage
currents. They generate low-voltage artifact at a particular fre-
quency, which may be noticeable at increased sensitivity record-
ings in comatose or anesthetized patients (38).
Electric Stimulators
Pacemakers produce very spiky, usually high-voltage artifact
(Fig. 13.36). These spikes are best seen in the EKG channel, but
may occasionally show a field extending to EEG electrodes.

Figure 13.35 Two-second burst of 30 cycle/ sec telephone artifact (electromagnetic artifact, vertical arrow) dif-
fusely, but highest amplitude in high impedance electrode Fp2. This leads to eye movements and EMG as the phone
is answered, obscuring a second “ring artifact.”
High-frequency artifacts can occur with deep brain stimulators,
the vagus nerve stimulator, spinal cord stimulators, and abdom-
inal electrostimulators. Again, these are usually best seen in the
EKG channel (38). The vagus nerve stimulator produces a scalp
negative evoked potential with very high amplitude
(Fig. 13.37). This potential is generated in the neck, and can be
recorded with electrodes over the stimulating electrodes in
the neck and sometimes in the EKG (39).
ARTIFACT DETECTION AND REJECTION
Artifacts are usually easily recognized by experienced EEGer. The
processes of visual analysis, remontaging, and digital filtering
allow identification of most physiologic and nonphysiologic arti-
facts. Reviewing an epoch of EEG in a different montage may
allow the EEGer to determine that a particular waveform does not
have a physiologic field, and thus be more certain of its artifactual
nature. Digital filters can be applied and removed multiple times,
and can significantly improve interpretation of EEG contami-
nated by artifacts by allowing specific frequencies to be removed
from the digital display. These filters may distort normal or abnor-
mal EEG activity, such as epileptiform discharges, so the ability to
remove the filters and view the “raw” EEG is essential.
Processing of EEG data to remove artifacts before digital
analysis (e.g., automatic spike or seizure detection, compressed
spectral analysis) is more difficult. Unrecognized artifacts can
Chapter 13 ■ Artifacts of Recording 263
significantly distort the processed data (40). Therefore, the raw
EEG should be available for comparison whenever quantitative
results are analyzed, and careful attention should be paid to
quality of data acquisition (41). Several commonly used meth-
ods to remove artifacts prior to digital processing are described
in the following sections.
Use of Band Pass Filters
If the analysis is restricted to certain frequency bands, an auto-
mated algorithm can be designed to only analyze activity in this
frequency band. For example, a 1- to 20-Hz band pass may be
used to remove muscle artifact (42). This method is not very
useful for analysis of the entire bandwidth of EEG, as artifacts
can occur at any frequency. Even for very narrow frequency
bands, there may be significant artifact remaining after band
pass filtering. The process of filtering may significantly alter the
appearance of EEG and make subsequent identification of arti-
facts more difficult.
Manual Rejection of Artifact Segments
In this case, a technologist or EEGer visually reviews the entire
EEG recording and marks segments with artifacts. This is a
reliable method, and may detect some artifact that would be
missed by automated techniques. It is time-consuming, how-
ever, and reader fatigue may become problematic for long or
multichannel recordings. Subtle or brief artifacts may not be
264 Part II ■ Normal EEG

Figure 13.36 Pacemaker artifact best noted in the EKG lead but seen diffusely on EEG leads.

Figure 13.37 VNS artifact.


identified, and different readers may have different thresholds
for rejection. This method is only possible for offline (not
real-time) digital analysis. Once the artifact segments are
identified, they are completely eliminated from the analysis,
including any intervening normal EEG that may have been of
interest.
Automatic Rejection of Artifact Segments
This technique rejects short segments of EEG if the segment
exceeds predefined thresholds. These thresholds can be simple
analysis of the EEG channels themselves such as amplitude,
numbers of zero crossings, or 60-Hz artifact. If a segment
shows very high amplitude, it is eliminated. Some techniques
use other special electrodes to identify artifact signals, such as
electrooculogram, EMG, EKG, or accelerometers (19). If the
signal in these channels exceeds a threshold, the segment of
EEG will be rejected. This technique can be used online or
offline. These automatic rejection techniques will not identify
all possible artifacts, especially for ambulatory patients or in
electrically hostile environments. Again, the entire EEG seg-
ment is rejected if the threshold is exceeded, so some useful
EEG may be eliminated.
Automated Subtraction of Artifact
These methods aim to identify artifact and to remove only the
artifact from the recording, leaving a “clean” EEG for digital
analysis (19). Simultaneously recorded EOG or EKG can
be subtracted from an EEG channel using a computer. The
algorithm must be adapted for each individual patient. These
techniques may result in distortion of the EEG signal, as
the EOG also contains some normal brain signals that will
be subtracted.
Source decomposition techniques aim to decompose EEG
signals into individual components that represent EEG and
others that represent artifact. Once the artifact component is
identified, it is removed, and the remaining signal is recom-
posed. Examples of these methods include spatial filtering (43),
principal component analysis, and independent component
analysis (44). Wavelet-based and neural network algorithms
(45) allow adaptation of this method to a wider variety or arti-
fact types. These techniques are discussed in full in Chapters 56
and 58.
CONCLUDING STATEMENTS
Numerous artifacts are encountered during routine or special
EEG recordings. While some artifacts are critical for interpreta-
tion of the EEG (i.e., sleep staging), many are not useful and
serve to “contaminate” the recording and interfere with the
interpretation. Technological advances for artifact identifica-
tion and rejection have improved the ability to eliminate the
vast artifact present in longer term monitoring of EEG such as
in the ICU and EMU (46). These special care neurology units
may house patients with skull asymmetries from craniotomies
and scalp edema requiring documentation to properly reflect
markedly asymmetric amplitudes on the recording (Fig. 13.14).
Chapter 13 ■ Artifacts of Recording 265
Patient manipulation through suctioning, dressing changes,
and movement during routine nursing activities or ICU proce-
dures (i.e., central lines, cardiopulmonary rescucitation (CPR))
produces copious artifacts and it can be helpful to train nursing
staff to annotate the record for events that occur at the patient’s
bedside. However, the EEG technologist and EEGer are both
crucial to identifying and recognizing sources as well as serving
to “troubleshoot” and eliminate artifact that may lead to limited
or false interpretations. The American Board of Technologists
in EEG and evoked potentials has established baseline stan-
dards for technological recording of EEG, and the American
Board of Clinical Neurophysiology holds similar standards for
EEG interpretation to help guide proper interpretation of the
EEG in the presence of artifact. Technologists and EEGer
remain challenged to recognize the source of each and every
artifact that might jeopardize interpretation of the recording
and therefore conservative interpretation is warranted when-
ever doubt exists. Furthermore, even when recognition and
identification of the source of artifact is accomplished, elimina-
tion is not always possible. Digital EEG has ushered in a new era
for clinical neurophysiologists. Yet, principles of noncerebral
waveforms, the electric properties, electric fields, and recording
techniques still require an orderly approach for the successful
visual analysis of artifact on the EEG. The systematic approach
of recognition, source identification, and elimination of artifact
is an important process to reduce the chance for misinterpreta-
tion of the EEG and limit the potential for adverse clinical con-
sequences.
REFERENCES
1. Aurlien H, Gjerde IO, Aarseth JH, et al. EEG background activity
described by a large computerized database. Clin Neurophysiol.
2004;115(3):665–673.
2. Krauss GL, Abdallah A, Lesser R, et al. Clinical and EEG features of
patients with EEG wicket rhythms misdiagnosed with epilepsy.
Neurology. 2005;64(11):1879–1883.
3. Benbadis SR, Tatum WO. Overinterpretation of EEGs and misdi-
agnosis of epilepsy. J Clin Neurophysiol. 2003;20(1):42–44.
4. Tatum WO. Normal EEG. In: Tatum WO, Husain A, Benbadis SR,
et al., eds. Handbook of EEG Interpretation. New York, NY: Demos;
2008:1–50.
5. Society ACN. Guideline one: minimum technical requirements for
performing clinical electroencephalography [online]. J Clin
Neurophysiol. 1994;11(1):2–5.
6. Fisch BJ. Fisch and Spehlmann’s EEG Primer: Basic Principles of
Digital and Analog EEG. 3rd ed. Amsterdam: Elsevier; 1999:631.
7. Alarcon G, Binnie CD, Garcia Seoane JJ, et al. Mechanisms
involved in the propagation of interictal epileptiform discharges in
partial epilepsy. Electroencephalogr Clin Neurophysiol Suppl.
1999;50:259–278.
8. Klass DW, Westmoreland BF. Electroencephalography: general prin-
ciples and adult electroencephalograms. In: Daube J, ed. Clinical
Neurophysiology. Philadelphia, PA: F.A. Davis; 1996:73–103.
9. Matsuo F, Shaner EE. EEG artifact due to a metallic prosthetic eye.
Am J EEG Technol. 1975;15:75–78.
10. Klem GH. Artifacts. In: Ebersole JS, Pedley TA, eds. Current
Practice of Clinical Electroencephalography. Philadelphia, PA:
Lippincott Williams & Wilkins; 2003:271–287.
266 Part II ■ Normal EEG
11. Tyner FS, Knott JR, Mayer W Jr. Fundamentals of EEG Technology:
Basic Concepts and Methods. New York, NY: Raven Press;
1983:280–311.
12. Brittenham Dorothea M. Artifacts. In: Daly DD, Pedley TA, eds.
Current Practice of Clinical Electroencephalography. New York, NY:
Raven Press; 1990:85–105.
13. Ebersole JS, Bridgers SL, Silva CG. Differentiation of epileptiform
abnormalities from normal transients and artifacts on ambulatory
cassette EEG. Am J EEG Technol. 1983;23:113–125.
14. Stern JM, Engel Jr J. Artifacts. In: Stern JM, Engel J Jr, eds. Atlas of
EEG Patterns. Philadelphia, PA: Lippincott Williams & Wilkins;
2005:55–86.
15. Guideline twelve: guidelines for long-term monitoring for
epilepsy. J Clin Neurophysiol. 2008;25(3):170–180.
16. Benbadis S, Agrawal V, Tatum W. How many patients with psy-
chogenic nonepileptic seizures also have epilepsy? Neurology.
2001;57(5):915–917.
17. Geddes LA. Bioelectrodes, part I. Am J EEG Technol. 1973;13:
195–203.
18. Geddes LA. Bioelectrodes, part III. Am J EEG Technol. 1973;15:
99–106.
19. Barlow J. Artifact processing (rejection and minimization) in EEG
data processing. In: Lopes da Silva F, van Leeuwen W, Remond A,
eds. Clinical Applications of Computer Analysis of EEG and Other
Neurophysiological Signals. Handbook of EEG. Amsterdam:
Elsevier; 1986:15–64.
20. Gordon M. Artifacts created by imbalanced electrode impedance.
Am J EEG Technol. 1980;20:149–160.
21. Ferree TC, Luu P, Russell GS, et al. Scalp electrode impedance,
infection risk, and EEG data quality. Clin Neurophysiol. 2001;
112(3):536–544.
22. Cooper R, Wosselton J, Shaw JC. EEG Technology. 3rd ed. London:
Butterworth-Heinemann; 1990.
23. Blum D. Computer-based electroencephalography: technical
basics, basis for new applications, and potential pitfalls.
Electroencephalogr Clin Neurophysiol. 1998;106(2):118–126.
24. Epstein CM. Aliasing in the visual EEG: a potential pitfall of video
display technology. Clin Neurophysiol. 2003;114(10):1974–1976.
25. Hirsch LJ, Kull L. Continuous EEG monitoring in the intensive
care unit. Am J Electroneurodiagn Technol. 2004;44(3):137–158.
26. Ford RG. A practical guide to EEG recording technique. Am J EEG
Technol. 1981;21:79–101.
27. Seaba PJ. Differential amplifiers and their limitations. Am J EEG
Technol. 1984;24(1):11–23.
28. Sethi NK, Torgovnick J, Sethi PK. Rhythmic artifact of physiother-
apy in intensive care unit EEG recordings. J Clin Neurophysiol.
2008;25(1):62.
29. MacGillivray BB. Section IV: artefacts, faults, and fault-finding. In:
Remond A, ed. Handbook of Electroencephalography and Clinical
Neurophysiology. Amsterdam: Elsevier; 1974:88–102.
30. Lininger AW, Volow MR, Gianturco DT. Intravenous infusion
motor artifact. Am J EEG Technol. 1981;21(4):167–173.
31. Khambatta HJ, Stone JG, Wald A, et al. Electrocardiographic artifacts
during cardiopulmonary bypass. Anesth Analg. 1990;71(1):88–91.
32. McGrath B, Columb M. Renal replacement therapy causing ECG
artefact mimicking atrial flutter. Br J Intensive Care. 2004; 14:49–52.
33. Young B, Osvath L, Jones D, et al. A novel EEG artifact in the inten-
sive care unit. J Clin Neurophysiol. 2002;19(5):484–486.
34. Saunders MG. Artifacts: activity of noncerebral origin in the EEG.
In: Klass DW, Daly DD, eds. Current Practice of Clinical
Electroencephalography. New York, NY: Raven Press; 1979:37–68.
35. Sethi PK, Sethi NK, Torgovnick J. Mobile phone artifact. Clin
Neurophysiol. 2006;117(8):1876–1878.
36. Nuwer M. Basic electrophysiology: evoked potentials and signal
processing. In: Nuwer M, ed. Evoked Potential Monitoring in the
Operating Room. New York, NY: Raven Press; 1986:5–49.
37. Egol AB, Guntupalli KK. Intravenous infusion device artifact in
the EEG—confusion in the diagnosis of electrocerebral silence.
Intensive Care Med. 1983;9(1):29–32.
38. Patel SI, Souter MJ. Equipment-related electrocardiographic arti-
facts: causes, characteristics, consequences, and correction.
Anesthesiology. 2008;108(1):138–148.
39. Hammond EJ, Uthman BM, Reid SA, et al. Vagus nerve stimula-
tion in humans: neurophysiological studies and electrophysiolog-
ical monitoring. Epilepsia. 1990;31(suppl 2):S51–S59.
40. Nuwer MR. Quantitative EEG: I. Techniques and problems of fre-
quency analysis and topographic mapping. J Clin Neurophysiol.
1988;5(1):1–43.
41. Anderer P, Roberts S, Schlogl A, et al. Artifact processing in com-
puterized analysis of sleep EEG—a review. Neuropsychobiology.
1999;40(3):150–157.
42. O’Donnell RD, Berkhout J, Adey WR. Contamination of scalp
EEG spectrum during contraction of cranio-facial muscles.
Electroencephalogr Clin Neurophysiol. 1974;37(2):145–151.
43. Ille N, Berg P, Scherg M. Artifact correction of the ongoing EEG
using spatial filters based on artifact and brain signal topogra-
phies. J Clin Neurophysiol. 2002;19(2):113–124.
44. Astolfi L, Cincotti F, Mattia D, et al. Removal of ocular artifacts for
high resolution EEG studies: a simulation study. Conf Proc IEEE
Eng Med Biol Soc. 2006;1:976–979.
45. Robert C, Gaudy JF, Limoge A. Electroencephalogram processing
using neural networks. Clin Neurophysiol. 2002;113(5):694–701.
46. Young GB, Campbell VC. EEG monitoring in the intensive care
unit: pitfalls and caveats. J Clin Neurophysiol. 1999;16(1):40–45.
 Part III Clinical EEG: General
Patterns of Unclear Significance
JONATHAN CHARLES EDWARDS AND EKREM KUTLUAY
T FAST ALPHA VARIANT/ SLOW
he subject of patterns of unclear significance encom-
passes a selection of electroencephalographic findings
that may look abnormal, but usually are not. These vari-
ants can be challenging, and may commonly lead to misinter-
pretation. These variants can be mainly grouped into two types:
variant rhythms and variant transients. Some of these findings
will also be discussed in the chapters on the normal EEG
(Chapters 9 and 10).
The difficulties of interpretation posed by these variants arise
for several reasons. These patterns possess features that are sim-
ilar in some ways to abnormal patterns. Beginning with variant
rhythms, as a general rule, a few common features help distin-
guish these rhythms from definitely abnormal findings: first, the
variant rhythms are, for the most part, monomorphic. The term
“monomorphic” simply means that each of the waves in the
rhythm looks distinctly like the other waves in the rhythm. The
waves have a characteristic shape, or “morphology,” that repeats.
Second, most of the variants have an arch-shaped appearance,
although some of the normal rhythms may have notches in some
or all of the waves. Third, and most importantly, the variant
rhythms for the most part do not evolve.
One important source of misinterpretation of these patterns
may simply be that many of these patterns are uncommon. A sig-
nificant number of clinicians who read EEGs in clinical practice
may have only 2 to 3 months of closely supervised EEG training.
The chances of seeing a rare pattern are quite low during this
short training. Without extensive preparation in the fundamen-
tals of electroencephalography and extensive supervised experi-
ence, misinterpretation of a rare pattern would not be surprising.
Finally, some confusion may arise from the fact that many of
these patterns have changed names since first being described.
In some cases, the nomenclature has changed numerous times.
For some of patterns, numerous names are still concurrently in
use today.
A particular challenge in assessing the true prevalence and
clinical significance of these patterns is the tendency toward
subject bias. Most patients who undergo EEG are doing so
because of neurological symptoms, and many of these patients
have a clinical history suggestive of epilepsy. While a few stud-
ies have looked at these findings in a general population, most
have reviewed their presence in a clinical population. Since EEG
is used to help diagnose epilepsy or other neurological condi-
tions, the true predictive value of these findings may be biased
(Tables 14-1 and 14-2).
CHAPTER
14
ALPHA VARIANT
At times, harmonics or subharmonics of the background
rhythm are seen. When this occurs, the resultant appearance is
of a rhythm that is twice as fast or half as fast as the awake back-
ground. When the appearance is of a rhythm that is twice the
frequency of the awake background, it is referred to as a “fast
alpha variant” (1,2). A “slow alpha variant” has an apparent fre-
quency that is about half that of the awake background (1–3).
These harmonics will often have a notched appearance, as if
many of the waves are simply being cut in half, but only partly
so. Slow and fast alpha variants are reactive to eye opening and
closure (Fig. 14.1).
Fast alpha variant is easy to detect, since awake posterior
background rhythms in the mid to upper beta range would
be quite unusual. True beta rhythms are most often seen in
the frontal, central, and parietal regions. When a posterior
background rhythm of 16, 18, or 20 is seen, it is easy to think
of fast alpha variant. However, slow alpha variant is more
challenging, and may easily be misinterpreted as occipital
slowing. This tendency to misread slow alpha variant as
abnormal stems mostly from the fact that slowing is such an
easily recognized abnormality, and all electroencephalogra-
phers look for signs of slowing. As a general helpful rule it is
good to remember that any time one sees occipital slowing,
one must think of slow alpha variant first, before calling the
record abnormal. One should look to find a sample of nor-
mal awake background within the record. If the slowing is
approximately half the frequency of the patient’s awake back-
ground, then the slowing is probably just a slow alpha vari-
ant. An additional clue can be small notches in some of the
“slow waves,” bearing in mind that the notches can be quite
subtle (Fig. 14.2).
ALPHA SQUEAK
Immediately on eye closure, the awake background is some-
times initially faster, and of lower voltage. Over approximately
0.5 to 2 seconds, the background slows, and increases in ampli-
tude to become the normal awake background. This initial
“speeding up” of the background immediately on eye closure is
referred to as an “alpha squeak,” and is a normal finding
(Fig. 14.3) (2,4–6).
267
 Tabl e 1 4 . 1

Summary of Characteristics for Patterns of Unclear Significance

Finding Peak Age State Location Frequency Duration Characteristics

Slow or fast alpha Children and Awake, eyes closed Occipital Half or double Brief or lasting Sinusoidal, archiform,
variant adults the posterior many seconds notched, flat-topped
dominant rhythm
Alpha squeak Children and Awake, immediately Occipital Initially fast (beta) Typically Sinusoidal,
adults on eye closure slowing rapidly to 0.5–2 seconds spindle-shaped
the posterior
dominant rhythm
Rhythmic of Younger to Drowsiness, light Midtemporal, 5–7 Hz Several seconds, Arched, notched,
midtemporal middle-aged sleep at times or up to a flat-topped
theta bursts of adults with spread to minute or more Notche may
drowsiness parasagittal or give the
(“RMTD”) occipital– waves a sharply
temporal region contoured
Bilateral or appearance
independent
over the right
or left side
Midline theta Children and Common in Midline (CZ, 4–7 Hz Brief or Arched, notched,
rhythm (“Cigánek adults drowsiness FZ), parasagittal lasting many flat-topped
rhythm”) and during seconds
mental activation.
Less common
in resting
wakefulness
Subclinical Adults, mostly Drowsiness and Maximal in May begin with Brief or Sharply contoured,
rhythmic age 40s–80s resting wakefulness temporal and delta frequency, lasting minutes archiform,
electrographic (mean 60s) (rare cases parietal regions but then notched, flat-topped
discharge in reported in sleep) Usually bilateral, increases to
adults (SREDA) but may be 4–7 Hz
asymmetric

268
 14- and 6-Hz Childhood and Drowsiness, Posterior 14 or 6 Hz 0.5–1 Archiform,
positive bursts adolescence light sleep temporal second comb-like,
low amplitude
6-Hz spike Children and Drowsiness, Highest 5–6 Hz 1–2 seconds Usually
and wave adults light sleep amplitude low-amplitude
bursts over fronto- ( 25 V) and
central short-duration
regions ( 30 msec) spike
component
Benign sporadic Mainly in Drowsiness, Anterior and Sporadic Usually 50 msec Mono- or diphasic
sleep spikes adults. Also light sleep midtemporal spike with steep
reported in descending arm
children and
adolescents
Wickets Adults Awake, sleep Temporal 6–11 Hz if Average of Archiform,
regions they occur 2–4-second monophasic
in runs bursts in
sleep
Frontal arousal Children Awakening, arousal Frontal 6.5–8.5 Hz Up to 13 Monomorphic
rhythm from sleep seconds

269
270 Part III ■ Clinical EEG: General

Tabl e 1 4 . 2 to the occipital–temporal region. Like mu rhythm, RMTD may be

seen bilaterally or may be independent over the right or left side.
Nomenclature Over Several Decades Like most rhythmic variants, it is typically arch-shaped, and is often
notched. These notches may give the waves a somewhat sharply
Other Terminologies contoured appearance, which may be mistaken for a sharp wave. At
EEG Pattern Used Previously times, the notched arches may have a flat-topped appearance
Rhythmic midtemporal Psychomotor variant (7) (1,7,8). RMTD may occur in runs that last several seconds, or even
theta bursts of drowsiness Rhythmic midtemporal up to a minute or more. However, like other normal rhythms,
(RMTD) discharge (8) RMTD can be clearly distinguished from seizure activity, because
Rhythmic theta bursts it is monomorphic, monorhythmic, and does not evolve.
of drowsiness (1) The reported overall prevalence of RMTD is quite low, rang-
Rhythmic temporal theta ing from 0.1% to 2% of records (9–13). The prevalence appears
bursts of drowsiness to be lower in age-matched normal controls than it is in EEGs
performed in neurological patients (9). While some studies (8)
Midline theta rhythm Theta-discharges in the have reported that a significant number of patients who are
middle line (15) found to have RMTD may have epilepsy, the majority of
Electrographic theta patients with this finding will not. RMTD has been reported in
discharges in the the presence of a structural abnormality (14). The overall clin-
midline (18) ical significance of this finding is unclear. It is most commonly
Cigánek rhythm considered a normal variant (Fig. 14.4).
Rhythmical midline
theta (23)
Subclinical rhythmic Paroxysmal discharge of the MIDLINE THETA RHYTHM
EEG discharge of adults temporo-parieto-occipital (CIGÁNEK RHYTHM)
(SREDA) junction (24)
14- and 6-Hz positive 14- and 6-Hz positive spikes Midline theta rhythm was first described by Cigánek, and is still
bursts (32), Ctenoids (33) commonly referred to as “Cigánek rhythm” (15). As is the case
with many variant rhythms, the midline theta rhythm was orig-
6-Hz spike and wave Spike and wave phantom (43) inally thought to be indicative of epilepsy, but was later found
bursts to be present in normal subjects also. Cigánek’s 1961 report is
Benign sporadic sleep Small sharp spikes (31), actually titled “Theta-discharges in the middle line—EEG
spikes Benign epileptiform symptom of temporal lobe epilepsy.” Midline theta is seen dur-
transients of sleep (50) ing wakefulness and drowsiness. Similar to RMTD, the midline
Wickets Wicket spikes (58) theta rhythm is typically a 4- to 7-Hz arch-shaped or sinusoidal
rhythm that may at times be notched or flat-topped (16). The
Frontal arousal rhythm Frontal arousal rhythm (63) main difference in appearance from RMTD is the location
(typically maximum over CZ), and presence in wakefulness.
The reported prevalence of midline theta rhythm has been
RHYTHMIC MIDTEMPORAL THETA highly variable, ranging from 2% to 35% of subjects (17,18). This
BURSTS OF DROWSINESS wide disparity is in part accounted for by subject selection.
Maulsby included the presence of midline theta during drowsiness
First described by Gibbs and Gibbs, this pattern has been given sev- and determined that it was present in 35% of normal adults. Many
eral names over nearly six decades (7). This pattern’s earlier name others have reported midline theta during mental tasks in a high
“psychomotor variant” is still commonly used. “Psychomotor” is number of patients (19–22). With the exclusion of midline theta
an older term for a complex partial seizure, typically of temporal rhythm that is seen only during drowsiness of mental activation,
lobe origin. Rhythmic midtemporal theta bursts of drowsiness the prevalence has been reported to be low in normal controls, and
(RMTD) has commonly been referred to as “psychomotor variant,” higher in patients with epilepsy. A more recent study (23) sug-
because it was felt to have a strong resemblance to a temporal lobe gested that when drowsiness and mental activation are excluded,
seizure. Subsequent names have included: “rhythmic midtemporal midline theta rhythm was common (26% overall) in their patients
discharge” (8), “rhythmic theta bursts of drowsiness,” and “rhyth- with epilepsy, and uncommon (0/54) in controls. Interestingly, in
mic temporal theta bursts of drowsiness” (1). the same series, and using the same exclusion criteria, midline
The name RMTD is widely accepted, and quite clear in listing theta rhythm was seen much more frequently in frontal lobe
the most distinctive characteristics. RMTD is seen mostly during epilepsy (48.1%) than in temporal lobe epilepsy (3.7%).
drowsiness and light sleep, although it may rarely be seen The clinical significance of midline theta rhythm is not univer-
during wakefulness. It occurs in bursts or trains of rhythmic 5- to sally agreed upon (16,23). It is typically regarded as a non specific
7-Hz activity. RMTD is maximal in the midtemporal leads, and finding, and is generally read as normal. Midline theta rhythm
may at times spread to the parasagittal regions, or more posteriorly during the awake recording and without mental tasks may
 Chapter 14 ■ Patterns of Unclear Significance 271

Figure 14.1 Fast alpha variant.

Figure 14.2 Slow alpha variant.

272 Part III ■ Clinical EEG: General

Figure 14.3 Alpha squeak.

Figure 14.4 Rhythmic midtemporal theta bursts of drowsiness.


Figure 14.5 Midline theta rhythm.
warrant further study. A potential practical challenge and con-
founding variable would be in actually knowing what a subject is
doing mentally in an awake, at rest recording, and ruling out that
particular mental tasks are taking place. Most evidence suggests
that the presence of midline theta rhythm during drowsiness or
during mental task is not an abnormality (Fig. 14.5).
SREDA (SUBCLINICAL RHYTHMIC
ELECTROGRAPHIC DISCHARGE IN ADULTS)
SREDA is a fairly rare pattern that can be easily misinterpreted. It
is perhaps the most challenging pattern among those that are
considered normal variants. First described in 1961 by Naquet et
al., this pattern was originally thought to be associated with
hypoxia (24). Westmoreland and Klass described 65 patients who
had 142 EEG recordings, and provided the term “Subclinical
Rhythmic EEG Discharge in Adults.” In this series, SREDA was
seen in patients aged 42 to 80, with an average age of 61 (25).
The pattern is seen mostly during relaxed awake or drowsy
states, although SREDA can be seen during sleep (25,26). The
typical appearance of SREDA is a diffuse, sharply contoured,
theta rhythm that is usually maximal in the temporal and
parietal regions. While SREDA is usually bilateral, it may be
asymmetric. An interesting feature of SREDA is that it often
occurs in long runs, such as 15 seconds to a minute or more
(25). The appearance on visual inspection is of a monomorphic
rhythm. However, a more recent detailed digital study using fre-
quency spectral analysis and Laplacian montages suggests that
SREDA is “composed of a complex mixture of multiple rapidly
shifting frequencies predominantly in the theta range, which show
poor spatial and temporal resolution” (Fig. 14.6) (27).
Chapter 14 ■ Patterns of Unclear Significance 273
This particular pattern has a distinctive feature that is differ-
ent from other non-epileptic variants and that makes it especially
difficult to recognize and easy to misdiagnose. The SREDA pat-
tern may initially evolve. SREDA may begin with a single sharp
transient or a series of sharp transients, initially at delta frequen-
cies, and then increasing to theta, before becoming a well-defined
4- to 7-Hz rhythm. A smaller number of cases remain in the 2- to
4-Hz frequency range. After the initial period in which the
rhythm is established, however, SREDA does not continue to
evolve, despite the fact that SREDA may be seen in prolonged
runs of 20 seconds to several minutes. The rhythm ends abruptly
in about half of patients, and in the other cases, the SREDA grad-
ually dissipates and merges with ongoing normal background
rhythms. Unlike seizures, SREDA is not followed by “postictal”
slowing, and no clinical changes accompany SREDA.
While this pattern may easily be mistaken for an ictal pat-
tern, evidence suggests that this pattern has no clear association
with seizures (13,25,28,29). In a large series, there was not only
a low incidence of clinical history of seizures or subsequent
development of seizures, but also a “uniform lack of any clinical
accompaniment during the bursts, even with the involvement of
both hemispheres and persistence of several minutes” (25). One
study of ictal single photon emission computer tomography
demonstrated no significant changes in uptake during the
bursts, further suggesting that the pattern is not ictal (30).
14- AND 6-HZ POSITIVE BURSTS
Like many other variants, this unique EEG phenomenon was
also described by Gibbs and Gibbs (31,32). Although it was pre-
viously called 14- and 6-Hz positive spikes, the current
274 Part III ■ Clinical EEG: General
Figure 14.6 Subclinical electrographic discharges of adults (SREDA).
preferred term is 14- and 6-Hz positive bursts. Lombroso et al.
also proposed the name “ctenoids” due to their appearance and
complex morphology resembling a comb (ktenos in Greek
translates to “comb”) (33).
The typical bursts last between 0.5 and 1 second and usually
occur during light sleep or drowsiness (34–36). However, a few
cases with occurrence in REM sleep have also been reported
(37). The bursts can be unilateral or bilateral but usually show
a shifting predominance if enough bursts are recorded in the
same patient (34–36). They are usually small in amplitude;
therefore, referential or long-distance electrode montages
reveal better waveforms (34–36). Although the laterality of
these bursts can appear to shift, the maximum amplitude is
expressed over the posterior temporal head regions (34–36).
There seems to be strong age predilection. This activity ini-
tially starts to appear during early childhood and reaches its
peak during adolescence (34–37). The coexistence of 14- and 6-
Hz positive bursts and 6-Hz spike and wave bursts was also
reported. Silverman hypothesized that a maturational sequence
might be responsible for this co-occurrence, and that most of
the 6-Hz spike and wave bursts are part of the 14- and 6-Hz pos-
itive bursts (38). The 14- and 6-Hz positive bursts are still seen
during adulthood but the incidence decreases with advanced age
(34). Lombroso et al. reported highest incidence of 58% in the
literature in teenage boys between the ages of 13 and 16 years
(33). In adults, the incidence is much lower. An incidence as low
as 4% was seen in a group of psychiatric patients (38), whereas
a study of EEG in asymptomatic normal adults revealed a rate
around 12% (39). In fact, the latter claimed that 14- and 6-Hz
positive bursts were the most common pattern of uncertain sig-
nificance in their series of 100 asymptomatic adults (39).
Several reports were published in the literature correlating
the existing of this pattern with certain symptoms and illnesses.
Wide variety of psychiatric conditions and neurovegetative
symptoms were subjected to research (33,34,40). Epilepsy was
also thought to be more frequent in patients with 14- and 6-Hz
positive bursts (41). However, most of the patients with behav-
ioral symptoms and organic CNS impairment were children or
adolescents, which is the age of peak incidence for this phe-
nomenon (42). One study suggests that a majority of patients
with seizures and 14- and 6-Hz positive spikes also showed
other epileptiform abnormalities in their EEGs (41).
This controversial pattern seems to be an age-related phe-
nomenon. Although there are arguments regarding a patholog-
ical significance, the reported incidence of up to 58% in normal
teenagers and 12% in healthy adults makes this relatively hard
to prove (Fig. 14.7).
SIX-HZ SPIKE AND WAVE BURSTS
Six-Hz spike and wave bursts were first described in 1950 as
“wave and spike phantom” (32,43). The first collected cases
revealed a unique EEG pattern of 1-second bursts of 5- to 6-Hz
spike and waves (43). The average amplitude of the discharge
was low, around 25 V or less, with the spike component less
than 30-msec duration (43–45). This pattern can be seen in
both children and adults, with amplitude usually higher over
the fronto-central regions (43,44). These bursts were mostly

Figure 14.7 Fourteen- and 6-Hz positive bursts.
seen during drowsiness and light sleep but they were also seen
during full wakefulness (43). Bursts are usually diffuse but may
show anterior or posterior predominance. Marshall suggested
the comparison linkage that measures the potential difference
between analogous scalp areas for better viewing rather than
traditional bipolar or referential montages (43).
Average incidence of 6-Hz spike and wave bursts is
between 1% and 2.5% of all EEG recordings (13,43,44,46,47).
Several authors tried to link this controversial pattern to
specific complaints, ranging from simple fainting to vegeta-
tive or dysautonomic pathologies, and from psychiatric dis-
turbances to seizures (42,43,44,48). However, occurrence in
normal people and absence of clinical symptoms during the
bursts make it difficult to label these discharges as a specific
indicator of any certain disease (34,48). As mentioned earlier,
Silverman suggested that there is relationship as a matura-
tional sequence and waveform spectrum between 6-Hz spike
and wave bursts and 14- and 6-Hz positive bursts (38). He
claimed that most of the 6-Hz spike and wave bursts are part
of the 14- and 6-Hz positive bursts and transition between
these two types may lead to occurrence of both patterns in
the same person (38).
On investigating more than 60,000 EEGs over 30 years,
Hughes proposed two distinct subtypes of this pattern (47).
The features of these two variants were summarized by
acronyms WHAM (Wake, High amplitude, Anterior location,
Male gender) and FOLD (Female gender, Occipital location,
Low amplitude, Drowsiness) (47). Hughes indicated that these
two variants are more extreme forms, and also suggested that
the more components of the WHAM form that exist in a sin-
gle person, the more likely is the association with seizures
Chapter 14 ■ Patterns of Unclear Significance 275
(47). Current consensus is that this EEG pattern is of unclear
clinical significance; however, high amplitude of the spikes,
slower rate than 6 Hz, and persistence of the discharges in
deeper stages of sleep are more likely associated with seizures
(Fig. 14.8) (35,49).
BENIGN SPORADIC SLEEP SPIKES
Small sharp spikes (SSS), the term originally described by Gibbs
and Gibbs, are also known as benign epileptiform transients of
sleep (BETS) or benign sporadic sleep spikes (BSSS) (32,34,50).
However, the latter became a more preferred term recently. As
indicated by the name SSS, these are short in duration and low in
amplitude. Although they are usually less than 50 msec and less
than 50 V, duration can be slightly longer and amplitude can be
slightly higher depending on the recording circumstances and
montage used (34–36). The shape is rather simple and consists of
mono- or diphasic spike with steep descending arm (34,35). An
aftergoing slow wave is not prominent and usually is lower in
amplitude than the spike component (34,36,50). Background
activity at the region of BSSS is not disrupted (34,35). These dis-
charges are best seen during drowsiness and light sleep (stage I
and II) and are best displayed over anterior and midtemporal or
ear electrodes (34,35). Although they have been reported in chil-
dren and adolescents, BSSS are mainly seen in adults (35,51).
The origin and cortical location of the BSSS have been the
subject of a few investigations. Recording with depth electrodes
in two patients revealed similarity to surface findings (52).
Invasively recorded spikes had widespread distribution and
similar simple morphological appearance (53). They occurred
as single sporadic transients without disrupting the ongoing
276 Part III ■ Clinical EEG: General
Figure 14.8 Six-Hz spike and wave discharges.
background activity; however, the amplitude was higher than
surface recordings (52). Another study with subdural electrodes
showed that BSSS were more widespread than on-surface
recordings, and were mainly seen over the posterior mesial tem-
poral area (53).
BSSS are generally not thought to be related to any certain
disease, and are accepted as a variant of normal EEG activity
(35,50,54). However, over the years, several reports tried to
find a connection between BSSS and seizures or certain psy-
chiatric conditions. BSSS were thought to be a sign of epilep-
togenicity. Different authors reported higher incidence of
BSSS in patients with seizures particularly in patients of
younger age (51,55). One study indicated that the highest
incidence is seen in patients younger than 20 years of age, with
a gradual decrease with advanced age leading to complete dis-
appearance after age 80 years (55). Still other studies claimed
a relationship between BSSS and certain psychiatric condi-
tions. Positive correlations were reported with manic depres-
sive psychosis and psychotic depressive episodes as well as
during acute mania (56,57).
BSSS were first described about half a century ago. They are
very common EEG findings seen with up to 25% incidence in
the normal population (36,50). Although several studies
reported increased frequency of BSSS in certain psychiatric
conditions and seizures because of their high incidence in nor-
mal population, we believe this peculiar EEG activity is of
unknown clinical significance with no statistical relationship to
a certain disease (Fig. 14.9).
WICKETS
Wickets are archiform waveforms first described as wicket
spikes by Reiher and Lebel in 1977 (58). They are simple
monophasic waveforms with surface negativity (58). Initial
descriptions revealed that wicket spikes are found in both
awake and sleep EEGs (58). Because of the intermixed back-
ground activities during the awake state, they are better recog-
nized during sleep (58,59). They are best seen during initial
stages of sleep but also reported during REM sleep (58,60).
Their location remains unchanged during different stages of
sleep (60). Wickets can occur as isolated “spikes” or they may
come in runs (34,36,58). If they occur in runs, the usual fre-
quency is between 6 and 11 Hz. The amplitude of a wicket
spike may range between 60 and 210 V and is maximally
expressed over the temporal regions (34,36,58). Although they
shift sides, often one side is more dominant than the other
(58).
Wickets are usually seen in adults (58). They are almost
exclusively reported in adults 30 years or older; however, cases
as young as 20 years of age have also been reported (12,36,58).
Wickets are reported to be an uncommon EEG pattern. Their
incidence in large EEG series has been reported to be less than
1%, ranging between 0.03% and 0.96% (12,13,58). In adults 30
years and older this incidence may rise up to 2.9% (58).
Wickets are considered to be a normal variant. In Reiher and
Lebel’s original description, wickets occurred four times more
in patients with no history of seizures than in those with a

Figure 14.9 Benign sporadic sleep spikes (BSSS).
history of seizures (58). An experienced electroencephalogra-
pher can differentiate wickets from temporal sharp waves easily
by the lack of aftergoing slow waves and the absence of distor-
tion of the background activities (36,49,58). However, one can
also over-read or misinterpret wickets as epileptiform sharp
waves, especially if they occur in isolation (61,62). Clues men-
tioned above, as well as observation of runs of wicket spikes
somewhere else in the same EEG, should alert the reader that
this is a non- epileptiform variant (Fig. 14.10) (61).
FRONTAL AROUSAL RHYTHM (FAR)
This unique EEG pattern was first described by White and
Tharp in 1974 although the pattern was first recognized in 1969
(63). In their original report, the authors described 8 cases out
of 4780 EEGs reviewed over a 4-year period. This group of eight
children consisted of four males and four females with a mean
age of 6.6 years, ranging from 2 to 14 years. Their initial diag-
noses were mild cerebral dysfunction and/or seizure disorder.
Representative EEG samples showed normal awake and sleep
patterns. However, during arousal from stage II sleep, a 6.5- to
8.5-Hz rhythm was seen over the frontal regions involving
mainly F3 and F4 electrodes with no or little spread to adjacent
regions. The overall amplitude of this activity was between 30
and 150 V, with duration of up to 13 seconds. This pattern was
originally described in children with evidence of minimal cere-
bral dysfunction and seizures but authors noted that the inci-
dence in a normal population was unknown at the time of the
publication (63).
Chapter 14 ■ Patterns of Unclear Significance 277
Twenty-five years after the initial description of this rare and
unusual EEG pattern, Hughes and Daaboul discussed their find-
ings in 50 cases (64). The incidence of FAR was 0.22% in their
series of 22,500 patients seen over an 8-year period. The charac-
teristics of the pattern were similar to those described before.
While FAR was the only atypical pattern in 58% of these patients,
nearly half (42%) also showed other abnormalities, including
focal and generalized epileptiform activities. Clinically, 70% of the
children had a seizure disorder, and 56% had both seizure and
cognitive/behavior disorders. Only 6% of the reported patients
had neither seizure nor cognitive/behavior disorders. However,
the incidence of cognitive/behavior disorders was not statistically
different from their control group whose EEG did not show a FAR
pattern. Their conclusion was that FAR is related to seizures. To
support this view, a recent case report described a 6-year-old boy
with mental retardation and possible seizure disorder (65). A typ-
ical FAR pattern was seen during few episodes associated with eye
fluttering followed by chewing, increased inspiration, and upper
lip quivering. These episodes were all recorded during arousal
from sleep and there were no clinical changes seen when the
patient was awakened without the FAR pattern present.
Few illustrative reports are available in the literature. The
incidence of FAR in general population is unknown. The gen-
eral consensus is that FAR is more of a non specific EEG pattern
with unknown clinical significance (2,35). However, the
reported case series as well as demonstration of FAR as an ictal
pattern in a reported case warrants careful review for other
interictal activity, and detailed questioning for possible seizure-
like activity (Fig. 14.11).
278 Part III ■ Clinical EEG: General

Figure 14.10 Wickets.

Figure 14.11 Frontal arousal rhythm (FAR) (courtesy of Dr William Tatum IV).

REFERENCES
1. Chatrian GE, Bergamini L, Dondey M, et al. A glossary of terms
most commonly used by clinical electroencephalographers.
Electroencephalogr Clin Neurophysiol. 1974;37:538–548.
2. Westmoreland BF, Klass DW. Unusual EEG patterns. J Clin
Neurophysiol. 1990;7:209–228.
3. Goodwin JE. The significance of alpha variants in the EEG, and
their relationship to an epileptiform syndrome. Am J Psychiatry.
1947;104:369–379.
4. Bekkering D, Kamp A, Storm van Leeuwen W, et al. Example of the
use of the spectrograph and the magnetograph—the squeak phe-
nomenon. Electroencephalogr Clin Neurophysiol. 1956;8:721.
5. Storm van Leeuwen W, Bekkering D. Some results obtained with
the EEG spectrograph. Electroencephalogr Clin Neurophysiol.
1958;10:563–570.
6. Storm van Leeuwen W, Kamp A, Kniper J. Concerning the
“squeak” phenomenon of the alpha rhythm. Electroencephalogr
Clin Neurophysiol. 1960;12:244.
7. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. 1st ed.
Cambridge, MA: Lew A. Cummings; 1941:55, 109.
8. Lipman IJ, Hughes JR. Rhythmic midtemporal discharges: an
electro-clinical study. Electroencephalogr Clin Neurophysiol. 1969;
27:43–47.
9. Gibbs FA, Rich CL, Gibbs EL. Psychomotor-variant type of seizure
discharge. Neurology. 1963;13:991–998.
10. Gibbs EL, Gibbs FA. Psychomotor-variant type of paroxysmal
cerebral dysrhythmia. Clin Electroencephalogr. 1989;20:147–152.
11. Maulsby RL. EEG patterns of uncertain diagnostic significance. In:
Klass DW, Daly DD, eds. Current Practice of Electroencephalography.
New York: Raven Press; 1979:411–419.
12. Radhakrishnan K, Santoshkumar B, Venugopal A. Prevalence of
benign epileptiform variants observed in an EEG laboratory from
South India. Clin Neurophysiol. 1999;110:280–285.
13. Santoshkumar B, Chong JJ, Blume WT, et al. Prevalence of benign
epileptiform variants. Clin Neurophysiol. 2009;120:856–861.
14. Hennessy MJ, Koutroumanidis M, Hughes E, et al. Psychomotor
EEG variant of Gibbs: an association with underlying structural
pathology. Clin Neurophysiol. 2001;112:686–687.
15. Cigánek L. Theta-discharges in the middle line —EEG symptom of
temporal lobe epilepsy. Electroencephalogr Clin Neurophysiol. 1961;
13:669–673.
16. Westmoreland BF, Klass DW. Midline theta rhythm. Arch Neurol.
1986;43:139–141.
17. Maulsby R. The Normative Electroencephalographic Data Reference
Library. Final Report, Contract NAS 9-1200. Washington, DC:
National Aeronautics and Space Administration; 1968.
^
18. Mokrán V, Cigánek L, Kabátnik Z. Electrographic theta discharges
in the midline. Eur Neurol. 1971;5:288–293.
19. Gevins A, Smith ME, McEvoy L, et al. High-resolution EEG map-
ping of cortical activation related to working memory: effects of
task difficulty, type of processing, and practice. Cereb Cortex.
1997;7:374–385.
20. Okada S, Urakami Y. Midline theta rhythm revisited. Clin
Electroencephalogr. 1993;24:6–12.
21. Takahashi N, Shinomiya S, Mori D, et al. Frontal midline theta
rhythm in young healthy adults. Clin Electroencephalogr. 1997;
28:49–54.
22. Inanaga K. Frontal midline theta rhythm and mental activity.
Psychiatry Clin Neurosci. 1998;52:555–566.
Chapter 14 ■ Patterns of Unclear Significance 279
23. Beleza P, Bilgin Ö, Noachtar S. Interictal rhythmical midline theta
differentiates frontal from temporal lobe epilepsies. Epilepsia.
2009;50:550–555.
24. Naquet R, Louard C, Rhodes J, et al. A propos de certaines
décharges paroxystiques du carrefour temporo-pariéto-occipital.
Leur activation par l’hypoxie. Rev Neurol. 1961;105:203–207.
25. Westmoreland BF, Klass DW. A distinctive rhythmic EEG dis-
charge of adults. Electroencephalogr Clin Neurophysiol. 1981;51:
186–191.
26. Fleming WE, Avidan A, Malow BA. Subclinical rhythmic electro-
graphic discharge of adults (SREDA) in REM sleep. Sleep Med.
2004;5:77–81.
27. O’Brien TJ, Sharbrough FW, Westmoreland BF, et al. Subclinical
rhythmic electrographic discharges of adult (SREDA) revisited: a
study using digital EEG analysis. J Clin Neurophysiol. 1998;15:
493–501.
28. Begum T, Ikeda A, Takahashi J, et al. Clinical outcome in patients
with SREDA (subclinical rhythmic EEG discharge of adults).
Intern Med. 2006;45:141–144.
29. Westmoreland BF, Klass DW. Unusual variants of subclinical
rhythmic electroencephalographic discharge of adults.
Electroencephalogr Clin Neurophysiol. 1997;102:1–4.
30. Thomas P, Migneco O, Darcourt J, et al. Single photon emission
computed tomography study of subclinical rhythmic electro-
graphic discharge in adults. Electroencephalogr Clin Neurophysiol.
1992;83:223–227.
31. Gibbs FA, Gibbs EL. Fourteen and six per second positive spikes.
Electroencephalogr Clin Neurophysiol. 1963;15:553–558.
32. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 3.
Reading, MA: Addison Wesley; 1964.
33. Lombroso CT, Chwartz IH, Clark DM, et al. Ctenoids in healthy
adults. Controlled study of 14- and 6-per-second positive spiking.
Neurology. 1966;16:1152–1158.
34. Klass DW, Westmoreland BF. Nonepileptogenic epileptiform elec-
troencephalographic activity. Ann Neurol. 1985;18:627–635.
35. Ebersole JS, Pedley TA. Current Practice of Clinical Electroence-
phalography. 3rd ed. Philadelphia, PA: Lippincott Williams &
Wilkins; 2003.
36. Tatum WO, Husain AM, Benbadis SR, et al. Normal adult EEG and
patterns of uncertain significance. J Clin Neurophysiol. 2006;23:
194–207.
37. Beun AM, van Emde Boas W, Dekker E. Sharp transients in the
sleep EEG of healthy adults: a possible pitfall in the diagnostic
assessment of seizure disorders. Electroencephalogr Clin
Neurophysiol. 1998;106:44–51.
38. Silverman D. Phantom spike-waves and the fourteen and six per
second positive spike pattern: a consideration of their relationship.
Electroencephalogr Clin Neurophysiol. 1967;23:207–213.
39. Jabbari B, Russo MB, Russo ML. Electroencephalogram of asymp-
tomatic adults. Clin Neurophysiol. 2000;111:102–105.
40. Boutros N, Fristad M, Abdollohian A. The fourteen and six posi-
tive spikes and attention-deficit hyperactivity disorder. Biol
Psychiatry. 1998;44:298–301.
41. Hughes JR, Cayaffa JJ. Positive spikes revisited—in the adult. Clin
Elecroencephalogr. 1978;9:52–59.
42. Small JG, Sharpley P, Small IF. Positive spikes, spike-wave phan-
toms, and psychomotor variants. A survey of these EEG patterns
in psychiatric patients. Arch Gen Psychiat. 1968;18:232–238.
43. Marshall C. Some clinical correlates of the wave and spike phan-
tom. Electroencaphalogr Clin Neurophysiol. 1955;7:633–636.
280 Part III ■ Clinical EEG: General
44. Tharp BR, Arsenal E. The 6-per-second spike and wave complex.
The wave and spike phantom. Arch Neurol. 1966;15:533–537.
45. Tharp BR. The six per second spike and wave complex (the wave
and spike phantom). Electroencephalogr Clin Neurophysiol. 1967;
23:291.
46. Olson SF, Hughes JR. The clinical symptomatology associated with
the 6 c/sec spike and wave complex. Epilepsia. 1970;11:383–393.
47. Hughes JR. Two forms of the 6/sec spike and wave complex.
Electroencephalogr Clin Neurophysiol. 1980;48:535–550.
48. Thomas JE, Klass DW. Six-per-second spike-and-wave pattern in
the electroencephalogram. Neurology. 1968;18:587–593.
49. Westmoreland BF. Epileptiform electroencephalographic patterns.
Mayo Clin Proc. 1996;71:501–511.
50. White JC, Langston JW, Pedley TA. Benign epileptiform transients
of sleep. Clarification of the small sharp spike controversy.
Neurology. 1977;27:1061–1068.
51. Saito F, Fukushima Y, Kubota S. Small sharp spikes: possible rela-
tionship to epilepsy. Clin Electroencephalogr. 1987;18:114–119.
52. Westmoreland BF, Reiher J, Klass DW. Recording of small sharp
spikes with depth electroencephalography. Epilepsia. 1979;20:
599–606.
53. McLachlan RS, Lubus N. Cortical location of benign paroxysmal
rhythms in the electrocorticogram. Can J Neurol Sci. 2002;29:
154–158.
54. Reiher J, Klass DW. “Small sharp spikes” (SSS): electroencephalo-
graphic characteristics and clinical significance. Electroencephalogr
Clin Neurophysiol. 1970;28:94.
55. Hughes JR, Gruener G. Small sharp spikes revisited: further data
on this controversial pattern. Clin Electroencephalogr. 1984;15:
208–213.
56. Small JG. Small sharp spikes: EEG signals of psychiatric signifi-
cance. Electroencephalogr Clin Neurophysiol. 1970;28:417–422.
57. Inui K, Motomura E, Okushima R, et al. Electroencephalographic
findings in patients with DSM-IV mood disorder, schizophrenia,
and other psychotic disorders. Biol Psychiatry. 1998;43:69–75.
58. Reiher J, Lebel M. Wicket spikes: clinical correlates of a previously
undescribed EEG pattern. Can J Neurol Sci. 1977;4:39–47.
59. Waveform Window #11. Wicket Spikes. Am J Electroneurodiagn
Technol. 2008;48:52–55.
60. Gelisse P, Kuate C, Coubes P, et al. Wicket spikes during rapid eye
movement sleep. J Clin Neurophysiol. 2003;20:345–350.
61. Benbadis SR, Tatum WO. Overinterpretation of EEGs and misdi-
agnosis of epilepsy. J Clin Neurophysiol. 2003;20:42–44.
62. Krauss GL, Abdallah A, Lesser R, et al. Clinical and EEG features of
patients with EEG wicket rhythms misdiagnosed with epilepsy.
Neurology. 2005;64:1879–1883.
63. White JC, Tharp BR. An arousal pattern in children with organic
cerebral dysfunction. Electroencephalogr Clin Neurophysiol.
1974;37:265–268.
64. Hughes JR, Daaboul Y. The frontal arousal rhythm. Clin
Electroencephalogr. 1999;30:16–20.
65. Hughes JR. The frontal arousal rhythm (FAR) is an ictal pattern: a
case report. Clin Electroencephalogr. 2003;34:13–14.

EEG of Degenerative Disorders
of the Central Nervous System
INTRODUCTION
Degenerative disorders of the central nervous system encom-
pass a wide range of genetic and metabolic conditions, which
result in progressive neurologic disability. Senile dementia can
be included under this heading, but is discussed elsewhere in
the text. Metabolic encephalopathies are distinct from chronic,
degenerative conditions in that they often result from systemic
disorders and are commonly reversible. Given the wide array of
etiologies for degenerative diseases, they can be categorized in
multiple ways. The most common classification system is to
separate these disorders based on the type of cellular organelle
that is primarily involved, as is done in this chapter. Special
attention is also given to the predominant neuroanatomical
structures involved, since this allows for an EEG to neu-
roanatomical correlation (Table 15-1).
LYSOSOMAL DISORDERS
Lysosomes are membrane-bound organelles that function as the
digestive system of the cell. They contain enzymes that break
down macromolecules originating either internally or exter-
nally. Lysosomal storage disorders occur when there is deficiency
Tabl e 1 5 . 1
Disorders affecting primarily cortex
Tay–Sachs disease
Neuronal ceroid lipofuscinosis
Mucopolysaccharidosis
Sialidosis
Rett syndrome
Menkes disease
Alpers–Hunttenlocher syndrome
Disorders affecting primarily white matter
Metachromatic leukodystrophy
Globoid leukodystrophy
X-linked adrenoleukodystrophy
Neonatal adrenoleukodystrophy
Disorders affecting primarily basal ganglia
Leigh syndrome
Neurodegeneration with brain iron accumulation
Infantile neuroaxonal dystrophy
Wilson disease
CHAPTER
JOHN GAITANIS
15
of one or more of these enzymes. Undegraded material accumu-
lates, leading to cellular and organ dysfunction. Most of these
conditions exhibit autosomal recessive inheritance.
GM2 Gangliosidoses/ Tay–Sachs Disease
The infantile form of Tay–Sachs disease is named for the British
ophthalmologist Warren Tay, who, in 1881, described the
cherry-red spot, and the American neurologist Bernard Sachs,
who reported on the clinical features and the enlarged pyrami-
dal neurons of the condition. The biochemical understanding
came about when Ernst Klenk found a novel group of glycol-
ipids in the brains of these patients, which was later discovered
to be GM2 ganglioside. Its accumulation in neurons results
from a deficiency of the lysosomal enzyme hexosaminidase A.
The hexosaminidase enzyme is comprised of alpha and beta
subunits. Hexosaminidase A consists of one alpha and one beta
subunit and hexosaminidase B is composed of two beta sub-
units. In Tay–Sachs disease, there is loss of the alpha subunit
resulting in absence of hexosaminidase A.
The result is a neurologically devastating disease that begins
in the first months of life. The initial symptom is excessive star-
tle myoclonus to noise, sound, or touch. The startle does not
attenuate with subsequent exposures, distinguishing it from a
normal Moro response. The myoclonic movements consist of
arm and leg extension. As the disease progresses, motor, visual,
and intellectual impairments develop. The affected infant
develops axial hypotonia with appendicular spasticity and
increased reflexes. Motor skills, such as head control, rolling,
and purposeful hand movements, are lost. Visual loss ensues as
lipid accumulates in the retinal ganglion cells, resulting in a
whitish discoloration around the fovea—the characteristic
cherry-red spot. With intraneuronal accumulations of GM2
ganglioside, macrocephaly develops. The children experience
progressive cognitive impairment and seizures. Feeding and
autonomic disturbances develop. Between 3 and 5 years of age,
the patient succumbs to cachexia or aspiration pneumonia.
The disease affects multiple regions of the brain. The basal
ganglia and cerebral white matter show low-density signal on
CT in early stages and enlargement of the caudate nucleus later
in the disease course (1). Early into the disease, the interictal
EEG may be normal. By age 1, there is a rapid and progressive
deterioration of the EEG (2). Paroxysmal features are not
prominent, but during myoclonic seizures, widespread spike
and sharp waves are seen (3). A progressive decline in voltage
occurs in later stages of the disease, likely resulting from contin-
281
282 Part III ■ Clinical EEG: General
ued neuronal loss. The electroretinogram remains normal even
in advanced stages (2,4), since only the ganglion cell layer of the
retina is affected. Progressive loss of the VEP occurs between 9
and 15 months (2).
Sandhoff disease, resulting from mutations of the beta sub-
unit of hexosaminidase A, is phenotypically similar to
Tay–Sachs disease, the only differences being the presence
of hepatosplenomegaly, cardiomyopathy, or N-acetylglu-
cosamine-containing oligosaccharides in the urine of patients
with Sandhoff disease. The AB variant occurs when there is a
deficiency of the GM2 activator, which is necessary for hydrol-
ysis of GM2 gangliosides by hexosaminidase A. Its phenotype
is indistinguishable from the infantile form of Tay–Sachs (5).
Late-onset variants of Tay–Sachs also exist, and can manifest in
childhood, adolescence, or adulthood. They may result in a
wide constellation of symptoms including spastic paraparesis,
ataxia with cerebellar atrophy, seizures, or psychiatric symp-
toms. Psychosis and depression can even be the initial manifes-
tation in adult patients (6). EEG changes in late-onset GM2
gangliosidosis are variable and unrelated to age or enzyme
defect (2).
GM1 Gangliosidosis
GM1 gangliosidosis results from a deficiency of -galactosi-
dase, which is responsible for cleavage of the terminal galac-
tose of GM1. The result is neuronal storage of
monosialoganglioside GM1. Histopathology reveals a marked
decrease in the number of oligodendrocytes and myelin
sheaths. As a result, spasticity, tonic spasms, and pyramidal
signs predominate. As in Tay–Sachs disease, a cherry-red spot
is seen and the ERG remains normal (2). The EEG also shows
a progressive deterioration as it does in GM2 patients. EEGs
universally reveal abnormal slowing, which worsens as the
disease advances. By 2 to 3 years of age, 4 to 5 cycle/sec rhyth-
mic activity is seen in the temporal regions. Paroxysmal dis-
charges, however, are not commonly observed in GM1
patients (7).
Neuronal Ceroid Lipofuscinoses
The neuronal ceroid lipofuscinoses (NCL) are a group of neu-
rodegenerative conditions caused by intralysosomal accumu-
lation of ceroid and lipofuscin (8). Originally described as a
form of “amaurotic familial idiocy,” the group of disorders
was renamed NCL by Zeman to better distinguish them from
gangliosidoses (9). Clinically, the disease group is character-
ized by cognitive decline, progressive myoclonic epilepsy, and
motor regression. Some forms also cause visual failure (8).
The four originally described phenotypes are organized
according to age of onset: infantile (INCL), late infantile
(LINCL), juvenile (JNCL), and adult neuronal ceroid lipofus-
cinosis (ANCL). Patients with INCL are normal at birth, and
exhibit developmental delays or myoclonic seizures between 6
and 24 months of age. In LINCL, patients experience develop-
mental regression and epilepsy between 2 and 4 years. JNCL
develops between 4 and 10 years with visual loss being the ini-
tial symptom and epilepsy following shortly afterwards.
ANCL appears sometime around 30 years of age and presents
with progressive myoclonic epilepsy, dementia, behavioral
abnormalities, ataxia, and pyramidal or extrapyramidal signs
(10). These disorders are inherited in an autosomal recessive
pattern and there are at least seven genes responsible for the
clinical phenotypes.
MRI findings in INCL include cerebral atrophy, callosal
thinning, T2-weighted hyperintensities in the thalami, and
cerebellar atrophy. After 4 years of age, the atrophy is severe and
the entire white matter shows abnormally high signal intensity.
SPECT studies show progressive cerebral and cerebellar hypop-
erfusion with relative sparing of the basal ganglia. LINCL and
JNCL reveal lesser degrees of atrophy (11).
The EEG findings reflect the predominance of cortical
abnormalities on MRI and SPECT. In INCL, the EEG initially
shows a lack of attenuation to eye opening and a disappear-
ance of sleep spindles. The background activity gradually
attenuates and becomes flat by 3 years of age. ERGs are
unrecordable by age 3 and VEPs are unrecordable by 4 years.
There is progressive attenuation of cortical somatosensory
potentials (12). In LINCL, seizures are the presenting symp-
tom, and the EEG reveals occipital spike waves in response to
photic stimulation at 1 to 2 Hz. The ERG is abnormal at pres-
entation and extinguishes in time. VEPs are enhanced, but
diminish in the final stages of the disease (13). In JNCL, the
EEG shows disorganization and spike- and slow-wave com-
plexes. ERG and VEPs are abnormal early into the course of
the disease. The EEG abnormalities of ANCL are nonspecific
(14). In one case series, the frequency of background slowing
in all types of NCL was 94.6% and epileptiform discharges
were seen in 81.1% (15). Slow-frequency photic stimulation
elicited a response in 5 out of 22 patients, a giant SSEP was
seen in 7 of 25 patients, and slow or absent waveforms were
seen in 9 of 25 patients on VEP (15).
Mucopolysaccharidoses
The mucopolysaccharidoses (MPSs) are a heterogeneous
group of disorders caused by deficiency of a lysosomal enzyme
involved in the degradation of glycosaminoglycans (16). As
glycosaminoglycans accumulate in lysosomes, cell and organ
dysfunction result. Seven types exist. The age of onset and
severity vary, but most exhibit a chronic, progressive course
characterized by multisystemic involvement, coarse facies,
organomegaly, and bony defects (dysostosis multiplex) (16).
The more severe forms (MPS I, MPS II, and MPS VIII) result
in progressive mental retardation, whereas patients with MPS
IV and MPS VI have normal cognition, even when somatic
manifestations are severe (16). The only variant without
somatic symptoms is MPS III, of which there are four sub-
types. MPS III presents between 1 and 6 years of age with
delayed psychomotor development and behavioral problems.
Progressive dementia, sleeplessness, and epilepsy follow (17).
In MPS III, the EEG is normal in half of patients, but shows
diffuse slowing in the rest (18). Low-amplitude 12- to 15-Hz
activity is seen intermixed with generalized delta frequencies
(19). These findings correspond to the severe cortical involve-
ment seen pathologically (19). VEPs and BAEPs are almost
always normal (18). Enzyme replacement therapy and
hematopoietic stem cell transplantation offer potential benefit
in these conditions (20,21).
 Chapter 15 ■ EEG of Degenerative Disorders of the Central Nervous System
Sialidosis Types I and II (Neuraminidase
Deficiency)
In sialidosis, a deficiency of the enzyme neuraminidase leads to
progressive storage of sialidated glycopeptides and oligosaccha-
rides. Sialidosis bears clinical and histopathologic resemblance
to the mucopolysaccharidoses, in that patients exhibit mild
Hurler-like facies, skeletal dysplasia, and psychomotor retarda-
tion. Sialidosis is divided into types I and II (22). Type I has a
later age of onset and patients exhibit little if any dysmorphol-
ogy. Visual failure, cherry-red spots, ataxia, and myoclonus are
common features (23,24). Type II patients have an earlier age of
onset and demonstrate coarse facies, hepatosplenomegaly, and
skeletal anomalies (dysostosis multiplex and spinal deformities)
(25). MRI findings include atrophy of the cerebral hemispheres,
cerebellum, and pontine region (26).
EEG findings in sialidosis types I and II are indistinguishable
(27). In type I, the background is low voltage and rhythmic
spikes are observed over the vertex, which increase in frequency
during sleep (27). Spike-wave discharges precede myoclonic
jerking in both types I and II (28,29). VEPs demonstrate low
amplitudes and prolonged latencies in type I (28), while SEPs
exhibit giant potentials in both conditions (28,29).
Niemann–Pick Disease
Niemann–Pick disease occurs when a deficiency of the enzyme
acid sphingomyelinase (ASM) causes accumulation of sphin-
gomyelin within cells of the monocyte–macrophage system (30).
Four phenotypic subtypes exist. In Niemann–Pick disease type A
Figure 15.1 EEG in a 10-year-old boy with Niemann–Pick type C. Note the frontal slowing with intermixed sharp
waves bilaterally.
283
(NPD-A), infants present in the first few weeks or months of life
with failure to thrive and hepatomegaly. Psychomotor regression
follows, with head lag, inability to sit, and loss of interest in sur-
roundings. In time, rigidity and opisthotonus predominate. Death
occurs in the second or third year. Startle myoclonus, like that seen
in Tay–Sachs, is not observed and there are no coarse features, like
those present in mucopolysaccharidosis. In Niemann–Pick type B
(NPD-B), visceral involvement predominates without significant
neurologic impairment (31). An intermediate form between
NPD-A and NPD-B may also exist with retinal degeneration and
slowed nerve conduction presenting in adolescence or adulthood
(30). In Niemann–Pick type C (NPD-C), patients present between
ages 3 and 8 years with ataxia, vertical supranuclear gaze palsy, and
hepatosplenomegaly. In time, they experience dementia, dystonia,
and seizures (32). Cataplexy occurs late in the disease course (33).
Niemann–Pick type D (NPD-D) occurs in French Canadian
patients who share a common founder mutation (34).
In late childhood or adult-onset NPD-C, the EEG may reveal
generalized slowing (Fig. 15.1) (35). Late into the course of
NPD-C, abnormally high-voltage and diffuse alpha activity is
seen and is enhanced by intermittent photic stimulation. In a
patient with myoclonus, EEG–EMG frequency analysis demon-
strated a cortical origin for the myoclonus (36).
Gaucher Disease
Gaucher disease is the most common lysosomal storage
disorder. It results from a deficiency of the lysosomal enzyme
beta-glucocerebrosidase, resulting in accumulation of
284 Part III ■ Clinical EEG: General
glucocerebroside in the mononuclear phagocyte system (37).
The characteristic features include hepatosplenomegaly, bony
disease, and neurologic decline. Three phenotypic subtypes
exist. Type I has minimal if any neurologic impairment. Bulbar
symptoms predominate in type II patients (38). In type III
patients, ocular motor apraxia is often the initial finding.
The pathologic hallmark of the disorder is the Gaucher cell,
which is an enlarged macrophage with cytoplasmic linear inclu-
sions resulting from excessive lysosomal accumulation of gluco-
cerebroside (38). Enzyme replacement therapy with recombinant
glucocerebrosidase (Cerazyme) is the mainstay of treatment in
Gaucher disease (39), but is not effective for the neurologic
symptoms since it does not cross the blood–brain barrier.
MRI in type II disease is often normal, but cerebral atrophy is
commonly observed in type III patients (40). BAERs and SSEPs
can demonstrate abnormalities, even when no lesions are
detectable by MRI (41). EEGs show rhythmical sharp waves (6 to
10/sec), polyspikes, and spike-wave discharges in type III patients
(42). These features are diffuse, but have a posterior predomi-
nance. EEGs are commonly normal in type II patients. Reduced
amplitude can also be an EEG feature of some patients (41).
Fabry Disease
Fabry disease is the only X-linked lysosomal disorder; the oth-
ers are inherited in an autosomal recessive fashion. Hence, the
disease affects men more severely than it does women (43). It
results from a deficiency of alpha-galactosidase A. This leads to
storage of glycosphingolipids in a variety of tissues, producing
cellular dysfunction, inflammation, and fibrosis. Symptoms
begin in early childhood with peripheral neuropathy and burn-
ing pain of the hands and feet, which can remain disabling
throughout life (43). Hypohydrosis, nausea, and postprandial
diarrhea are also seen in childhood. By adulthood, proteinuria
and renal failure develop. Left ventricular hypertrophy and
coronary artery disease are also common. Angiokeratoma
develop on the skin. Neurologically, cryptogenic strokes occur.
Enzyme replacement therapy can clear microvascular endothe-
lial deposits from the kidneys, heart, and skin, reversing the
pathogenesis (44). MRI scans reveal gray and white matter
lesions. The anterior circulation generally appears normal, but
tortuous basilar arteries are common (45). EEGs can reveal
either focal or diffuse slowing (46).
Metachromatic Leukodystrophy
Metachromatic leukodystrophy (MLD) is characterized by an
inability to degrade sulfated glycolipids due to a deficiency of
the lysosomal enzyme arylsulfatase A. The disorder affects cen-
tral and peripheral myelin. There is widespread loss of myeli-
nated oligodendroglia in the CNS and segmental demyelination
of peripheral nerves (47). Clinically, it presents with motor
regression and ataxia at around 2 years of age (47) and leads to
a decerebrate state some years after the initial onset of symp-
toms. Epileptic seizures can occur late into the course of the ill-
ness. A late-onset form exists and presents in the third decade
with psychiatric symptoms.
Earlier into the course of MLD, EEGs are often normal. If
abnormalities are seen at presentation, they consist of mild
slowing of the background activity (48). As the clinical severity
of the disease progresses, so does the degree of slowing. Later
into the course of the illness, as the patients develop epileptic
seizures, focal spike waves appear. In the final stages of the dis-
ease course, hypersynchronous activity is seen (49). Both the
clinical and EEG features of the disease can improve following
bone marrow transplantation (50).
Globoid Leukodystrophy (Krabbe Disease)
Krabbe disease (KD) results from a deficiency of the lysosomal
enzyme galactosylceramidase. Severe loss of myelin occurs and
globoid cells are seen in the CNS white matter. Early infantile
KD accounts for 90% of all cases and presents in the first 6
months of life (51) with irritability and rapidly progressive
rigidity and tonic spasms. MRI of the brain shows symmetric
signal abnormalities in the periventricular regions of the poste-
rior cerebral hemispheres. Peripheral nerves are also affected,
but no visceromegaly is observed. The remaining 10% of
patients have late-infantile or juvenile KD, which presents with
ataxia and later exhibits dystonia and visual failure (52).
EEGs are abnormal in a majority of the early infantile cases,
but the opposite is true for late-infantile and juvenile-onset
patients. Abnormalities consist of slowing without epileptiform
activity in 46%, epileptiform activity without slowing in 15%,
and both slowing and epileptiform features in 38% of patients
(53). Likewise, BAEPs are abnormal in 88% of the early infan-
tile patients and only 40% of the late-onset cases. The most fre-
quent BAEP abnormality is absence of waves III and V (53).
PEROXISOMAL DISORDERS
Peroxisomal disorders are a group of heterogeneous conditions
that share dysfunction of peroxisomal function. Peroxisomes are
bound by a single membrane and contain a fine granular matrix.
They are histologically identified by the presence of catalase.
They are present in all human tissues except erythrocytes. They
carry out over 40 metabolic functions, including beta-oxidation
of very-long-chain fatty acids (VLCFA) and biosynthesis of plas-
malogen and cholesterol. Some of those reactions can occur in
other organelles, but many are exclusive to peroxisomes.
X-linked Adrenoleukodystrophy
X-linked adrenoleukodystrophy (X-ALD) is the most common
of the peroxisomal disorders. It is characterized by accumula-
tion of VLCFA in the brain and adrenal tissues (54). VLCFA are
also detected in fibroblasts, blood cells, and plasma, allowing
for identification of the condition (55). X-ALD affects the CNS
or adrenal glands and can have diverse presentations. The cere-
bral form presents between 4 and 8 years of age with symptoms
mimicking ADHD. Psychomotor regression and cortical blind-
ness follow and seizures are observed late into the disease
course. MRI shows lesions of the parieto-occipital white matter
with contrast enhancement at the leading edge (56). The mean
interval between the first symptom and an apparent vegetative
state is 2.4 years (48). Adrenomyeloneuropathy is a variant that
affects primarily the spinal cord and does not present until the
third decade. Since X-ALD is an X-linked condition, it presents
 Chapter 15 ■ EEG of Degenerative Disorders of the Central Nervous System
earlier and with greater severity in males, but heterozygous
women can be symptomatic.
During early stages of the disease, EEGs are normal in about
half of patients and show posterior slowing in the remainder. The
location of the slowing corresponds to the region of MRI abnor-
malities (48). As the disease progresses, the slowing becomes
more widespread and paroxysmal discharges are observed (48).
In one reported case, the ictal EEG during a tonic seizure revealed
an electrodecrement of the background activity (48).
Zellweger Syndrome
Zellweger syndrome (ZWS), also known as cerebrohepatorenal
syndrome, results from deficiencies of VLCFA beta-oxidation,
phytanic acid oxidation, and plasmalogen synthesis. As a result,
plasma levels of VLCFAs and phytanic acid are high, whereas
erythrocyte concentrations of plasmalogen are reduced. ZWS
presents in infancy with craniofacial dysmorphism (high fore-
head, large anterior fontanelle, separated cranial sutures,
hypoplastic supraorbital ridges, and epicanthal folds) and
hepatomegaly (57). Neurologic findings include hypotonia
with absent reflexes, profound developmental delays, seizures,
and impaired vision and hearing (57). MRI reveals both corti-
cal (polymicrogyria and pachygyria) and white matter (delayed
myelination) abnormalities (58).
Interictal EEGs in ZWS show bilateral independent multifo-
cal spike waves, predominantly in the frontal motor cortex (59).
Continuous negative vertex sharp and spike waves are com-
monly observed. They were found in 9 of 11 patients in one
study (60) and seen in a single case in a separate report (61).
This finding, observed during wakefulness and sleep, is pathog-
nomonic for ZWS (61). BAEPs and VEPs can have absent or
delayed responses (60).
Neonatal Adrenoleukodystrophy
Neonatal adrenoleukodystrophy (NALD) is a disorder of perox-
ysomal biosynthesis resulting from impaired ATPases that are
necessary for importing proteins into peroxisomes. NALD pres-
ents at birth with facial dysmorphism (midface hypoplasia)
(57). Hepatomegaly and adrenal cortical atrophy can also be
observed. Neurologically, patients present with hypotonia, optic
nerve atrophy, and seizures (62). Psychomotor regression
occurs in early childhood. MRI reveals a severe deficiency of
white matter in addition to heterotopias and polymicrogyria.
Interictal EEGs show high-voltage slow waves and bilateral
independent multifocal spike waves. Later in the disease course,
background suppression can be observed (59).
MITOCHONDRIAL DISORDERS
Mitochondrial diseases are a heterogeneous group of conditions
that result from dysfunction of the mitochondrial respiratory
chain. The mitochondria’s main function is generating ATP to
power the cells they reside in. Because mitochondria are present
in every cell type, with the sole exception of erythrocytes, mito-
chondrial dysfunction can have wide-ranging effects. The
mitochondrial genome contains 37 genes, of which 13 encode
structural proteins for the respiratory chain. The remaining
285
3000 or so genes are required to make a mitochondrion reside in
the nuclear DNA. Inheritance patterns are therefore varied with
both maternal and Mendelian patterns observed.
Leigh Syndrome
In 1951, Leigh described a case of subacute necrotizing
encephalomyelopathy in a 7-month-old infant. The patient had
symmetric lesions of the thalamus, midbrain, pons, medulla,
and posterior columns of the spinal cord (63). Leigh syndrome
results from defects in cytochrome oxidase or the pyruvate dehy-
drogenase complex. Inheritance patterns can be maternal, X-
linked, or autosomal recessive. Patients presenting prior to 12
months are likely to exhibit psychomotor retardation, vomiting,
weight loss, and weakness. In patients presenting after 18
months, symptoms include coma, ophthalmopareis, or nystag-
mus (64). Neuroimaging reveals abnormal signal in the basal
ganglia and brain stem with sparing of the mamillary bodies.
There is a strong predilection for the putamen; absence of puta-
minal involvement should call the diagnosis into question (65).
The interictal EEG can reveal focal or multifocal epilepti-
form activity, which becomes more frequent during sleep.
During partial seizures, the ictal EEG reveals posterior or hemi-
spheric background attenuation or ictal fast activity (66).
MELAS Syndrome (Mitochondrial Myopathy,
Encephalopathy, Lactic Acidosis, and Stroke)
MELAS is characterized by mitochondrial myopathy,
encephalopathy, lactic acidosis, and recurrent stroke-like
episodes. Many patients also exhibit seizures, migraines, corti-
cal blindness, short stature, or ragged-red fibers on muscle
biopsy (67). The mechanism of stroke in this condition remains
controversial. Some implicate a mitochondrial angiopathy,
resulting in small vessel occlusion, and others postulate that the
impairment of oxidative metabolism results in lactic acidosis
and brain injury (68). Imaging studies support the later
hypothesis (68,69). MRI abnormalities are most pronounced in
the parietal–occipital regions (69).
Likewise, interictal EEGs reveal diffuse sharp waves with an
occipital predominance. The epileptiform abnormalities can be
enhanced with photic stimulation (66). PLEDs can also be seen
over the affected occipital lobe (70). In one report of 11 patients
with a syndrome of overlapping MERRF–MELAS phenotype,
EEG abnormalities correlated with the severity of the clinical
phenotype. In milder cases, they included disorganization of
the background activity and focal or generalized delta–theta
waves (Fig. 15.2). More severely affected patients exhibited
bursts of generalized spike or polyspike and slow-wave com-
plexes (71). MELAS patients can develop epilepsia partialis con-
tinua. EEGs at those times reveal pseudoperiodic spike waves in
the contralateral temporal–occipital regions (72).
MERRF Syndrome (Myoclonus, Epilepsy,
Ragged-red Fibers)
MERRF is characterized by myoclonic epilepsy and myopathy
(with ragged-red fibers on muscle biopsy). Dementia, ataxia,
and generalized tonic–clonic seizures are also observed. It
results from an A-to-G mutation at nucleotide 8344 in the
286 Part III ■ Clinical EEG: General
Figure 15.2 EEG in a 13-year-old boy with MELAS syndrome and seizures. Note the continuous semirhythmic right
occipital slowing.
mitochondrial DNA (73). The mutation impairs mitochondrial
protein synthesis and thus impairs oxidative phosphorylation.
MERRF is transmitted by maternal inheritance, but the clinical
phenotype varies, even within the same family.
In one series, EEG abnormalities were seen in six of nine
patients. They included slowing of the background activity and
generalized epileptiform discharges (15). Polyspike- and spike-
and slow-wave complexes are enhanced by photic stimulation
and may occur in conjunction with myoclonic jerking (66).
Alpers–Huttenlocher Syndrome (Diffuse
Degeneration of Cerebral Gray Matter With
Hepatic Cirrhosis)
In 1931, Alpers reported the clinical features and neuropathology
of a 4-month-old girl with intractable generalized epilepsy (74).
He termed the condition “diffuse progressive degeneration of the
gray matter of the cerebrum.” The neuropathology is character-
ized by extensive brain atrophy with loss of cerebral neurons.
Psychomotor retardation and intractable epilepsy are universally
seen. Liver failure is common but sometimes absent. The syn-
drome is caused by mutations of the nuclear gene encoding the
catalytic subunit of mitochondrial polymerase gamma (POLG).
Mitochondrial DNA Depletion Syndrome
The EEG reveals slow or absent posteriorly dominant rhythms
(75). Interictal discharges can be focal, multifocal, or
generalized, but the occipital regions are most commonly
involved (75). Periodic (approximately 1 Hz) epileptiform dis-
charges can be seen (76) (Fig. 15.3) and high-amplitude slow-
ing with polyspikes is characteristic for the condition (77).
PROGRESSIVE MYOCLONIC EPILEPSIES
The progressive myoclonic epilepsies are a group of unrelated
disorders that share myoclonic epilepsy in common.
Generalized tonic–clonic, atypical absence, and simple and
complex partial seizures can be seen in these conditions. They
are generally refractory to medical management and worsen
over time. Cognitive impairment and ataxia are common
among these conditions. MERRF is usually included among the
progressive myoclonic epilepsies, but is discussed in the section
“Mitochondrial Disorders.”
Lafora Disease
Lafora disease (LD) generally presents between 12 and 17
years of age (78). Generalized tonic–clonic seizures are the
presenting symptom in 71% (15). Cognitive decline, visual
impairment, occipital seizures, and myoclonic seizures are
also observed. The occipital seizures can cause transient
blindness, visual hallucinations, or a photoconvulsive
response. Myoclonus worsens as the disease progresses.
Continuous generalized myoclonus can result in early wheel-
 Chapter 15 ■ EEG of Degenerative Disorders of the Central Nervous System
Figure 15.3 EEG in a 10-month-old boy with Alpers–Hunttenlocher syndrome. Note the periodic 1-Hz sharp waves
over the right frontal region.
chair dependency. Most patients die from neurologic compli-
cations within 10 years of disease onset (78). It is inherited in
an autosomal recessive fashion. The pathologic hallmark of
LD is the presence of Lafora bodies in neurons, liver, and mus-
cle. Lafora bodies stain positive with periodic-acid Schiff and
have a dense core with a less dense periphery. EEGs in LD
reveal background slowing in 97.4% of patients and general-
ized epileptiform discharges in 84.2% (15). Focal and multifo-
cal discharges can be observed. Photosensitivity with a fast
stimulus frequency is seen in 25%. EEG changes can precede
clinical symptoms by up to 6 years (79). As the disease pro-
gresses, the background rhythm slows and the alpha rhythm
and sleep morphologies are lost (78). Generalized irregular
spike-wave discharges with an occipital predominance domi-
nate the background activity. Unlike many primary general-
ized epilepsies, epileptiform discharges diminish during sleep
in LD. In one case series, giant SSEPs were seen in 24 of
31 patients and VEPs revealed prolonged or absent waveforms
in 12 of 31 patients (15).
Unverricht–Lundberg Disease
Unverricht–Lundberg disease (ULD) presents between 6 and 15
years of age with stimulus-sensitive myoclonus or generalized
tonic–clonic seizures (80). The myoclonus may be focal, multi-
focal, or generalized. It progresses over the course of the illness
and can be disabling. Tonic–clonic, absence, and complex
287
partial seizures can also increase over the course of the disease.
These symptoms are accompanied by ataxia, tremor, and
dysarthria. Cognitive dysfunction and visual impairment are
absent or mild (15). Although the condition clinically resembles
LD, Lafora bodies are not seen in ULD. It is inherited in an
autosomal recessive fashion and results from an unstable
expansion of a 12-nucleotide repeat within the CSTB gene on
chromosome 21q22.3 (81).
The EEG reveals generalized epileptiform discharges (15)
with a frequency of 3 to 5/sec. Photic stimulation elicits gener-
alized spikes and polyspikes. As the disease progresses, so does
the frequency of the epileptic discharges (80).
Severe Myoclonic Epilepsy of Infancy
(Dravet Syndrome)
Severe myoclonic epilepsy of infancy (SMEI) presents in the
first year of life with febrile seizures (82). The seizures are gen-
eralized or focal. Status epilepticus in the setting of fever is
common (82). The early age of onset helps distinguish it from
other forms of progressive myoclonic epilepsy. It is character-
ized by generalized tonic, clonic, tonic–clonic, absence, or
myoclonic seizures, which are often refractory to medications.
A ketogenic diet provides benefit in some patients (82). Ataxia
and psychomotor regression are common, although cognitive
development is normal prior to 1 year. The condition results
from a mutation of the sodium channel gene, SCN1A, which
288 Part III ■ Clinical EEG: General
also causes generalized epilepsy with febrile seizures plus
(GEFS ) syndrome (83). SMEI may therefore represent a vari-
ant of GEFS .
During the first 2 years of life, EEGs are normal (82). After
the second year, generalized spike and polyspike waves are
observed (82). The discharges increase in frequency during
sleep. Photic stimulation elicits discharges in about one fourth
of patients (82).
AMINOACIDEMIAS AND
ORGANIC ACIDEMIAS
Amino and organic acidemias are conditions that cause an
abnormal build-up of amino acids in the blood.
Aminoacidemias occur when there are elevated blood levels of
any amino acid, whereas organic acidemias refer specifically to
elevated levels of branched chain amino acids. These conditions
are generally diagnosed in infancy. They present with acute
metabolic decompensation, during which somnolence, hypoto-
nia, vomiting, and seizures are observed. Multisystemic involve-
ment is common. Untreated, these conditions can result in
coma and death. Treatment can include hemofiltration, dietary
restriction of particular nutrients or a class of nutrients, and
specific supplements designed to remove toxic metabolites.
Late-onset forms also exist and may present without overt
metabolic crisis. These conditions are diagnosed by chromatog-
raphy or mass spectroscopy. Laboratory investigations are best
performed during acute episodes.
Phenylketonuria
Phenylketonuria results from deficient activity of phenylalanine
hydroxylase (PAH), the enzyme that converts phenylalanine to
tyrosine. As a result, phenylalanine accumulates and is excreted
in the urine in large amounts as phenylketones. Phenylalanine
is neurotoxic and untreated patients develop mental retarda-
tion. Restricting phenylalanine in the diet can normalize
plasma phenylalanine levels and improve outcome. If instituted
in infancy, before the patient is symptomatic, mental retarda-
tion can be averted. For this reason, newborn screening has
been mandatory since the 1960s.
Some infants exhibit EEG abnormalities at the time of diag-
nosis, consisting of focal or multifocal spike and sharp waves
(84). Many of those same patients will have normal EEGs 1 year
after initiating dietary restriction of phenylalanine. Overall,
patients who begin treatment sooner are more likely to have
normal EEGs (85). Some patients with EEG abnormalities in
infancy will develop abnormalities later in childhood, even
though they remain on a restricted diet. EEG abnormalities in
such patients are indicative of higher phenylalanine levels (gen-
erally greater than 20 mg/dL) (86). Older patients and those who
initiate treatment later in childhood are more likely to exhibit
abnormalities on EEG (86). Focal epileptic discharges are the
most common abnormality observed (86).
Maple Syrup Urine Disease
In 1954, Menkes described four siblings with feeding difficulty,
irregular respirations, hypertonia, and opisthotonus presenting
in the first days of life. All of the infants had urine with the
smell of maple syrup (87). The condition, now termed maple
syrup urine disease, is caused by deficiency of branched chain
alpha-ketoacid dehydrogenase complex. This results in an ele-
vation of branched chain amino acids in the blood and urine.
Five clinical phenotypes of the condition exist (88). In the clas-
sic form, symptoms develop in the first week of life. Seizures are
common and coma and death can result if not treated early.
Initially, neuroimaging is normal, but does reveal generalized
edema in severe cases (89).
During acute metabolic decompensation, EEG reveals
spikes, polyspikes, spike-wave complexes, triphasic waves,
severe slowing, and bursts of periodic suppression (90). A char-
acteristic comb-like pattern of rolandic sharp waves is seen on
EEG between the second and third weeks and may persist after
the acute metabolic decompensation has passed (90). Spike-
wave discharges are seen with photic stimuli in just over half of
patients (90).
Homocystinuria
Homocystinuria is a metabolic disorder caused by a cystathio-
nine [beta]-synthase gene deficiency, resulting in increased
serum levels of homocysteine and methionine. Patients have
skeletal changes resembling Marfan syndrome. Cognitive
delays, psychiatric disturbances, strokes, and extrapyramidal
findings can result. Seizures occur in 21% of patients and are
usually generalized tonic–clonic (91).
Del Giudice et al. (92) reported EEG findings in 19 patients.
Ten of those patients had abnormalities, which generally con-
sisted of background slowing. Mid-temporal paroxysmal activ-
ity was seen in two patients. Centrotemporal spikes, increasing
in drowsiness and sleep, can occur and may mimic benign
childhood epilepsy with centrotemporal spikes (BECTS) (93).
Propionic Acidemia
Propionic acidemia results from a deficiency of the mitochon-
drial enzyme propionyl-CoA carboxylase, which converts pro-
pionyl-CoA to D-methylmalonyl CoA. Biotin is a cofactor for
this reaction. Most patients present as neonates with an acute
metabolic crisis with hypotonia, lethargy, or seizures. The
seizures may result from NMDA receptor-mediated mecha-
nisms. Later onset patients present with developmental delays
without metabolic decompensation, acute encephalopathy, or
episodic ketoacidosis (94). Acute metabolic crises can be treated
with total protein restriction and hemofiltration (peritoneal
dialysis and exchange transfusion). Carnitine supplementation
and antibiotics (to reduce intestinal production of propionic
acid by normal gut flora) are also used. Sass et al. (94) reported
EEG findings in 32 patients. Of those, nine showed moderate
abnormalities and six were clearly abnormal. The remaining 17
patients had normal EEGs. A separate case report (95) revealed
disorganized background activity at the time of acute metabolic
decompensation.
Methylmalonic Acidemia
Methylmalonic acidemia (MMA) is a heterogeneous group of
conditions characterized by accumulation of methylmalonic
 Chapter 15 ■ EEG of Degenerative Disorders of the Central Nervous System
acid and its byproducts. It results from impairments of methyl-
malonate or cobalamin (vitamin B12) metabolism. Hence,
some patients respond to cobalamin supplementation and oth-
ers do not. Patients with defects of cobalamin metabolism have
simultaneous homocystinuria.
MMA typically presents in the first weeks of life with vomit-
ing, lethargy, hypotonia, and seizures. Later onset cases also
exist and may present with hypoglycemia, myopathy, or throm-
bosis (due to homocystinuria). In one case report of an infant
with MMA and tonic seizures, EEG showed polyspike bursts
(96). Another report described sharp transients superimposed
on a low-voltage and discontinuous tracing with long flat peri-
ods. Following treatment with hydroxycobalamin (vitamin
B12), blood homocysteine levels declined and a reduction was
seen in the length of the interburst periods (97).
Isovaleric Acidemia
Isovaleric acidemia is caused by a deficiency of the mitochon-
drial enzyme isovaleryl-CoA dehydrogenase, resulting in accu-
mulation of isovaleryl-CoA derivatives. It generally presents in
infancy with poor feeding, vomiting, obtundation, and seizures.
The affected infants have a characteristic odor, which resembles
“dirty socks” and is best appreciated in sweat and cerumen.
Patients presenting after the newborn period exhibit failure to
thrive, developmental delays, or mental retardation. The initial
therapeutic approach is to decrease leucine intake and promote
an anabolic state with increased calories. EEG may reveal low-
voltage activity (less than 20 V) with bursts of delta slowing or
sharp waves (98). In an experimental animal model, infusion of
isovaleric acid causes slowing on EEG, which worsens as the
concentration increases (99).
Glycine Encephalopathy (Nonketotic
Hyperglycinemia)
Glycine encephalopathy is caused by dysfunction of the mul-
timeric glycine cleavage enzyme system, resulting in accumu-
lation of glycine in all body tissues, including the brain. The
disorder presents in the first week of life with apnea, lethargy,
hypotonia, myoclonus, seizures, and feeding difficulty (100).
Brain imaging is normal in half of cases. When abnormalities
are seen, they include agenesis of the corpus callosum, pro-
gressive atrophy, or delayed myelination (100). Diagnosis is
confirmed by the detection of elevated cerebral spinal fluid
(CSF) glycine levels and an elevation of the CSF-to-plasma
glycine ratio (101). Classic neonatal-onset patients have ratios
greater than 0.2, but atypical, later onset cases also exist, and
have ratios of roughly 0.09 (101). Atypical presentations can
include seizures, developmental delays, and movement disor-
ders outside of the neonatal period. In classical neonatal pre-
sentations, EEG demonstrates a burst suppression pattern or
hypsarrhythmia (100). In later onset variants, the EEG can
reveal generalized spike and slow waves (102). Since glycine
can stimulate NMDA glutamate receptors, treatment with
NMDA antagonists (dextromethorphan and ketamine) has
been tried. Sodium benzoate decreases CSF glycine levels and
can further improve seizure control and alertness (103). In
one case, clinical and electrophysiologic improvements were
289
seen with a combined regimen of dextromethorphan and
sodium benzoate (103).
Hartnup Disease
Hartnup disease is an autosomal dominant disorder of amino
acid transport within the renal tubules and intestinal tract. It
results in malabsorption of neutral monoamino monocar-
boxylic amino acids and aminoaciduria. Symptoms include pel-
lagroid dermatitis, photosensitivity, ataxia, and neuropsy-
chiatric abnormalities. EEG can reveal epileptiform features
(104), but is normal in most cases.
Lowe Syndrome (Oculocerebrorenal Syndrome)
Lowe syndrome was originally described in 1952 by Lowe et al.
(105), who reported the histories of three unrelated male
infants with congenital cataracts, glaucoma, developmental
retardation, hyporeflexia with low musculature, osteopenia,
decreased ammonia production, and a “peculiar, high-pitched,
irritating cry.” Affected children also exhibit hyperactivity,
pseudotumor cerebri, ophthalmoplegia, or hypoglycemia. The
condition is X-linked and is caused by mutations of the OCRL1
gene on chromosome Xq26. OCRL1 encodes a protein product
that plays a role in cytoskeletal remodeling and cellular traffick-
ing. Lowe syndrome has a high mortality rate in the first
months from metabolic derangements and metabolic acidosis.
Most patients will survive into the second or third decade, but
will develop chronic renal disease. MRI reveals small cyst-like
changes in the periventricular white matter and H-MRS
demonstrates prominent myoinositol peaks indicating gliosis
(106,107). EEG shows an irregular, poorly sustained posteriorly
dominant rhythm and bursts of high voltage slowing.
Independent bitemporal spike- and slow-wave discharges are
seen (108). Generalized tonic–clonic, tonic, atonic, and com-
plex partial seizures have been described. In one report of a
patient with atonic seizures, the ictal EEG revealed secondarily
generalized epileptiform discharges (109).
INBORN ERRORS OF UREA SYNTHESIS
Urea cycle disorders occur when there is deficiency of one of the
enzymes in the cycle that converts nitrogen to urea. There are
six enzymes involved: carbamyl phosphate synthetase I (CPSI),
ornithine transcarbamylase (OTC), arginosuccinate synthetase
(ASS), argininosuccinate lyase (ASL), N-acetyl glutamate syn-
thetase (NAGS), and arginase. Dysfunction of any of the
enzymes, with the exception of arginase, results in hyperam-
monemia and metabolic decompensation in infancy. Newborns
appear well in the first 24 to 48 hours of life. After the infant
begins taking protein from formula or breast milk, vomiting,
lethargy, irritability, and poor feeding develop (110). Cerebral
edema results in hyperventilation and metabolic alkalosis
(111). Half of the affected patients have seizures.
Patients with partial enzyme deficiencies have later onset
presentations. This occurs most commonly in female carries of
OTC, since OTC is an X-linked condition. Late-onset patients
exhibit developmental delays, seizures, psychiatric disorders,
and chronic vomiting (112). Following high protein loads or
290 Part III ■ Clinical EEG: General
catabolic stress, headache, vomiting, and ataxia develop. The
diagnosis is suspected when an elevated plasma ammonia level
is seen in conjunction with a normal blood glucose and a nor-
mal anion gap. The diagnosis is confirmed by performing
enzyme analysis from liver, fibroblasts, or red blood cells.
In one study of four infants with urea cycle disorders
(113), all of the patients demonstrated abnormalities on at
least one EEG. The most common findings included multifo-
cal spike- and spike- and slow-wave discharges. Clancy and
Chung (114) reported burst suppression in cases of neonatal
citrullinemia. They observed that the length of the EEG inter-
burst interval correlated with serum levels of ammonia, sug-
gesting that hyperammonemia may directly contribute to
encephalopathy. Nagata et al. (115) noted that EEG abnor-
malities were more often observed in patients with CT abnor-
malities, and that patients with a normal CT were less likely
to exhibit EEG findings.
DISORDERS OF HEAVY METAL
ACCUMULATION
Neurodegeneration w ith Brain Iron Accumulation
Neurodegeneration with brain iron accumulation (NBIA) is a
collection of phenotypically similar conditions, which share in
common brain iron deposition. Included within this heading
are pantothenate kinase-associated neurodegeneration
(PKAN), infantile neuroaxonal dystrophy (INAD), neurofer-
ritinopathy, and aceruloplasminemia. Patients present with
posture and gait abnormalities, bradykinesia, and rigidity
(116). MRI distinguishes these disorders, but all share in com-
mon a hypointense globus pallidus on T2 sequencing (117).
More definitive confirmation can be accomplished through
genetic testing.
PKAN is caused by a mutation of the pantothenate kinase 2
gene (PANK2). It is phenotypically classified into classical and
atypical forms. Classical PKAN presents around age 3 with
clumsiness and gait abnormalities. Dystonia begins in the lower
extremities. Dysarthria, spasticity, visual impairment, and
behavioral problems follow. Pigmentary retinopathy is com-
mon (116). MRI reveals T2 hyperintense signal surrounded by
a rim of hypointensity in the globus pallidus, creating the “eye-
of-the-tiger” sign, which is pathognomonic for the condition
(116). Atypical presentations have a later onset and are charac-
terized by psychiatric symptoms and movement disorders. The
EEG in PKAN is normal in some cases (118,119). Observed
abnormalities can include low voltage slowing and fast activity
(119). Swaiman et al. (119) reported a single patient with
seizures and bursts of generalized slowing and sharp waves on
EEG. Bilateral multifocal spikes and excessive posterior slowing
were also noted.
Infantile Neuroaxonal Dystrophy
(Seitelberger Disease)
INAD is characterized pathologically by axonal swelling and
spheroid bodies in the central nervous system. Onset is between
6 months and 2 years. Patients present with neuromotor delay
including lack of head control, inability to sit or walk, and poor
eye contact. Myoclonic seizures occur. Dysmorphism (microg-
nathia, a prominent forehead, and low-set ears) is sometimes
observed (121). INAD is distinguished from PKAN by the ear-
lier age of onset and greater pyramidal signs (spasticity and
hyperreflexia) (122). Like PKAN, however, INAD is associated
with high iron levels in the globus pallidus on MRI (123).
Cerebellar atrophy is also observed. The two conditions are
genetically distinct; INAD is caused by mutations of the
PLA2G6 gene. Death usually occurs by age 10.
Interictal EEG reveals high-voltage fast rhythms (16 to 22
Hz) in sleep and wakefulness. Diffuse slow spike-wave com-
plexes are also observed (124). In a 4-year-old boy with tonic
spasms, the ictal EEG revealed diffuse, high-voltage, irregular
sharp and slow-wave complexes followed by desynchronization
(124). Visual evoked potentials are abnormal and EMG demon-
strates chronic degeneration.
Wilson Disease (Hepatolenticular Degeneration)
In Wilson disease, a defect of copper transport and excretion
leads to copper accumulation in the liver. As cirrhosis develops,
copper leaks into plasma and accumulates in the brain, kidneys,
and cornea where it results in injury. The earliest pathology in
Wilson disease occurs in the liver; the initial deposition of
hepatic copper is clinically silent.
There is wide variability in the age of disease onset, but the
mean is 16 years (125). Younger patients tend to present with
hepatic disease and older patients are more likely to exhibit
neuropsychiatric symptoms at presentation (126). In one
review of 282 patients (124), 69.1% presented with neurologic
symptoms. Of those, the predominant features were parkin-
sonism (62.3%), dystonia (35.4%), cerebellar dysfunction
(28%), pyramidal signs (16%), chorea (9%), athetosis (2.2%),
myoclonus (3.4%), and behavioral abnormalities (16%). All of
the patients with neurologic symptoms had Kayser–Fleischer
rings on ophthalmologic assessment. Serum ceruloplasmin was
decreased in 93% and 24-hour copper excretion was increased
in 70%. MRI reveals bilateral T2 hyperintensities within the
basal ganglia and thalamus as well as mild to moderate atrophy.
EEG changes in Wilson disease include a reduction in alpha
frequency (17.6%), and increased beta (14.7%), theta (20.5%),
or delta (2.9%) activity. Focal epileptiform discharges are seen
in 11.7% of patients. In one series of 23 patients, only 3 exhib-
ited EEG abnormalities (127), while in a literature review of 80
patients, over half showed EEG abnormalities (128). In general,
the degree of EEG abnormality parallels the severity of clinical
involvement. The EEG may improve during or after treatment
(128). Prolonged latencies and abnormal waveforms are
observed with visual and auditory evoked potentials.
Menkes Disease (Kinky Hair Disease)
Menkes disease (MD) is a fatal X-linked condition caused by a
mutation of the copper-transport gene ATP7A, which is neces-
sary for copper excretion from enterocytes. Copper deficiency
results in impaired function of copper-dependent enzymes.
The neurologic symptoms result mainly from deficiency of
cytochrome c oxidase. The clinical features can therefore
 Chapter 15 ■ EEG of Degenerative Disorders of the Central Nervous System
resemble those in Leigh disease. Dysfunction of lysyl oxidase
leads to connective tissue abnormalities and predisposes
patients to arterial aneurysms and intracranial hemorrhage.
Tyrosinase deficiency accounts for hypopigmentation of hair
and skin. Patients typically present at 2 to 3 months of age with
loss of developmental milestones, truncal hypotonia, failure to
thrive, temperature instability, and epilepsy. The physical exam-
ination is notable for abnormally kinky hair, which is shorter
and sparser over the sides and back. Skin and hair are hypopig-
mented. Minor dysmorphology, including sagging cheeks, a
depressed nasal bridge, and high arched palate are noted. MRI
reveals intracranial vessel tortuosity, white matter changes, and
ischemic lesions of the deep gray matter (129). Treatment with
copper replacement may improve clinical outcomes if started
early in the course of the disease (130).
Bahi-Buisson et al. (131) classified three stages of MD based
on seizure types and EEG findings. In early MD, patients
develop focal clonic status epilepticus, which is accompanied on
EEG by slow spike waves in the posterior head regions. Interictal
EEG reveals multifocal polymorphic slow waves. During the
intermediate stage of the disease, infantile spasms occur and a
modified hypsarrhythmic pattern is seen on EEG. In late-stage
MD, patients exhibit multifocal seizures, tonic spasms, and
myoclonus. EEG reveals multifocal high-amplitude activity
mixed with irregular slow waves (Fig. 15.4) (131). The degree of
EEG abnormality relates in part to serum copper levels (132).
Figure 15.4 EEG in a 9-month-old boy with Menkes disease. Note the right temporal delta slowing and spike-
and slow-wave discharges.
291
OTHER METABOLIC CONDITIONS
Glut-1 Deficiency Syndrome (De Vivo Syndrome)
Glut-1 is expressed on vascular endothelial cells and is the
major transporter of glucose into the central nervous system.
Dysfunction of Glut-1 results in low CSF glucose levels in the
absence of hypoglycemia. Patients present in infancy with pro-
gressive microcephaly, developmental delays, and seizures
(133). Atypical forms exhibit ataxia or mental retardation with-
out epilepsy.
Seizure types include generalized tonic, clonic, myoclonic,
atonic, and atypical absence. In infancy, the EEG reveals multi-
focal spike-wave discharges, which localize to the temporal and
occipital regions. With maturation, generalized 2.5- to 4-Hz
spike- and slow-wave discharges are observed (133). The
seizures can worsen with phenobarbital, since it inhibits Glut-1
function. The ketogenic diet is the only effective therapy since it
allows the brain to utilize ketone bodies as a fuel source.
Pyridoxine-Dependent Seizures
Pyridoxine-dependent seizures (PDS) are a rare autosomal
recessive condition, presenting with neonatal epilepsy. A variety
of seizure types, including infantile spasms, is observed. The
seizures are resistant to traditional anticonvulsants, but are well
controlled by a daily pharmacologic dose of pyridoxine (vitamin
B6). Patients with PDS require pyridoxine supplementation
292 Part III ■ Clinical EEG: General
throughout life. In a related condition, pyridoxine-responsive
seizures, patients are able to successfully wean off of pyridoxine
without seizure recurrence. PDS is caused by deficiency of the
enzyme -aminoadipic semialdehyde ( -AASA) dehydroge-
nase, which is encoded by the antiquitin gene (ALDH7A1). -
AASA acts within the lysine catabolism pathway (134). Elevated
levels of -AASA and pipecolic acid (PA) result, and can serve as
biomarkers for the condition.
In PDS, the interictal EEG is abnormal in both the awake
and asleep states. The background is discontinuous with a sup-
pression burst-like pattern. Spike and polyspike waves are seen
over the central and temporal regions. Continuous high-voltage
rhythmic delta slowing occurs (135). Administration of pyri-
doxine (100 to 500 mg intravenously) results in rapid normal-
ization of the EEG, but should be given under close supervision
since it may result in apnea and bradycardia. Folinic acid sup-
plementation and lysine restriction may also play a role in treat-
ment of PDS.
Folinic Acid-Responsive Seizures
Folinic acid-responsive seizures present with myoclonic, clonic,
or subtle (apnea and irritability) seizures in infancy. The condi-
tion is confirmed by two characteristic peaks of CSF
monoamine metabolic analysis. The identity of the peaks
remains unknown. MRI reveals white matter abnormalities and
atrophy. EEG demonstrates a discontinuous background pat-
tern with multifocal spike or sharp waves (135). In one report
of two patients with folinic acid-responsive seizures, both were
found to have increased CSF -AASA and PA (137). Those
patients were subsequently found to have mutations of the
ALDH7A1 gene, indicating that folinic acid-responsive seizures
and PDS may be the same condition.
CEREBELLAR DISORDERS
Spinocerebellar Ataxias
Spinocerebellar ataxias (SCAs) are a heterogeneous group of
progressive cerebellar syndromes with varying combinations of
ataxia, dysmetria, and oculomotor deficits (138). Additional
clinical findings include pigmentary retinopathy, extrapyrami-
dal movement disorders (parkinsonism, dystonia, and chorea),
pyramidal signs, dementia, seizures, and peripheral neuropathy
(138). MRI reveals cerebellar, olivopontocerebellar, or global
brain atrophy. The SCAs are genetically heterogeneous autoso-
mal dominant conditions. A gene-based classification scheme is
used, in which the numbering system represents the order of
gene discovery (SCA 1 to 8, SCA 10 to 23, and SCA 25). Genes
for fibroblast growth factor 14 (FGF14) and dentatorubral-pal-
lidoluysian atrophy (DRPLA) are also included within the fam-
ily of SCAs. Many forms result from trinucleotide repeat
disorders. Anticipation, progressively worsening phenotypes of
subsequent generations within the same family, is observed.
The age of onset and response to symptomatic treatment differ
among the various types of SCA (138).
EEG findings in SCA vary among the different types. Variants
that are commonly associated with epilepsy are more likely to
reveal EEG abnormalities. In one report of 18 patients with SCA,
temporal theta and delta intrusions were noted (139). Epilepsy
is common in SCA10, and a case series of 18 patients found EEG
abnormalities in all 18 (140). The most common finding was
slow and disorganized background activity. A case report of a
father and son with complex partial seizures and SCA2 noted
temporal epileptiform discharges and slowing in both patients
(141). Patients with DRPLA also develop epilepsy, and may
experience complex partial, myoclonic, atypical absence, and
generalized tonic–clonic seizures. In 63.6% of patients with
DRPLA, the EEG shows epileptiform features. Photosensitivity
was observed in 27.3%, and those patients with a photoparoxys-
mal response also had progressive myoclonic epilepsy (142).
Friedreich Ataxia
Friedreich ataxia (FRDA1) is an autosomal recessive trinu-
cleotide repeat disorder of the frataxin gene on chromosome
9p. The frataxin gene acts to regulate mitochondrial iron con-
tent. Iron overload results in damage to respiratory chain com-
plexes I, II, and III through generation of oxygen free radicals
(143). FRDA1 therefore has features of a mitochondrial disor-
der. It presents prior to adolescence with progressive cerebellar
dysfunction and hypoactive knee and ankle reflexes. Patients
also exhibit dysarthria, nystagmus, impaired position and
vibratory sensation, and a Babinski sign. Cardiac dysfunction
(subaortic stenosis and hypertrophic cardiomyopathy) and dia-
betes are common medical complications and one half of
patients die of heart failure (144).
No qualitative EEG pattern is seen in FRDA1 (144). EEG also
fails to provide prognostic information, since there is no rela-
tionship between the degree of EEG abnormality and the sever-
ity of the disease.
GENETIC CONDITIONS
Rett Syndrome
The clinical features of Rett syndrome were defined by a con-
sensus panel held in 2001 (146). Necessary criteria for the diag-
nosis include female sex, normal development through the first
6 months, deceleration of head growth, psychomotor regres-
sion, loss of purposeful hand movements, and hand-wringing
stereotypies. Pathologically, Rett syndrome is characterized by
reduced dendritic arborization. Seventy percent of patients
who meet clinical criteria carry a mutation of the X-linked
methyl CpG-binding protein (MECP2) gene (147), which is
critical for postnatal neuronal maturation. The X-linked cyclin-
dependent kinase-like 5 (CDKL5) gene can also result in a Rett
syndrome (RTT)-like phenotype. Patients with CDKL5 muta-
tions are less likely to exhibit normal development in the first
year and more likely to have early onset intractable epilepsy,
infantile spasms, and severe hypotonia (148).
EEG criteria for Rett syndrome have been proposed (149,150)
and match the clinical stages of the disease (151). Stage I occurs
between 6 and 18 months of age and is characterized by develop-
mental arrest and decelerating head growth. The EEG is normal
or slightly slow and epileptiform activity is absent. In stage II, the
 Chapter 15 ■ EEG of Degenerative Disorders of the Central Nervous System
patient exhibits further developmental deterioration with loss of
hand skills, severe dementia, and seizures. On EEG, there is slow-
ing or loss of the occipitally dominant rhythm and central spike
waves are seen during sleep (Fig. 15.5). During stage III, patients
stabilize, but continue to have psychomotor delay and epilepsy.
The EEG reveals multifocal epileptiform discharges and general-
ized spike and slow waves with an absence of the posteriorly
dominant rhythm. Stage IV occurs after 8 years of age and
involves decreasing mobility and loss of ambulation. Seizures
become less prominent. The EEG may reveal continuous, gener-
alized, slow spike-wave activity in sleep. Rhythmic frontal theta
activity can be seen in wakefulness and drowsiness.
CDKL5 patients more commonly have normal interictal
EEGs. In a study of 20 CDKL5 patients, 15 had a normal inter-
ictal EEGs at onset and the other 5 revealed only background
slowing (148). In CDKL5 patients with intractable seizures,
EEG abnormalities include high-amplitude delta waves with
bursts of pseudoperiodic high-amplitude spikes, polyspikes, or
spike and slow waves (148).
Angelman Syndrome
Angelman syndrome is characterized by developmental delay,
speech and motor impairment, ataxia, epilepsy, and an appar-
ently happy demeanor (152). It is caused by a lack of expression
of the UBE3A gene resulting from abnormalities of chromo-
some 15q11-q13. Approximately 70% of Angelman syndrome
cases result from a de novo 15q11-q13 deletion of the maternal
Figure 15.5 EEG in a 3-year-old girl with Rett syndrome. Note the bilateral central spike waves.
293
chromosome. Five percent to 10% of cases result from a muta-
tion of the maternal UBE3A gene. Three percent to 5% have an
abnormality of DNA methylation in the maternal copy of the
gene. Two percent to 3% of patients inherited both copies of
chromosome 15 from the father and none from the mother
and, therefore, have no functional copy of UBE3A from the
mother (152). Abnormalities of the same region on 15q11-q13
also result in Prader–Willi syndrome. The phenotype is deter-
mined by the parental origin of the chromosome 15 defect with
maternal abnormalities resulting in Angelman syndrome and
paternal abnormalities leading to Prader–Willi. Hence, these
conditions illustrate the phenomenon of genomic imprinting.
Ninety percent of patients with Angelman syndrome have
seizures. Epilepsy is generally more severe in patients with a
deletion of the 15q11-q13 region, which includes encoding
information for GABAa receptor subunits. Epilepsy begins
between 1 and 3 years in most patients. Seizure types include
epileptic spasms, atypical absence, myoclonic, tonic, atonic, and
tonic–clonic. Multiple seizure types occur in approximately half
of patients with a 15q11-q13 deletion. The EEG resembles hyp-
sarrhythmia and demonstrates runs of rhythmic 2- to 3-Hz
high-amplitude activity (greater than 300 mV) over the frontal
regions. Multifocal spikes or sharp waves can be intermixed
(153). This pattern is distinguished from hypsarrhythmia by a
lack of fragmentation during sleep. Rhythmic runs of slow
spike and slow waves occur and may indicate nonconvulsive
status epilepticus (152) (Fig. 15.6). This pattern can mimic the
294 Part III ■ Clinical EEG: General
Figure 15.6 EEG in a 3-year-old boy with Angelman syndrome. Note the continuous, rhythmic, bifrontally predominant
delta slowing with intermixed sharp features.
changes seen in Lennox–Gastaut syndrome, but the clinical
phenotype helps distinguish between these conditions. Tremor
is common and correlates with 4- to 10-Hz electromyographic
bursts, but no EEG changes are seen, indicating that the tremor
is not due to seizure activity (154). Likewise, bouts of laughter
do not have any accompanying EEG changes and are unlikely to
represent gelastic seizures (152).
Dow n Syndrome (Trisomy 21)
Down syndrome is a chromosomal disorder caused by all or
part of an extra 21st chromosome. It is characterized by intel-
lectual impairment, short stature, and facial dysmorphism.
Hypotonia is seen on the neurologic examination. Older
patients develop Alzheimer-type dementia. Other medical fea-
tures include congenital heart disease, obstructive sleep apnea,
frequent bouts of otitis media, and thyroid dysfunction.
Epileptic seizures are more prevalent in patients with Down
syndrome than in the general population. All major seizure
types have been reported, but infantile spasms are particularly
common (155). The EEG shows hypsarrhythmia, which can
resolve following administration of diazepam. Many patients go
on to develop other seizure types following infantile spasms,
including myoclonic, atonic, tonic–clonic, or absence (156).
Older Down syndrome patients have increased risks of epilepsy
and dementia. Myoclonic epilepsy associated with cognitive
decline, termed senile myoclonic epilepsy, occurs late into the
condition. Myoclonus is most common in the early morning
hours. The EEG reveals generalized spike-wave discharges asso-
ciated with bilateral myoclonic jerks, whereas segmental jerks
can occur without accompanying EEG changes (157). Spectral
analysis of EEG reveals reduced alpha power in the posterior
cortical regions and diffuse EEG slowing (158).
REFERENCES
1. Fukumizu M, Yoshikawa H, Takashima S, et al. Tay Sachs disease:
progression of changes on neuroimaging in four cases.
Neuroradiology. 1992;34:483–486.
2. Pampiglione G, Harden A. Neurophysiological investigations in
GM1 and GM2 gangliosidoses. Neuropediatrics. 1984;15:74–84.
3. Kellaway P, Petersen I, eds. Clinical Electroencephalography in
Children. New York and London: Grune and Stratton; 1968:287–306.
4. Cooperhaver RM, Goodman G. The electroretinogram in infan-
tile, late infantile, and juvenile amaurotic familial infancy. Arch
Ophthalmol. 1960;63:559–566.
5. Sakuraba H, Itoh K, Shimmoto M, et al. GM2 gangliosidosis AB
variant: clinical and biochemical studies of a Japanese patient.
Neurology. 1999;52:372.
6. Zaroff CM, Neudorfer O, Morrison C, et al. Neuropsychological
assessment of patients with late onset GM2 gangliosidosis.
Neurology. 2004;62:2283.
 Chapter 15 ■ EEG of Degenerative Disorders of the Central Nervous System
7. Harden A, Martinovic Z, Pampiglione G. Neurophysiological
studies in GM1 gangliosidosis. Ital J Neurol Sci. 1982;3:201–206.
8. Cooper JD. Progress towards understanding the neurobiology of
Batten disease or neuronal ceroid lipofuscinosis. Curr Opin Neurol.
2003;16:121.
9. Vinken PJ, Bruyn GW, eds. Handbook of Neurology. Vol 10.
Amsterdam: North Holland Publishers; 1970:588–679.
10. Wisniewski K. Neuronal ceroid-lipofuscinoses. Gene Rev. 2006:
1–26.
11. Groebel HH, Mole SE, Lake BD, eds. The Neuronal Ceroid
Lipofuscinoses (Batten Disease). Amsterdam: IOS Press; 1999:16–36.
12. Santavuori P, Lauronen L, Kirveskari E, et al. Neuronal ceroid lipo-
fuscinoses in childhood. J Neurol Sci. 2000;21:s35–s41.
13. Groebel HH, Mole SE, Lake BD, eds. The Neuronal Ceroid
Lipofuscinoses (Batten Disease). Amsterdam: IOS Press; 1999:37–54.
14. Goebel H, Wisniewski KE. Current state of clinical and morpho-
logical features of human NCL. Brain Pathol. 2004;14:61–69.
15. Sinha S, Satishchandra P, Gayathri N, et al. Progressive myoclonic
epilepsy: a clinical, electrophysiological and pathological study
from South India. J Neurol Sci. 2007;252:16–23.
16. Muenzer J. The mucopolysaccharidoses: a heterogeneous group of
disorders with variable pediatric presentations. J Pediatr. 2004;
I44:s27–s34.
17. Ruijter GJG, Valstar MJ, van de Kamp JM, et al. Clinical and
genetic spectrum of Sanfilippo type C (MPS IIIC) disease in the
Netherlands. Mol Genet Metab. 2008;93:104–111.
18. Husain AM, Escolar ML, Kurtzberg J. Neurological assessment
of mucopolysaccharidosis III. Clin Neurophysiol. 2006;117:
2059–2063.
19. Kriel RL, Hauser WA, Sung JH, et al. Arch Neurol. 1978;35:838–843.
20. Tolar J, Gewal SS, Bjoraker KJ, et al. Combination of enzyme
replacement and hematopoietic stem cell transplantation as therapy
for Hurler syndrome. Bone Marrow Transplant. 2008;41:531–535.
21. Wrait JE, Clarke LA, Beck M, et al. Enzyme replacement therapy
for mucopolysaccharidosis I: a randomized, double-blinded,
placebo-controlled, multinational study of recombinant human
alpha-L-iduronidase (laronidase). J Pediatr. 2004;144:581–588.
22. Lowden JA, O’Brien JS. Sialidosis: a review of human neu-
raminidase deficiency. Am J Hum Genet. 1979;31:1–18.
23. Thomas PK, Abrams JD, Swallow D, et al. Sialidosis type I: cherry-
red spot myoclonus syndrome with sialidase deficiency and altered
electrophoretic mobilities of some enzymes known to be glyco-
proteins. J Neurol Neurosurg Psychiat. 1979;42:873–880.
24. Swallow DM, Evans L, Stewart G, et al. Sialidosis type I: cherry red
spot-myoclonus syndrome with sialidase deficiency and altered
electrophoretic mobility of some enzymes known to be glycopro-
teins. Ann Hum Genet. 1979;43:27–35.
25. Kelly TE, Graetz G. Isolated acid neuraminidase deficiency: a dis-
tinct lysosomal storage disease. Am J Med Genet. 1977;1:31–46.
26. Palmeri S, Villanova M, Malandrini A, et al. Type I sialidosis: a
clinical, biochemical, and neuroradiological study. Eur Neurol.
2000;43:88–94.
27. Engel J, Rapin I, Giblin DR. Electrophysiological studies in two
patients with cherry red spot-myoclonus syndrome. Epilepsia.
1977;18:73–87.
28. Louboutin JP, Nogues B, Caillaud C, et al. Multimodality evoked
potentials and EEG in a case of cherry red spot-myoclonus syn-
drome and alpha-neuraminidase deficiency (sialidosis type I). Eur
Neurol. 1995;35:175–177.
29. Tobimatsu S, Fukui R, Shibasaki H, et al. Electrophysiological
studies of myoclonus in sialidosis type 2. Electroencephalogr Clin
Neurophysiol. 1985;60:16–22.
295
30. Wasserstein MP, Aron A, Brodie SE, et al. Acid sphingomyelinase
deficiency: prevalence and characterization of an intermediate
phenotype of Niemann–Pick disease. J Pediatr. 2006;149:554–559.
31. Kolodney EH. Niemann–Pick disease. Curr Opin Hematol.
2000;7:48.
32. Vanier MT, Millat G. Niemann–Pick disease type C. Clin Genet.
2003;64:269–281.
33. Vankova J, Stepanova I, Jech R, et al. Sleep disturbances and
hypocretin deficiency in Niemann–Pick disease type C. Sleep.
2003;15:427–430.
34. Greer WL, Riddell DC, Murty S, et al. Linkage disequilibrium
mapping of the Nova Scotia variant of Niemann–Pick disease. Clin
Genet. 1999;55:248–255.
35. Sevin M, Lesca G, Baumann N, et al. The adult form of
Niemann–Pick disease type C. Brain. 2007;130:120–133.
36. Canafoglia L, Bugiani M, Uziel G, et al. Rhythmic cortical myoclonus
in Niemann–Pick disease type C. Mov Disord. 2006;21:1453–1436.
37. Chen M, Wang J. Gaucher disease: review of the literature. Arch
Pathol Lab Med. 2008;132:851–853.
38. Zhao H, Grabowski GA. Gaucher disease: perspectives on a proto-
type lysosomal disease. Cell Mol Life Sci. 2002;59:694–707.
39. Brady RO. Enzyme replacement for lysosomal diseases. Annu Rev
Med. 2006;57:283–296.
40. Chang YC, Huang CC, Chen CY, et al. MRI in acute neuropathic
Gaucher’s disease. Neuroradiology. 2000;42:48–50.
41. Miyata R, Watanabe A, Hasegawa T, et al. Neurophysiological
analysis in an 18-month-old girl with Gaucher’s disease type 2. No
To Hattatsu. 2006;38:289–293.
42. Nishimura R, Omos-Lau N, Ajmone-Marsan C, et al.
Electroencephalographic findings in Gaucher disease. Neurology.
1980;30:152–159.
43. Zarate YA, Hopkin RJ. Fabry’s disease. Lancet. 2008;372:1427–1435.
44. Eng CM, Guffon N, Wilcox WR. Safety and efficacy of recombi-
nant human alpha-galactosidase A-replacement therapy in Fabry’s
disease. N Engl J Med. 2001;345:9–16.
45. Low M, Nicholls K, Turbridy N, et al. Neurology of Fabry’s disease.
Int Med J. 2007;37:436–447.
46. Lou HOC, Reske-Nielsen E. The central nervous system in Fabry’s
disease. Arch Neurol. 1971;25:351–359.
47. Gieselmann V, Franken S, Klein D, et al. Metachromatic leukodys-
trophy: consequences of sulphatide accumulation. Acta Paediatr
Suppl. 2003;443:74–79.
48. Wang PJ, Hwu WL, Shen YZ. Epileptic seizures and electroen-
cephalographic evolution in genetic leukodystrophies. J Clin
Neurophysiol. 2001;18:25–32.
49. Eggers C, Lederer H, Scheffner D. EEG changes in the course of
progressive cerebral disorders in children. Monatsschr Kinderheilkd.
1977;125:8–15.
50. Solders G, Celsing G, Hagenfeldt L, et al. Improved peripheral
nerve conduction, EEG, and verbal IQ after bone marrow trans-
plantation for adult metachromatic leukodystrophy. Bone Marrow
Transplant. 1998;22:1119–1122.
51. Scriver CR, Beaudet AL, Sly WS, et al., eds. The Metabolic and
Molecular Bases of Inherited Disease. 8th ed. New York: McGraw-
Hill; 2001:3669–3693.
52. Lyon F, Hagberg B, Evrand P, et al. Symptomatology of late onset
Krabbe’s leukodystrophy: the European experience. Dev Neurosci.
1991;13:240.
53. Husain AM, Altuwaijri M, Aldosari M. Krabbe disease: neurophys-
iologic studies and MRI correlations. Neurology. 204;63:4.
54. Igarashi M, Schaumburg HH, Powers J, et al. Fatty acid abnormal-
ity in adrenoleukodystrophy. J Neurochem. 1976;26:851–860.
296 Part III ■ Clinical EEG: General
55. Moser HW, Moser AB, Frayer KK, et al. Adrenoleukodystrophy:
increased plasma content of saturated very long chain fatty acids.
Neurology. 1981;31:1241–1249.
56. Kumar AJ, Rosenbaum AE, Naidu S, et al. Adrenoleukodystrophy:
correlating MR imaging with CT. Radiology. 1987;165:496.
57. Percy AK, Rutledge SL. Adrenoleukodystrophy and related disor-
ders. Ment Retard Dev Disabil Res Rev. 2001;7:179–189.
58. Barkovich AJ, Peck WW. Am J Neuroradiol. 1997;18:1163–1170.
59. Takahashi Y, Suzuki Y, Kumazaki K, et al. Epilepsy in peroxisomal
diseases. Epilepsia. 1997;38:182–188.
60. Govaerts L, Colon E, Rottevell J, et al. A neurophysiological study
of children with the cerebro-hepato-renal syndrome of Zellweger.
Neuropediatrics. 1985;16:185–190.
61. Panjan DP, Meglic NP, Neubauer D. A case of Zellweger syndrome
with extensive MRI abnormalities and unusual EEG findings. Clin.
Electroencephalogr. 2001;32:28–31.
62. Chang YC, Huang CC, Huang SC, et al. Neonatal adrenoleukodys-
trophy presenting with seizure at birth: a case report and review of
the literature. Pediatr Neurol. 2008;38:137–139.
63. Leigh D. Subacute necrotizing encephalomyelopathy in an infant.
J Neurol Neurosurg Psychiatry. 1951;14:216–221.
64. Pincus JH. Subacute necrotizing encephalomyelopathy (Leigh’s
disease): a consideration for clinical features and etiology. Dev
Med Child Neurol. 1971;14:78–87.
65. Medina L, Chi TL, DeVivo DC, et al. MR findings in patients with
subacute necrotizing encephalopathy (Leigh syndrome). Am
J Neuroradiol. 1989;11:379–384.
66. Canafoglia L, Franceschetti S, Antozzi C, et al. Epileptic phenotypes
associated with mitochondrial disorders. Neurology. 2001;56:
1340–1346.
67. Pavlakis SG, Phillips PC, DiMauro S, et al. Mitochondrial myopa-
thy, encephalopathy, lactic acidosis, and strokelike episodes: a dis-
tinctive clinical syndrome. Ann Neurol. 1984;16:481–488.
68. Stoquart-ElSankari S, Lehmann P, Perin B, et al. MRI and diffu-
sion-weighted imaging follow-up of a stroke-like event in a patient
with MELAS. J Neurol. 2008;255:1593–1595.
69. Ito H, Mori K, Harada M, et al. Serial brain imaging analysis of
stroke-like episodes in MELAS. Brain Dev. 2008;30:483–488.
70. Bell AJ. MELAS syndrome: overview and case study. Am
J Electroneurodiagn Technol. 1985;35:83–91.
71. Serra G, Piccinnu R, Tondi M, et al. Clinical and EEG findings in
eleven patients affected by mitochondrial encephalomyopathy
with MERRF–MELAS overlap. Brain Dev. 1996;18:185–191.
72. Ribacoba R, Salas-Puig J, Gonzalez C, et al. Characteristics of status
epilepticus in MELAS. Analysis of four cases. Neurologia. 2006;21:
1–11.
73. Shoffner JM, Wallace DC. Mitochondrial genetics: principles and
practice [editorial]. Am J Hum Genet. 1992;51:1179–1186.
74. Alpers BJ. Diffuse progressive degeneration of gray matter of cere-
brum. Arch Neurol Psychiatry. 1931;25:469–505.
75. Hunter M, Mackay MT, Peters H, et al. Alper’s syndrome with
POLG mutations: clinical, EEG, and radiological features. J Clin
Neurosci. 2008;15:341.
76. Joshi CN, Greenberg CR, Mhanni AA, et al. Ketogenic diet in
Alpers–Hutternlocher syndrome. Pediatr Neurol. 2009;40:314–316.
77. Boyd SG, Harden A, Egger J, et al. Progressive neuronal degenera-
tion of childhood with liver disease (“Alpers’ disease”): character-
istic neurophysiological features. Neuropediatrics. 1986;17:75–80.
78. Minassian BA. Lafora’s disease: towards a clinical, pathologic, and
molecular synthesis. Pediatr Neurol. 2001;25:21–29.
79. Acharya JN, Satishchandra P, Shankar SK. Familial progressive
myoclonus epilepsy: clinical and electrophysiologic observations.
Epilepsia. 1995;36:429–434.
80. Koskiniemi M, Donner M, Majuri H, et al. Progressive
myoclonus epilepsy. A clinical and histopathological study. Acta
Neurol Scand. 1974;50:307–332.
81. Pennacchio LA, Lehesjoki AE, Stone NE, et al. Mutations in the
gene encoding cystatin B in progressive myoclonus epilepsy
(EPM1). Science. 1996;271:1731–1734.
82. Caraballo RH, Fejerman N. Dravet syndrome: a study of 53
patients. Epilepsy Res. 2006;70s:S231–S238.
83. Del-Favero CL, Ceulemans J, Ceulemans B, et al. De novo muta-
tions in the sodium-channel gene SCN1A cause severe myoclonic
epilepsy of infancy. Am J Hum Genet. 2001;68:1327–1332.
84. Blaskovics M, Engel R, Podosin RL, et al. EEG pattern in
phenylketonuria under early initiated dietary treatment. Arch
Pediatr Adolesc Med. 1981;135:802–808.
85. Mina I, Hirotaka Y, Misao O. EEG pattern in phenylketonuria.
J Jpn Pediatr Soc. 2004;108:1366–1371.
86. Gross PT, Berlow S, Schuett VE, et al. EEG in phenylketonuria.
Arch Neurol. 1981;38:122–126.
87. Menkes JH, Hurst PL, Craig JM. A new syndrome: progressive
familial infantile cerebral dysfunction associated with an unusual
urinary substance. Pediatrics. 1954;14:462.
88. Scriver CR, Beaudet AL, Sly WS, et al., eds. The Metabolic and
Molecular Bases of Inherited Disease. Vol II. New York: McGraw-
Hill; 2001.
89. Brismar J, Aqeel A, Brismar G, et al. Maple syrup urine disease:
findings on CT and MR scans of the brain in 10 infants. Am
J Neurorad. 1990;11:1219.
90. Korein J, Sansaricq C, Kalmijn M, et al. Maple syrup urine disease:
clinical, EEG, and plasma amino acid correlations with a theoretical
mechanism of acute neurotoxicity. Int J Neurosci. 1994;79:21–45.
91. Mudd SH, Skovby F, Levy HL, et al. The natural history of homo-
cystinuria due to cystathionine beta-synthase deficiency. Am
J Hum Genet. 1985;37:1–3.
92. Del Giudice E, Striano S, Andria G. Electroencephalographic
abnormalities in homocystinuria due to cystathionine synthase
deficiency. Clin Neurol Neurosurg. 1983;85:165–168.
93. Buoni S, diBartolo RM, Molinelli M, et al. Atypical BECTS and
homocystinuria. Neurology. 2003;61:1129–1131.
94. Sass JO, Hofmann M, Skladal D, et al. Proprionic acidemia revis-
ited: a workshop report. Clin Pediatr. 2004;43:837–843.
95. Akman I, Imamoglu S, Demirkol M, et al. Neonatal onset propri-
onic acidemia without acidosis: a case report. Turk J Pediatr.
2002;44:339–342.
96. Aikoha H, Sasakia M, Sugaia K, et al. Effective immunoglobulin
therapy for brief tonic seizures in methylmalonic acidemia. Brain
Dev. 1997;19:502–505.
97. Bellieni CV, Ferrari F, DeFelice C, et al. EEG in assessing hydrox-
ycobalamin therapy in neonatal methylmalonic aciduria with
homocystinuria. Biol Neonate. 2000;78:327–330.
98. Gilbert-Barness E, Barness LA. Isovaleric acidemia with promye-
locytic myeloproliferative syndrome. Pediatr Dev Pathol. 1999;2:
286–291.
99. Teychenne PF, Walters I, Claveria LE, et al. The encephalopathic
action of five-carbon-atom fatty acids in the rabbit. Clin Sci Mol
Med. 1976;50:463–472.
100. Hoover-Fong JE, Shah S, Van Hove JL, et al. Natural history of non-
ketotic hyperglycinemia in 65 patients. Neurology. 2004;63:1847.
101. Applegarth DA, Toone JR. Nonketotic hyperglycinemia (glycine
encephalopathy): laboratory diagnosis. Mol Genet Metab. 2001;
74:139–146.
102. Flusser H, Kormann SH, Sato K, et al. Mild glycine encephalopa-
thy (NKH) in a large kindred due to a silent exonic GLDC splice
mutation. Neurology. 2005;64:1426–1430.
 Chapter 15 ■ EEG of Degenerative Disorders of the Central Nervous System
103. Hamosh A, McDonald JW, Valle D, et al. Dextromethorphan and
high-dose benzoate therapy for nonketotic hyperglycinemia in
an infant. J Pediatr. 1992;121:131–135.
104. Amladi TS, Kohil M. Hartnup disease. Indian J Dermatol Venereol
Leprol. 1994;60:105–107.
105. Lowe CU, Terrey M, MacLachan EA. Organic aciduria, decreased
renal ammonia production, hydrophthamos[hydrophthalamos],
and mental retardation: a clinical entity. Am J Dis Child. 1952;
83:164–184.
106. Schneider JF, Boltshauser E, Neuhaus TJ, et al. MRI and proton
spectroscopy in Lowe syndrome. Neuropediatrics. 2001;32:
45–48.
107. Sener RN. Lowe syndrome: proton MR spectroscopy and diffu-
sion MR imaging. J Neuroradiol. 2004;31:238–240.
108. Ono J, Harada K, Mano T, et al. MRI findings and neurologic
manifestation in Lowe oculocerebrorenal syndrome. Pediatr
Neurol. 1996;14:162–164.
109. Erdogan F, Ismailogullari S, Soyuer I, et al. Different seizure types
and skin lesions in oculocerebrorenal syndrome of Lowe. J Child
Neurol. 2007;22:427–431.
110. Leonard JV, Morris AA. Urea cycle disorders. Semin Neonatol.
2002;7:27.
111. Butterworth RF. Effects of hyperammonaemia on brain function.
J Inherit Metab Dis. 1998;21(suppl 1):6.
112. Maestri NE, Lord C, Glynn M, et al. The phenotype of ostensibly
healthy women who are carriers for ornithine transcarbamylase
deficiency. Medicine (Baltimore). 1998;77:389.
113. Verma NP, Hart ZH, Kooi KA. Electroencephalographic findings
in urea-cycle disorders. Electroencephalogr Clin Neurophysiol.
1984;57:105–112.
114. Clancy RR, Chung HJ. EEG changes during recovery from acute
severe neonatal citrullinemia. Electroencephalogr Clin Neurophysiol.
1991;78:222–227.
115. Nagata N, Matsuda I, Matsuura T, et al. Retrospective survey of
urea cycle disorders: part 2. Neurological outcome in forty-nine
Japanese patients with urea cycle enzymopathies. Am J Med Genet.
1991;40:477–481.
116. Hayflick SJ. Neurodegeneration with brain iron accumulation: from
genes to pathogenesis. Semin Pediatr Neurol. 2006;13:182–185.
117. McNeill A, Birchall D, Hayflick SJ, et al. T2 and FSE MRI distin-
guishes four subtypes of neurodegeneration with brain iron
accumulation. Neurology. 2008;70:1614.
118. Cossu G, Melis M, Floris G, et al. Hallervorden Spatz syndrome
(pantothenate kinase associated neurodegeneration) in two
Sardinian brother with homozygous mutation in PANK 2 gene.
J Neurol. 2002;249:1599–1600.
119. Remond A, ed. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 15A. Amsterdam: Elsevier; 162–191.
120. Swaiman K, Smith SA, Trock GL, et al. Sea blue histiocytes, lym-
phocytic cytosomes, movement disorder. 59Fe uptake in basal
ganglia: Hallervorden-Spatz disease or ceroid storage disease
with abnormal isotope scan? Neurology. 1983;33:301–305.
121. Seven M, Ozkilic A, Yuksel A. Dysmorphic face in two siblings
with infantile neuroaxonal dystrophy. Genet Counsel. 2002;13:
465–473
122. Hortnagel K, Nardocci N, Zorzi G, et al. Infantile neuroaxonal
dystrophy and pantothenate kinase-associated neurodegenera-
tion: locus heterogeneity. Neurology. 2004;63:922–924.
123. Morgan NV, Westaway SK, Morton JEV, et al. PLA2G6, encoding
a phospholipase A2, is mutated in neurodegenerative disorders
with high brain iron. Nat Genet. 2006;38:752–754.
124. Wakai S, Asanuma H, Hayasaka H, et al. Ictal video-EEG analysis
of infantile neuroaxonal dystrophy. Epilepsia. 1994;35:823–826.
297
125. Taly AB, Meenakshi-Sundaram S, Sinha S, et al. Wilson disease:
description of 282 patients evaluated over 3 decades. Medicine
(Baltimore). 2007;86:112–121.
126. Saito T. Presenting symptoms and natural history of Wilson dis-
ease. Eur J Pediatr. 1987;146:261–265.
127. Giagheddu M, Tamburini G, Piga M, et al. Comparison of MRI,
EEG, EPs, and ECD-SPECT in Wilson’s disease. Acta Neurol
Scand. 2001;103:71–81.
128. Heller GL, Kooi KA. The electroencephalogram in hepatolentic-
ular degeneration (Wilson’s disease). Electroencephalogr Clinical
Neurophysiol. 1962;14:520–526.
129. Hsich GE, Robertson RL, Irons M, et al. Cerebral infarction in
Menkes disease. Pediatr Neurol. 2000;23:425–428.
130. Kaler SG, Holmes CS, Goldstein DS, et al. Neonatal diagnosis and
treatment of Menkes disease. N Engl J Med. 2008;358:605–614.
131. Bahi-Buisson N, Kaminska A, Nabbout R, et al. Epilepsy in Menkes
disease: analysis of clinical stages. Epilepsia. 2006;47:380–386.
132. White SR, Reese K, Sato S, et al. Spectrum of EEG findings in
Menkes disease. Electroencephalogr Clin Neurophysiol. 1993;87:
57–61.
133. Wang D, Pascual JM, Yang H, et al. Glut-1 deficiency syndrome:
clinical, genetic, and therapeutic aspects. Ann Neurol. 2005;
57(1):111–118.
134. Bennett CL, Chen Y, Hahn S, et al. Prevalence of ALDH7A1
mutations in 18 North American pyridoxine-dependent seizure
(PDS) patients. Epilepsia. 2009;50:1167–1175.
135. Nabbout R, Soufflet C, Plouin P, et al. Pyridoxine dependent
epilepsy: a suggestive electroclinical pattern. Arch Dis Child Fetal
Neonatal Ed. 1999;81:F125–F129.
136. Torres OA, Miller VS, Buist NMR, et al. Folinic acid-responsive
neonatal seizures. J Child Neurol. 1999;14:529–532.
137. Gallagher RC, Van Hove JL, Scharer G, et al. Folinic acid-respon-
sive seizures are identical to pyridoxine-dependent epilepsy. Ann
Neurol. 2009;65:550–556.
138. Manto MU. The wide spectrum of spinocerebellar ataxia (SCAs).
Cerebellum. 2005;4:2–6.
139. Sack G, Gruss B, Lossner J, et al. EEG studies of spinocerebellar
ataxia. Psychiatr Neurol Med Psychol (Leipz). 1977;29:521–528.
140. Rasmussen A, Matsuura T, Ruano L, et al. Clinical and genetic
analysis of four Mexican families with spinocerebellar ataxia type
10. Ann Neurol. 2001;50:234–239.
141. Tan NCK, Zhou Y, Tan ASC, et al. Spinocerebellar ataxia type 2
with focal epilepsy—an unusual association. Ann Acad Med
Singapore. 2004;33:103–106.
142. Inazuki G, Baba K, Naito H. Electroencephalographic findings of
hereditary dentatorubral-pallidoluysian atrophy (DRPLA). Jpn J
Psychiatry Neurol. 1989;43:213–220.
143. Rötig A, de Lonlay P, Chretien D, et al. Aconitase and mitochon-
drial iron–sulphur protein deficiency in Friedreich ataxia. Nat
Genet. 1997;17:215–217.
144. Hewer RL. Study of fatal cases of Friedreich’s ataxia. Br Med J.
1968;3:649–652.
145. Bourchard RW, Bouchard JP, Bouchard R, et al. Electro -
encephalographic findings in Friedreich’s ataxia and autosomal
recessive spastic ataxia of Charlevoix-Saguenay (ARSACS). Can J
Neurol Sci. 1979;6:191–194.
146. Hagberg B, Hanefeld F, Percy A, et al. An update on clinically appli-
cable diagnostic criteria in Rett syndrome: comments to Rett
Syndrome Clinical Criteria Consensus Panel Ssatellite to European
Paediatric Neurology Society meeting, Baden Baden, Germany,
September 11, 2001. Eur J Paediatr Neurol. 2002;6:293–297.
147. Percy A, Lane J. Rett syndrome: clinical and molecular update.
Curr Opin Pediatr. 2004;16:670–677.
298 Part III ■ Clinical EEG: General
148. Bahi-Buisson N, Nectoux J, Rosas-Vargas H, et al. Key clinical
features to identify girls with CDKL5 mutations. Brain. 2008;131:
2647–2661.
149. Glaze DG, Frost JD Jr, Zoghbi HY, et al. Rett’s syndrome: correla-
tion of electroencephalographic characteristic with clinical stag-
ing. Arch Neurol. 1987;44:1053–1056.
150. Glaze DG. Neurophysiology of Rett syndrome. Ment Retard Dev
Disabil Rev. 2002;8:66–71.
151. Hagberg B, Witt-Engerstrom I. Rett syndrome: a suggested stag-
ing system for describing impairment profile with increasing age
towards adolescence. Am J Med Genet. 1986;26:47–59.
152. Pelc K, Boyd SG, Cheron G, et al. Epilepsy in Angelman syn-
drome. Seizure. 2008;17:211–217.
153. Valente KD, Andrade JQ, Grossmann RM, et al. Angelman syn-
drome: difficulties in EEG pattern recognition and possible mis-
interpretations. Epilepsia. 2003;44:1051–1063.
154. Stecker MM, Myers SM. Reserpine responsive myoclonus and
hyperpyrexia in a patient with Angelman syndrome. Clin Neurol
Neurosurg. 2003;105:183–187.
155. Stafstrom CE, Patxot OF, Gilmore HE, et al. Seizures in children
with Down syndrome: etiology, characteristics and outcome. Dev
Med Child Neurol. 1991;33:191–200.
156. Silva ML, Cieuta C, Guerrini R, et al. Early clinical and EEG fea-
tures of infantile spasms in Down syndrome. Epilepsia. 1996;37:
977–982.
157. De Simone R, Daquin G, Genton P. Senile myoclonic epilepsy in
Down syndrome: a video and EEG presentation of two cases.
Epileptic Disord. 2006;8:223–227.
158. Kaneko WM, Phillips EL, Riley EP, et al. EEG findings in fetal alco-
hol syndrome and Down syndrome children. Electroencephalogr
Clin Neurophysiol. 1996;98:20–28.

The EEG in Congenital Malformations of
Cortical Development, Neurocutaneous
Disorders, Cerebral Palsy, Autism/ Mental
Retardation, and ADHD/ Learning
Disabilities of Childhood
T
his chapter reviews numerous disorders of infancy and
childhood considered congenital with a known genetic
or familial basis or acquired by insults to the developing
brain. These include various structural abnormalities, cerebral
palsy, autism, mental retardation, attention-deficit hyperactiv-
ity disorder (ADHD), and learning disabilities.
CONGENITAL MALFORMATIONS
OF CORTICAL DEVELOPMENT
Anencephaly is a severe anomaly with absence of the cerebrum
and superior skull. These infants survive only a very brief time;
thus, EEG studies are not usually obtained for practical and
technical reasons.
Hydranencephaly refers to nearly complete absence of the
cerebral hemispheres that is thought most likely due to bilat-
eral cerebrovascular occlusions or intrauterine infections with
cytomegalovirus (CMV) or toxoplasmosis after the fourth
month of gestation. The infant is often initially normo-
cephalic and nondysmorphic. The EEG is usually attenuated
(1) but Velasco et al. reported a hydrancephalic patient with
seizures considered consistent with the Lennox Gastaut syn-
drome, though the typical slow spike-wave pattern was not
seen (2).
Hydrocephalus refers to a dilatation of the ventricular sys-
tem due to a variety of underlying brain structural anomalies
or secondarily acquired from bleeding or infection. It is often
treated by a neurosurgical procedure of shunting, typically
with a catheter from the right lateral ventricle to the peri-
toneal cavity. The underlying abnormalities of brain struc-
ture, the presence of the shunt (the burr hole as a
craniotomy), and at times complications of the shunt with
infection, hemorrhage, and multiple revisions can give rise to
EEG abnormalities and seizures. Ines and Markand (3)
reviewed the EEGs of 92 hydrocephalic patients and found a
higher incidence of EEG abnormalities in the shunted (95%)
as compared to the nonshunted patients (47%), with a 65.4%
versus 18.2% incidence of seizures in the two groups. Many of
CHAPTER
16
MARY REPOLE ANDRIOLA
the EEG abnormalities localized to the area of the shunt itself
and were first noted after surgery, though asynchrony of sleep
spindles was noted in both groups (3). Bartoshesky et al. (4)
found in studying a group of 111 children with meningomye-
locele that 24 out of 98 children with shunted hydrocephalus
had seizures. More of these children had other brain malfor-
mations besides Arnold Chiari, shunt infections, and perhaps
shunt revisions and these children were more likely to be
developmentally delayed than the rest (4). Saukkonen et al.
(5) following 168 shunted hydrocephalic children found
seizures in 22% of children prior to surgery in addition to
abnormal EEGs. Some of these patients presented with
seizures as newborns and some had infantile spasms with
hypsarrhythmic EEGs but there was no specific localization or
EEG pattern otherwise. The subsequent occurrence of
seizures, 25.6% after shunting for a total of 47.6% over a long
follow-up period with a mean of 8.9 years, they found related
to the lack of use of prophylactic antiepileptic medication
causing them to recommend treatment in those hydrocephalic
children prior to shunting with abnormal EEGs with epilepti-
form discharges. They did not find a relationship of EEG
abnormalities to the site of the shunt, number of revisions,
nor infections, but they did find the development of seizures
related to the occurrence of the slit ventricle syndrome (5). A
similar incidence of seizures (48.5%) was reported by Noetzel
and Blake (6) in 68 children with congenital hydrocephalus
not associated with myelomeningocele. Brain malformations,
developmental impairment, and spike, spike-wave, or slow-
wave paroxysmal EEG activity at seizure onset focal or diffuse
correlated with intractable seizures while age of shunting,
revisions, or infections did not. The few children without
mental retardation were able to come off their medication
after 3 years seizure free (6). Seizures may also occur in these
children as a symptom of shunt malfunction (7). Faillace and
Canady described the various EEG findings in 9 out of 15
patients with seizure activity in the presence of shunt mal-
function. These included diffuse dysfunction and focal spike
activity (Fig. 16.1) (8).
299
300 Part III ■ Clinical EEG: General

Figure 16.1 A 37-year-old woman with hydrocephalus, Arnold Chiari II, and myelodysplasia. Patient has had multiple
shunts and revisions—initially on the right, now on the left. History of seizures as a child. EEG: slowing and higher volt-
age on the right plus independent sharp activity from either hemisphere.

Holoprosencephaly (HPE) is a syndrome of congenital brain
malformations with incomplete separation of the hemispheres
and basal ganglia of varying severity often associated with midline
facial dsymorphisms, hypothalamic pituitary dysfunction,
developmental delay, abnormal EEGs, and epilepsy in approxi-
mately half the patients. A dorsal cyst may be present produc-
ing hydrocephalus that requires shunting. The disorder has
multiple causes: teratogenic and chromosomal including tri-
somy 13 and 15 and the sonic hedgehog mutation (SHH), SIX,
TGIG, and ZIC2. This can be considered an abnormality of
neural induction in the ventrodorsal plane (9). There are four
types: alobar, semilobar, lobar, and middle interhemispheric
variant (10,11). Clegg et al. (12) reviewed the literature and
reported on 32 children, approximately half of whom had
epilepsy with various forms of the disorder. Of the 20 patients
who had EEGs the abnormalities noted were hypersynchronous
theta during waking and sleep, prominent spindle-like beta
during drowsing and sleep, multiple epileptiform discharges,
and periods of attenuation. Posterior gradient flattening was
seen in one child with a dorsal cyst associated with alobar HPE
(Figs. 16.2 and 16.3) (12).
Schizencephaly refers to a full-thickness gray matter-lined
cleft of a hemisphere that may be unilateral or bilateral, open
(communication between ventricle and subarachnoid space) or
closed. The gray matter is abnormal, often composed of
polymicrogyri or heterotopic gray matter. This congenital dis-
order of brain development occurring at the end of the first and
the beginning of the second trimester may be associated with
other cerebral anomalies, particularly agenesis of the corpus
callosum, absence of the septum pellucidum, polymicrogyria,
and gray matter heterotopias, neurologic and developmental
impairment, epilepsy, and EEG abnormalities. Packard et al.
(13) reported on 47 children with all types of which 57% had
seizures, one third of which were difficult to control. The most
common seizure type was complex partial (70%). Though
seizures presented earlier and were harder to control in the
bilateral open lip type, there was no correlation of seizure type
or occurrence with the type of schizencephaly. Of the 27
patients with seizures, EEGs were unavailable in 1, normal in 3,
contained spike and sharp wave discharges or sharp waves
localizing to the cleft in 15, multifocal discharges in 5, and dif-
fuse or focal slowing in 3. The neurodevelopmental outcome of
these children was dependent on the extent of cortex involved.
They often have motor and language impairment (13). The
EEGs in nine patients with schizencephaly were reviewed by
Granata et al. (14). Focal epileptiform discharges that correlated
with their clinical seizures and the location of the cleft were the
most common findings. Their six patients with unilateral clefts

had IQs in the normal range, hemiparesis, and incompletely
controlled partial seizures that were postural at onset followed
by turning or complex partial but did not generalize (Fig. 16.4)
(14). A report of two cases by Iannetti et al. (15) raises the pos-
sibility of congenital CMV infection causing schizencephaly.
These patients were negative for a mutation in the homeobox
gene EMX2 that has been found in many patients with schizen-
cephaly (15).
A
Chapter 16 ■ The EEG in Congenital Malformations 301
Figure 16.2 A 10-year-old girl with semi-
lobar holoprosencephaly. The patient has
marked developmental retardation and at
times appears to have myoclonic spasms
involving her upper trunk but her parents
prefer not having her on medication. Her
EEG is without age-appropriate waking
and sleeping patterns and frequent high-
voltage paroxysmal synchronous notched
delta discharges as illustrated.
Hemimegalencephaly refers to congenital hypertrophy of
one cerebral hemisphere with abnormal chaotic cytoarchitec-
ture of gray and white matter with an ipsilateral enlarged ven-
tricle. Clinically, the patients have hemiparesis, mental
retardation, and seizures. The EEG data in 12 patients reported
by Paladin et al. (16) found partial seizures in all patients and 3
also had infantile spasms. In addition, they also reviewed the lit-
erature. Three EEG patterns were observed, tended not to
Figure 16.3 An 8-year-old girl with the
middle interhemispheric variant of holo-
prosencephaly. She has mild facial dys-
morphism, developmental delay, and
symptomatic epilepsy. Her EEG reveals
(A) asynchrony and multifocal spikes, (B)
and marked activation by sleep consistent
with ESES. (C) MRI demonstrates fused
frontal lobes, absent septum pellucidum,
and falx. (D) MRI shows agenesis of the
body of the corpus callosum with radia-
tion of the gyri.(continued)
302 Part III ■ Clinical EEG: General

C D

Figure 16.3 (continued)

change over time, and were correlated with the patient’s clinical
history:
1. Triphasic spike complexes of large amplitude maximum in
the abnormal hemisphere that appeared early as did seizures.
No significant developmental milestones were achieved and
one patient died at age 7 days.
2. Isolated “alpha-like” activity appeared at approximately
6 months of age in the central area and extended to the
frontal by 2 years of age. Though seizure onset occurred
before 6 months, these children learned to walk and talk and
could attend special schools.
3. Suppression burst occurred at birth or within the first few
months of life. This pattern seemed to respond to steroid ther-
apy with a relatively favorable outcome in two patients (16).
Aicardi syndrome as first described in 1965 occurs in girls with
severe mental retardation and ocular abnormalities who present

Figure 16.4 A 39-year-old man with incompletely controlled partial epilepsy and right hemiparesis secondary to left pari-
etal open lip schizencephaly who nevertheless holds a job and is married. EEG findings of low-voltage left temporal sharp
waves and underlying slowing.
with infantile spasms (17). Their brains have a thin unlayered cor-
tex with diffuse polymicrogyria with fused molecular layer, nodu-
lar heterotopias in the periventricular region and centrum
semiovale, and agenesis of the corpus callosum. The patients may
also have chorioretinal lacunae, coloboma, and vertebral and
costal abnormalities. They usually have a shortened life span. The
EEG findings are that of completely asynchronous activity
between the hemispheres consistent with agenesis of the corpus
callosum plus multifocal discharges on a burst suppression pat-
tern. There is also absence of recognizable stage II sleep activity
such as spindles (18). There are more than 100 cases reported (19).
Agenesis of the corpus callosum may be seen as a total or
partial phenomenon with the severity of clinical and EEG find-
ings often related to other brain anomalies. Lacey (20) reviewed
40 children prospectively followed for up to 15 years. More than
half of the patients were retarded and half had seizures. EEGs
were done in all patients and abnormalities were consistent
with the clinical picture from hypsarrhythmia in the seven
patients with infantile spasms to background slowing and mul-
tifocal and generalized discharges in the patients with retarda-
tion and symptomatic epilepsy (Fig. 16.5) (20). Agenesis of the
corpus callosum can also be caused by metabolic disorders (21).
When a lipoma occurs in the corpus callosum, epilepsy, usu-
ally partial and difficult to control, and an abnormal EEG, begins
in childhood. The EEG abnormality may appear temporal or
Chapter 16 ■ The EEG in Congenital Malformations 303
centrotemporal. Some of the patients are retarded or develop-
mentally delayed. Gastaut et al. proposed that the seizures were
related to the disconnection of the corpus callosum (22).
Intrauterine infections with viruses under the acronym
TORCH (toxoplasmosis, rubella, CMV, and herpes simplex
virus) plus syphilis can cause some of the congenital anomalies
reviewed previously in this section. Infants with infections close
to the time of delivery may have systemic symptoms. Sequelae
include chorioretinitis, sensorineural hearing loss, hydro-
cephalus, microcephaly, developmental retardation, various neu-
rologic deficits, and seizures with abnormal EEGs (23).
A number of teratogens are known to adversely affect the
fetus. Fetal exposure to alcohol is a now-known preventable ter-
atogen causing fetal alcohol syndrome (FAS) with microcephaly
(as part of the intrauterine growth retardation), developmental
retardation, and irritability and restlessness in infants.
Craniofacial anomalies are also reported (24). A study by
Kaneko et al. (25) comparing the EEG findings in 18 matched
children with the FAS and Down syndrome versus normal con-
trols using power spectra analysis found significant reductions
in the mean power of alpha frequencies for both clinical groups.
While Down syndrome children had diffuse slowing, FAS chil-
dren had decreased posterior alpha especially on the left (25).
Another common modern teratogen inflicting harm on the
fetal brain is cocaine. Infants exposed to cocaine have been
304 Part III ■ Clinical EEG: General
Figure 16.5 A 13-month-old female infant with agenesis of the corpus
callosum, microcephaly, congenital cataracts, symptomatic intractable
epilepsy, and severe developmental delay. The patient has a brother
with the same disorder. Her EEG is markedly abnormal with lack of age-
appropriate background rhythms, asynchrony, and multifocal spikes.
Hemispherical asynchrony is the electrographic hallmark of this disor-
der. Her EEG at 5 months when she presented with infantile spasms
had a hypsarrhythmic pattern.
reported to have smaller head circumferences from birth
through at least 2 years of age, and developmental delay when
evaluated at 2 years (26). A report of 10 children with intrauter-
ine exposure to cocaine and other street drugs found significant
intracranial and ophthalmologic anomalies including absence
of the corpus callosum, schizencephaly, hydrancephaly, poren-
cephaly, septo-optic dysplasia, and optic atrophy (27). Half of
the patients had seizures. Many of these malformations and
their associated EEG findings have been described above.
Congenital blindness may be associated with cerebral mal-
formations, developmental retardation, or be purely on an ocu-
lar basis. The EEG of a blind child usually lacks occipital alpha
while central rhythms are preserved. Occipital spikes may be
seen in the absence of clinical seizures. These findings tend to
decrease with age (28).
We should note, however, that the earlier theory that visual
deprivation can cause electrical abnormalities over the visual
cortex was not supported by some data, namely by the study of
Jan and Wong (29) who investigated 104 children with ocular
visual loss. In this study although most totally blind children’s
records lacked alpha, five had spontaneously appearing and dis-
appearing alpha rhythm.
A more recent study (30) showed a significant reduction of
alpha power in the blind compared to the sighted controls over
parieto-occipital recording sites and concluded that brain
structures, associated with the generation of posterior alpha
rhythms in sighted adults, might be either reorganized or atro-
phied following blindness from birth.
Congenital porencephaly is residua of an infarct that may have
occurred during the pre- or perinatal period. The patient usually
has a hemiparesis, developmental delay, and epilepsy. Ho et al.
(31) describe the seizure localization in 14 patients with poren-
cephaly and intractable seizures. The porencephaly was in the dis-
tribution of the middle cerebral artery in eight, posterior cerebral
in three, internal carotid in one, and multiple vessels in two. The
majority of patients had complex partial seizures. Scalp EEGs
demonstrated abnormal background with focal slowing over the
cyst in four patients. Eight patients had interictal discharges noted
anterior temporally (one bilateral) and four extratemporally. Ictal
monitoring confirmed temporal localization as did one intracra-
nial monitoring. A total of 13 patients had hippocampal forma-
tion atrophy ipsilateral to the side of the lesion, 2 of whom
underwent surgery and became seizure free (31).
Arachnoid cysts are intracranial congenital lesions in which
redundant infoldings of the arachnoid membrane appears to
trap cerebral spinal fluid into a cyst. These are often found
“incidentally” when an imaging study is obtained post-trauma
or as an evaluation for another diagnosis such as epilepsy. There
has been a question raised as to whether or not these cysts play
a causative role in a patient’s epilepsy and therefore should be
treated surgically. The consensus appears to be that cyst
removal is not likely to reduce seizures, but instead these
patients need to be treated with medication even if epileptiform
activity is noted in the region of the cyst (32,33).
Infantile nonprovoked epilepsy after the neonatal period is
often related to cortical dysplasias. The pathogeneses and the
pathology of malformations of cortical development are
reviewed by Barkovich et al. (34). Focal cortical dysplasia refers
to a localized area of abnormal cortical structure that may be
appreciated now preoperatively by a combination of improved
imaging and EEG studies. Patients tend to have intractable
epilepsy with focal interictal discharges in about half the cases.
They may be neurologically normal or impaired and may be
candidates for seizure surgery.
Tassi et al. in a review of histologic sections from a large
epilepsy surgery series correlated the clinical, EEG, MRI, and
neuropath findings in patients with cortical dysplasia. They sep-
arate the pathology into the following three subtypes: (i) archi-
tectural dysplasia with disruption of cortical lamination and
ectopic neurons in the white matter, (ii) cytoarchitectural dyspla-
sia with abnormal cortical lamination plus giant neurofilament-
enriched neurons, and (iii) (Taylor type) cortical dysplasia,
abnormal cortical layering plus giant dysmorphic neurons, and
balloon cells. Seizure frequency was more severe with the
increased severity of the pathology. The epileptogenic zone was
mainly extratemporal and the best seizure free rate after surgery
was in the Taylor type (35). Krsek et al. (37) reviewing the results
of 200 children who underwent seizure surgery for cortical dys-
plasia used the more complex classification system proposed in
2004 by Palmini et al. (36) that has more subgroups and corre-
lated history, clinical findings, EEG, MRI, neuropsychological,
surgical variables, and outcome data. Mild malformations of cor-
tical development and focal cortical dysplasia type I were more
frequently associated with perinatal risk factors. Ictal EEGs were
more frequently precisely localized in the mild cortical dysplasia
and the focal cortical dysplasia II group b (37). Kim et al. (38)
report on 166 consecutive adult patients at the time of surgery

followed over a 10-year span. Forty additional patients were
excluded as they also had hippocampal sclerosis. Scalp video EEG
monitoring, interictal (42.7%), and ictal (76.5%) correctly local-
ized to the resected lobe, while the MRI had a detection rate of
42.2% that was greatest the more severe the cortical dysplasia
(38). For both series, the best surgical outcome was related to the
completeness of the resection (37,38).
Congenital abnormalities of brain structure now known to
be or likely to be on a chromosomal (genetic basis) such as neu-
ronal migrational disorders often lead to EEG abnormalities
and symptomatic epilepsy. Examples are periventricular nodu-
lar heterotopia (PNH), subcortical band heterotopia (SBH),
and lissencephaly (34).
PNH are subependymal nodules of gray matter along the
superior lateral walls of the lateral ventricles from the frontal
horns to the trigones. Bilateral PNH occurs in females as an X-
linked dominant (Xq28) usually with mild epilepsy onset in the
second decade and normal to borderline IQ. The interictal EEG
is normal or there are temporal discharges (Fig. 16.6). When the
periventricular nodules are seen in association with other brain
malformations, the patients usually are males, have a younger
age at presentation, mental retardation, more difficult to con-
trol seizures (often atonic), and EEGs with focal and bisynchro-
nous discharges (39).
Battaglia et al. (40) discuss the spectrum of PNH in 54
patients including the various types with known genetic point
mutations of the FLNI gene in bilateral, symmetrical familial
occurring mostly in females and familial patients with bilateral
PNH not related to the gene and sporadic cases. Unilateral PNH
is usually found in the posterior paratrigonal region of the lat-
eral ventricle but can extend into white matter and cortical tis-
sue. In this study, 12 out of 14 were females with 4 patients
having abnormal overlying neocortical areas and 8 having ipsi-
Chapter 16 ■ The EEG in Congenital Malformations 305
lateral hippocampal hypoplasia. Three patients were either
autistic or had mild mental retardation (Fig. 16.7). Most
patients had epilepsy with focal seizures difficult to control and
the three who did not yet were under 15 years. Background
rhythms were normal as were stage II sleep patterns. Photic
driving was bilateral except in the unilateral cases where it was
ipsilateral. Multifocal epileptiform activity was seen interictally
even in the unilateral patients in addition to bilateral asynchro-
nous abnormalities though in slow-wave sleep, bilaterally dif-
fuse bursts of polyspikes were seen. Three ictal patterns are also
described all partial onset. There is some evidence that seizure
activity may arise from the heterotopic nodules (40,41). The
patients may have hippocampal sclerosis; however, surgery to
remove the hippocampus does not achieve seizure control (41).
Double cortex syndrome or SBH is characterized by a sub-
cortical band of gray matter and an abnormal DCX gene (XLIS)
in most cases (34). Mosaic mutations of the LISI gene that is
usually the responsible gene in severe lissencephaly can be
found in patients with posteriorly prominent SBH (42). These
patients, more commonly female, are retarded with usually
intractable epilepsy. The EEG pattern most typically seen is that
of hypsarrhythmia with clinical infantile spasms followed by a
Lennox Gastaut pattern though multifocal spikes may be noted
(Figs. 16.8–16.10) (43,44). In a 13-year-old female patient who
had generalized tonic and grand mal seizures and the DCX gene
mutation by DNA analysis, a persistent Lennox Gastaut slow
spike-wave pattern was seen on EEG associated with SBH on
MRI and increased PET FDG uptake in the band (45).
Mutations of the X-linked gene DCX (XLIS) usually cause SBH
in females and lissencephaly in males (46).
Lissencephaly that means a smooth cerebral surface is also
caused by deletions involving the L1S1 gene on chromosome
17 p13.3 (34). This can be isolated or part of a syndrome such
Figure 16.6 A 12-year-old girl with
developmental delay, partial-onset
seizures, and bilateral frontal and lateral
periventricular heterotopic gray matter
nodules. EEG: right mid-temporal–central
sharp and slow waves with wide field acti-
vated by drowsiness.
306 Part III ■ Clinical EEG: General

Figure 16.7 An 8-year-old girl with

macrocephaly, developmental delay/ autis-
tic spectrum disorder, and partial-onset
symptomatic epilepsy with head turning
and eye deviation to the left, lip smacking,
and confusion. A heterotopic gray matter
lesion close to the right posterior lateral
ventricle is noted on MRI while the EEG
demonstrates focal slowing from the right
posterior quadrant.

Figure 16.8 A 3-year-old girl who presented at 5 months with infantile spasms that responded to ACTH treatment. She
then developed prolonged complex partial seizures at 2.5 years. Initial MRI was reported as normal; however, subsequent
MRI revealed double cortex syndrome. EEG at 5 months—high-voltage bursts of polyspike and sharp wave discharges are
noted in both the waking and sleeping states. These bursts are often bilateral and are followed by attenuation. There are
also sharp wave transients seen in a multifocal manner. Often, when the patient has these discharges, she shows no clin-
ical change. At other times when she has these discharges in the waking state, she may stop and blink, she may have a
head drop, or she may have an actual infantile spasm.
 Chapter 16 ■ The EEG in Congenital Malformations 307

Figure 16.9 Polymorphic predominantly delta activity, occasionally sharp contoured is seen bilaterally but maximal in
voltage in right temporal region. These transients were associated with two clinical episodes of staring and unresponsive-
ness. Right temporal sharp and slow waves are seen throughout with phase reversals at T4 and T6.

as the Miller–Dieker syndrome. The patients are usually
severely retarded, initially hypotonic, and then spastic. Most
have intractable seizures and abnormal EEGs with diffuse fast
rhythms of high amplitude though diffuse slow is also seen in
the first year of life (47). A study of 17 patients in the
Netherlands by de Rijk van Andel et al. (48) with serial EEG
recordings found high-amplitude beta at times cycling with
sharp- and slow-wave complexes especially when the patients
had more seizures. Other patterns included high-voltage slow
and diffuse rhythmic high-voltage alpha and beta patterns
(48). Sébire et al. (49) reported 43 children with macrogyri or
agyri (lissencephaly type I). Rapid rhythms of high-voltage
and high-amplitude delta to theta complexes were seen in the
latter group (49).
Patients with bilateral perisylvian polymicrogyria syn-
dromes have pseudobulbar palsy, mental retardation, and usu-
ally symptomatic epilepsy with most typically the Lennox
Gastaut syndrome clinically and the pattern on EEG. MRI
reveals bilateral perisylvian microgyri while pathologically
small prominent convolutions are seen without proper cortical
lamination (50). A more widespread bilateral generalized
polymicrogyria syndrome was reported by Chang et al. (51)
and there are other regional polymicrogyri syndromes. Thin
excessively folded cortex without the normal six-layer struc-
ture is seen in children with developmental delay and seizures
that may be intractable. EEGs may show multifocal spikes.
Guerrini et al. described nine patients with multilobar polymi-
crogyri and electrical status epilepticus during sleep (ESES)
with infrequent focal motor seizures but eight out of nine had
atonic drop attacks. These patients however who ranged from
severely mentally retarded to one patient with normal cogni-
tive function outgrew their ESES and were either seizure free
off or on medication except for two patients with infrequent
seizures (52).
NEUROCUTANEOUS DISORDERS
Neurofibromatosis I (NFI) is a neurocutaneous disorder associ-
ated with skin, eye, brain, and systemic lesions. Seizures and asso-
ciated EEG abnormalities occur in a minority of patients. Young
et al. reviewing their experience in an NF clinic over a 7-year
period found 5% of the patients had seizures diagnosed mainly in
childhood (53). Korf et al. (54) found 22 out of 359 persons with
NFI followed in a children’s hospital over a 10-year period, to have
one or more seizures, with onset ranging from newborn to over
20 years. There was a variety of seizure types including febrile
308 Part III ■ Clinical EEG: General
Figure 16.10 MRI: there is a bilateral and symmetrical mantle of tis-
sue within the white matter bilaterally that follows gray matter signal
on all sequences (arrows) and is consistent with band heterotopias
(double cortex syndrome).
seizures and one infant with infantile spasms but in none did the
seizures appear to correlate with a structural lesion in the brain.
EEG findings also varied depending on the seizure type from nor-
mal to focal to generalized discharges (54). Kulkantrakorn and
Geller obtained similar results in their retrospective study of 499
NFI patients (55), while Vivarelli et al. (56) reported a correlation
of seizures in their patients with structural lesions seen on imag-
ing in 64%. A small percentage (7%) of their 198 patients with
NFI was found on retrospective analysis to have epilepsy (56).
Tuberous sclerosis (TS) is an autosomal dominant disorder
that involves the skin, eyes, multiple organs, and the brain. It is
a dominant trait with variable low penetrance, but more than
50% of cases are de novo mutations. There are two genes
involved, TSC1 (hamartin 9q34) and TSC2 (tuberin 16p13).
More severe neurologic involvement is reported with the latter
(57). The patients may have abnormal EEGs often presenting
with seizures in the first year of life as infantile spasms with
hypsarrhythmia or atypical hypsarrhythmia or partial seizures
with the subsequent development of the Lennox Gastaut syn-
drome (58). The ultimate cognitive, behavioral (autistic spec-
trum), and developmental outcome is directly related to the
earlier onset of epilepsy and the severity of the abnormality of
the EEG and proportional to the control of the epilepsy
(Fig. 16.11) (59).
Seizures and neurobehavioral deficits appear directly related
to the cerebral tubers (number and tuber burden) that are pres-
ent in the majority of patients and are characterized by a loss of
the normal cortical six-layer structure and dysmorphic neu-
rons, astrocytes, and giant cells that have a pathologic appear-
ance similar to cortical dysplasia (60). The greater the number
of tubers and the tuber burden, the more likely an infant is to
have infantile spasms and to be cognitively impaired. This
makes it even more imperative that they are treated early with
vigabatrin, perhaps over ACTH and even surgery on occasion
(57,61–63). Major et al. (64) studying the epileptogenicity of
the tubers and the surrounding cortex in three patients with TS
undergoing seizure surgery using subdural grids on the tubers
and cortex and depth electrodes into the tubers found signifi-
cant epileptiform activity in the surrounding tissue while the
tuber was cold. Patient seizures were markedly reduced after the
tuber and local tissue were removed (64,65).
Sturge Weber syndrome is a neurocutaneous syndrome
with leptomeningeal angiomatosis most commonly in the
occipital, parietal region, facial cutaneous vascular malforma-
tion in the ophthalmic division especially involving the skin
around the eye and eye lid, glaucoma, and hemianopsia. This is
presumably based on an embryogenic error in the formation of
the cephalic venous microvasculature that occurs at 5 to 8
weeks gestation. Venous stasis occurs leading to ischemia of the
underlying cortex with laminar cortical necrosis and calcifica-
tion. Symptomatic partial epilepsy that is often progressive
occurs in 75% to 90% of patients by age 3 with the initial
seizures often associated with fever. The severity of the epilepsy
has not been directly related to prognosis but instead 50% to
75% of the patients have mental retardation, ADHD, and
recurrent stroke-like episodes. Depending on the size of the
malformation, 25% to 60% develop a hemiparesis and a vascu-
lar-type headache occurs in about half (65). The initial EEG
may reveal attenuation ipsilateral to the vascular malformation
and then subsequent focal and then bilateral epileptiform
activity (66).
Incontinentia pigmenti (Bloch–Sulzberger syndrome) is a
rare neurocutaneous disorder that is X-linked (Xq28) occur-
ring predominantly in females as it is lethal in the male. The
disorder involves the skin with a rash that may present at birth
on the flexor surface of the limbs and trunk, eventually leading
to first linear hyperpigmentation and then hypopigmentation.
Abnormalities of the hair, teeth, and eyes with macular lesions,
retinal detachments, and strabismus are reported. Focal
seizures with onset in the newborn period, abnormal EEGs
with focal epileptiform discharges, and residual neurologic
deficits including developmental retardation and hemiparesis
can occur. Microvascular infarcts noted on MRI in the periven-
tricular white matter appear to begin as early as the first week
of life (Fig. 16.12) (67).
Hypomelanosis of Ito (HI) or incontinentia pigmenti achro-
mans is a neuroectodermal disorder characterized by large
hypopigmented areas that are whorled with various neuronal
migrational abnormalities such as hemimegencephaly, pachy-
gyria, cortical dysplasia, gray matter heterotopias, gray/white
matter hamartomas plus abnormal white matter, and neuronal
morphology. Mental retardation occurs in the majority, and
symptomatic epilepsy with focal EEG abnormalities occurs in
approximately half the patients. Placantonakis et al. reported
two patients with hemimegencephaly, HI, and intractable
seizures with focal discharges who successfully underwent
seizure surgery (68).

Figure 16.11 A 5-year-old girl with intractable partial-onset symptomatic epilepsy secondary to tuberous sclerosis. EEG:
prominent beta is seen with the patient in the sedated sleeping state. Slow and sharp activity is seen from multiple foci
including left anterior temporal and left parietal regions.
Linear nevus sebaceous syndrome is a neurocutaneous dis-
order with epidermal sebaceous facial or scalp nevus associated
with many of the cerebral malformations reviewed in the sec-
tion “Congenital Malformations of Cortical Development,”
neurologic deficits, mental retardation, and symptomatic
epilepsy in more than half the patients and EEG abnormalities.
Seizures may present in infancy as infantile spasms with hypsar-
rhythmic EEG or as focal seizures. Other epileptic syndromes
may develop and epilepsy is usually intractable (69).
CEREBRAL PALSY
Neonatal seizures have a high association with subsequent
seizures and developmental disabilities including cerebral palsy
and developmental delay. Guillet and Kwon found no statistical
difference in outcome in infants discharged home after neona-
tal seizures on prophylactic phenobarbital or not; 24% had
seizures, 19% of their patients developed cerebral palsy, and
54% were developmentally delayed (70). As phenobarbital has
been shown harmful to newborn brain in the experimental ani-
mal model, concern has been expressed regarding treatment,
but it is clear that neonatal seizures are also harmful and that
neonates with seizures have a poor prognosis. Severely abnor-
Chapter 16 ■ The EEG in Congenital Malformations 309
mal neonatal EEGs such as a burst suppression pattern in
infants who have suffered hypoxic ischemic encephalopathy are
associated with a poor outcome for mortality and morbidity as
reported by Sinclair et al. (Fig. 16.13) (71).
Cerebral palsy by definition “is an umbrella term covering a
group of nonprogressive but often changing motor impairment
syndromes secondary to lesions or anomalies of the brain aris-
ing in early stages of its development” (72). The insult or brain
anomaly may occur prenatal, perinatal, or postpartum. There is
no specific etiology. The severity may vary, but there is no pro-
gression or improvement, though motor manifestations may
change with age. There are common clinical features such as
spasticity, but these are not universal and comorbidities occur
including abnormal EEGs and symptomatic epilepsy
(Fig. 16.14). Motor findings may be unilateral (hemiplegic) or
bilateral (para- or quadriplegic) and abnormal tone may occur
as hypotonia or abnormal movements (choreoathetotic).
Abnormal EEGs are seen in patients with cerebral palsy without
clinical seizures on no medication, though this may only be for
a point in time. The more abnormal the brain is, structurally,
the more vulnerable it is to electrographic disturbances and the
clinical manifestation of epilepsy. Russman and Ashwal
reviewed studies including 1918 children with cerebral palsy
310 Part III ■ Clinical EEG: General

Figure 16.12 A 6-year-old girl with typical skin lesions of incontentia pigmenti, a disorder inherited
from her mother. She had the onset of symptomatic epilepsy at 3 months of age. Right focal motor
seizures remain difficult to control, her development has been delayed, and she has developed a left
hemiparesis. EEG demonstrates poorly organized background rhythms for age and multifocal spikes
seen in waking and sleeping states.

Figure 16.13 A 3-day-old full-term infant with hypoxic ischemic encephalopathy secondary to pro-
lapsed cord with Apgars of 1, 1, and 4. Seizures developed within a few hours of life and were treated
with phenobarbital and fosphenytoin. He is on the ventilator in a coma with a burst suppression EEG
pattern.
 Chapter 16 ■ The EEG in Congenital Malformations 311

Figure 16.14 A 6-year-old boy with cerebral palsy left hemiparesis, developmental delay, and partial
symptomatic epilepsy. His MRI shows areas of tissue loss with gliosis in the inferior posterior right
frontal lobe most likely the sequela of a prior infarct. EEG findings: persistent epileptiform activity as
spike and slow waves from the right centroparietal region activated by sleep.

and found an average of 43% with epilepsy. Severity was related
to the type of cerebral palsy, onset of seizures as neonate, num-
ber of seizure medications, intractability, abnormalities on
imaging studies, and lower intelligence (73). When periventric-
ular leukomalacia is present, these children are likely to have
cerebral palsy, cognitive impairment, visual difficulties, and
symptomatic intractable epilepsy usually of multifocal onset
and multiple seizure types (74).
Ulegyria is scaring of the cortex with thinning and small
mushroom-shaped gyri and atrophy at the base of the convolu-
tion and better preserved apex. This is frequently associated
with epilepsy and is considered related to perinatal ischemia.
Kuchukhidze et al. reported on a group of 25 patients with the
typical MRI findings and the comorbidities of cerebral palsy
and neurologic disabilities and the possibility of seizure surgery
for their pharmacoresistant seizures (75). Usui et al. (76) car-
ried out seizure surgery on 10 patients with posterior cortex
epilepsy secondary to ulegyria. Nine had the classic perinatal
history of hypoxia. Engel’s class I outcome was achieved in
seven (76).
AUTISM/ MENTAL RETARDATION
Autism is now considered a developmental disorder with its
specific features listed in the Diagnostic and Statistical Manual
of Mental Disorders, 4th Edition (DSM-IV) (77). Rapin
reviewed the disorder’s broad clinical manifestations, evalua-
tion, and therapeutic approaches. She writes “EEGs are indi-
cated for children in whom epilepsy is suspected but it should
be kept in mind that nonepileptic staring spells are much more
common than absence seizures” (78). A study by Kagan-
Kushnir et al. to establish guidelines for the use of screening
EEGs in autism spectrum disorders to discover subclinical
epileptiform activity reviewed 25 studies and resulted in a con-
clusion that there is insufficient evidence to recommend for or
against routine screening EEGs in autistic patients (79). It has
been recommended to obtain overnight EEG recordings to cap-
ture prolonged sleep in children even without known seizures
but with language or autistic regression, fluctuating deficits or
mute children (76).
Frank epileptiform abnormalities seen on EEGs in the autis-
tic children such as “rolandic” spikes or benign epileptiform
discharges with centrotemporal spikes (BECTs) have been
treated with antiepileptic drugs, high-dose steroids, high-dose
Valium before bed, and surgery even in patients without appar-
ent clinical seizures. There are no evidence-based guidelines
available. Approximately one third of individuals with autism
have seizures and these often present after early childhood in
the second and third decades (Fig. 16.15).
Less than 10% of children with autistic symptoms have a
known etiology such as in the case of Rett syndrome that is
included in the autistic spectrum. Some children with
Angelman, fragile X (FXS), and Down syndromes and TS also
have autistic symptoms.
312 Part III ■ Clinical EEG: General
Figure 16.15 A 17-year-old boy with autism and onset of seizures. Waking EEG was normal. Sleeping
EEG: left temporal spike- and slow-wave transients with underlying slow waves. His seizures were
responsive to Lamictal.
Canitano et al. (80) studied 46 consecutive children who met
the DSM-IV criteria for autism with waking and chloral
hydrate-induced sleep. Epileptic abnormalities were seen in
35% with only 6 (13%) out of the 16 patients having epilepsy of
various categories. Autism with regression was found in
approximately half and 20% showed epileptiform abnormali-
ties on EEG, but only one had seizures. The children with
epilepsy were reported as “severely retarded” (80).
Problems also arise in distinguishing the moderate to severely
retarded child in whom the development of language and social
function is impaired merely due to retardation without neces-
sarily attributing another diagnosis or disorder such as autism.
Mental retardation refers to a significant impairment in cogni-
tive and general adaptive abilities manifested during the develop-
mental period. Retardation may be mild (IQ 50 to 70), moderate
(IQ 30 to 50), or severe (IQ 30) and the more severe the more
likely there are other associated neurologic signs and symptoms
including seizures and abnormal EEGs. In a population-based
study, Steffenburg et al. (81) studied epilepsy in retarded children
6 to 13 years of age. The more severe the retardation, the younger
the age of seizure onset. There were a variety of seizures and some
children had more than one type. Approximately half had fre-
quent, daily to weekly seizures. This group also included children
with cerebral palsy and other neurologic impairment (81).
Nonsyndromic autosomal mental retardation has been iden-
tified in patients with various genetic synaptic abnormalities
representing de novo truncating mutations (K138X, R579X,
and L81RFsX22). Of the 3 nondysmorphic children found out
of a group of 94 mentally retarded patients by Hamdan et al. in
whom they sequenced the gene SYNGAP1, 2 had seizures and
abnormal EEGs “with bioccipital spikes during intermittent
light stimulation” (82).
There are many etiologies though mental retardation is more
common in boys in part thought to be related to mutations in X-
linked genes first described with the FXS. The clinical description
of X-linked mental retardation by Martin and Bell in 1943 (83)
was followed by a fragile site on the X chromosome found in four
boys and an unaffected female by Lubs in 1969 (84). Subsequent
improvement of the culture method by the 1980s allowed the
identification of what is now considered the most common chro-
mosomal inherited form of mental retardation and developmen-
tal delay for which genetic testing is possible. FXS is due to an
expansion of a trinucleotide (CGG) repeat of the FMR1 gene on
the X chromosome. It is the lack of messenger RNA causing lack
of FMRP protein that causes the syndrome.
It is much more common in males who are dysmorphic with
large ears, and who in adolescence develop macroorchidism.
Their developmental delay is greatest in the area of speech/lan-
guage and they are often hyperactive. Some are diagnosed as
autistic. Approximately 20% will have usually infrequent
seizures easy to control as was my experience with a boy I fol-
lowed in the 1970s initially with developmental delay and
hyperactivity. At age 9, he experienced a brief seizure in his
sleep with a typical EEG with centrotemporal spikes noted in
sleep. His family preferred not to treat. His next and last seizure
occurred 4 years later. Another son was born much more lan-
guage impaired but he never experienced a seizure. They both
were subsequently diagnosed with FXS in the 1980s.

Musumeci et al. reported on 192 patients with FXS of whom
18% had epilepsy mostly focal with a pattern of benign epilepsy
with centrotemporal spikes (85). Berry-Kravis (86) reviewed data
of 136 patients with the FXS and found 11% had seizures easily
to control. Twenty of 120 without seizures had EEGs. The most
common EEG abnormality was centrotemporal spikes. Those
with this abnormality had no seizures after childhood (86).
Rett syndrome was described in 1966 by Rett (87) and is the
second most common cause of X-linked mental retardation now
known to be caused by MECP2 gene. It occurs predominantly in
females who have progressive retardation, microcephaly, and
characteristic loss of meaningful use of their hands but with
stereotypical hand movements. Wark (88) studied the EEG in 14
patients, 8 of whom had simultaneous videos. In one younger
patient, the hand movements had a 1:1 time locked relationship
with “rolandic” spikes perhaps eliciting the spikes while the older
patient’s spikes were less frequent with the hand movements.
Focal central spike-wave discharges were more common in the
younger patients while rhythmic spike-wave discharges probably
ictal were more common in the older (Fig. 16.16) (88)
A patient with Rett syndrome confirmed by genetic analysis
as having a nonsense mutation in the MECP2 gene had deteri-
oration of her seizure control at age 6. While her EEG was
“rolandic” at age 2, her EEG at age 6 was considered typical of
the “typical” EEG described for Angelman syndrome (89).
Angelman syndrome described by Angelman consists of
severe mental retardation, microcephaly, ataxic gait, and tremu-
lousness jerky movements described as “puppet-like” (90).
Presenting symptoms are failure to thrive, feeding problems,
Figure 16.16 A 9-year-old girl with autistic spectrum and hand wringing is found to have Rett syn-
drome. EEG: independent centrotemporal spikes right greater than left activated by sleep.
Chapter 16 ■ The EEG in Congenital Malformations 313
developmental delay, seizures, and microcephaly in the first
year of life. Children have an abnormal EEG with diffuse sharp
and slow waves of 2 to 4 Hz, laugh frequently, are overexcitable,
and extremely active (91). The disorder is related to an abnor-
mality of the ubiquitin ligase gene UBE3A located at 15q11.2
(92). In certain brain regions, this gene has monoallelic expres-
sion from maternal chromosome 15. Involvement of three
GABAa receptor subunit genes may explain the severe epilepsy.
The EEG pattern is typical of a Lennox Gastaut-type pattern
with continuous slow spike-wave activity. The EEG patterns are
considered characteristic and able to lead to the correct diagno-
sis. Minassian et al. (93) studied the prolonged video EEGs of
20 patients with the various pheno- and genotypes of
Angelman syndrome. The majority (nine) had maternally
inherited 15q11-13 deletions (Class I) with severe intractable
epilepsy: atypical absences, myoclonus, atonic seizures, and
flexor spasms. Four had uniparental disomy (Class II): two had
myoclonic seizures. Five had methylation imprinting abnor-
malities (Class III): one had mild epilepsy and another frequent
absences and myoclonus, tonic, and atonic seizures. Two had
mutations in the UBE3A gene (Class IV), with infrequent atyp-
ical absence, myoclonus, and tonic seizures, and in 20% a
genetic cause could not be identified. They found that “all had
diffuse bifrontally dominant high amplitude 1–3 Hz notched or
triphasic or polyphasic slow waves or slow and sharp waves.”
The youngest (2 years) Class I had hypsarrhythmia while older
patients with eyes closed had alpha background and a 15-year-
old had hypnagogic hypersynchrony an immature pattern for
drowsing at this age (93).
314 Part III ■ Clinical EEG: General
Korff et al. (94) looked for notched delta at a large tertiary
care children’s hospital and found retrospectively 1.1% of video
EEGs had this pattern. The specificity for Angelman was 38%
with the youngest 14 months of age. Seven of the 14 patients
who had Angelman syndrome features and notched delta did
not appear to have genetic testing for this disorder even though
they were returning for outpatient visits to a tertiary care cen-
ter (94). This chapter highlights how an EEG finding can alert
to a syndrome diagnosis, though of course it is not necessarily
pathonemonic for the syndrome.
Valente et al. (95) reviewed 26 patients with Angelman syn-
drome with 70 EEGs and 15 videos. The major abnormality was
a delta pattern seen in 22 including a hypsarrhythmic variant in
addition to the notched delta. Other less-specific abnormalities
such as theta seen diffusely or posteriorly and posterior dis-
charges were described (95). Laan et al. (96) follow the evolu-
tion of epilepsy and EEG findings in Angelman syndrome in 36
patients (15q11–13 deletion) over a 1 to 39 years follow-up.
Figure 16.17 A 12-year-old girl with Angelman syndrome and seizures. EEG is lacking age-appropriate background and
has high-voltage 2- to 3-Hz delta with notching or spike activity that is diffuse and often not symmetrical.
Thirty (83%) had seizures, 43% febrile as the initial seizure.
Ninety-two percent of the adults had seizures: myoclonic, atyp-
ical absences. Sixty-six percent of the 150 EEGs contained
triphasic anterior delta in 83% of patients. This was found prior
to the diagnoses in almost half. The authors note that high-
amplitude rhythmic 4- to 6-Hz theta up to 12 years in retarded
patients should suggest Angelman syndrome (Fig. 16.17) (96).
Guerrini et al. studied the almost continuous rhythmic
myoclonus involving hands and face accompanied by rhythmic
5- to 10-Hz EEG activity in 11 patients with Angelman syn-
drome and concluded that this is related to a unique pattern of
fast-bursting cortical myoclonus (97).
Down syndrome (trisomy 21) is the most common of the
autosomal trisomy syndromes. The children have a characteristic
facies, mental retardation from mild to severe, and at times dis-
orders of other organ systems. Seizures such as infantile spasms
or the Lennox Gastuat syndrome can occur in the more severely
impaired but the incidence of epilepsy is low (Fig. 16.18) (98). A

study of 18 matched Down syndrome patients with normal con-
trols and FAS found slower activity but 7 out of 8 kept their eyes
open tending to block age-appropriate alpha (25). The EEG in
patients with Down syndrome can be quite normal, especially in
childhood to young adulthood, if they can cooperate and are not
markedly cognitively impaired (Fig. 16.19).
Mental/developmental retardation may be caused by a num-
ber of endocrine and metabolic disorders that at this time in
many countries are screened for at the time of birth such as
hypothyroidism. One still has to be aware of neonates present-
ing with myoclonus and burst suppression EEGs due to nonke-
totic hyperglycinemia, or other inborn errors such as in the
urea cycle, or an infant with infantile spasms with a hypsar-
rhythmic EEG having B6 deficiency or dependency. Other rare
disorders such as glucose transporter type I deficiency syn-
drome can present with intractable seizures early in life with
severely abnormal EEGs. It is beyond the scope of this chapter
to review these entities (99).
ADD/ ADHD
ADD/ADHD is a controversial diagnosis with guidelines avail-
able from the American Academy of Pediatrics and specific
diagnostic criteria under DSM-IV (77). Previous terms now
discarded were minimal brain damage or dysfunction (MBD).
Patients may also have learning disabilities including dyslexia or
behavioral disorders (oppositional defiant disorder) and psy-
chiatric disorders such as bipolar disorder. The diagnosis con-
sists of a list of symptoms noted by parents and teachers with
onset prior to age 7. These are mainly inattention, impulsivity,
and hyperactivity. These symptoms are more common in boys,
are frequently familial, and have a maturational aspect though
some patients never seem to outgrow the disorder (100).
Chapter 16 ■ The EEG in Congenital Malformations 315
Figure 16.18 A 28-year-old man with
Down syndrome and moderate mental
retardation who has intractable drop
seizures associated with a Lennox Gastaut
EEG pattern. EEG: note poor development
of age-appropriate waking rhythms and
slow spike-wave discharges with intermit-
tent attenuations.
Objections have been raised regarding the true scientific
validity of this diagnosis as “the most common neurobehavioral
condition of childhood” not based on the disorder as a disease,
but on a group of symptoms secondary to a result of the behav-
ioral manifestations of a host of psychological, psychiatric, and
educational difficulties. There is a tendency to use apparent
response to medication with stimulants or atomoxetine as a
“proof” of the diagnosis (101).
Specific EEG studies regarding ADD/ADHD have been
unrewarding. Recent literature regarding EEG and ADHD has
focused on quantitative EEG (QEEG). A meta-analysis of these
numerous studies culled down to nine observed increased theta
and decreased beta power compared to controls. As these inves-
tigators observed this change in theta/beta ratio may arise with
other conditions, a prospective study is indicated especially for
control of many aspects such as mental state, alertness, and
medication (102).
Epileptiform abnormalities in the EEGs of children with
ADD/ADHD have been described in the reports of Hughes et
al. (103) and Richer et al. (104). Hughes reported on all
ADHD children without a clinical history of seizures, retarda-
tion, or psychiatric symptoms referred to a children’s hospital
pediatric neurology clinic and found 30.1% had definite
“non- controversial” spike activity. The EEGs usually included
sleep with focal spikes more common than generalized.
Richer’s group reported epileptiform activity in twice as many
children with ADHD as controls 1.3% versus 0.6% mostly
elicited by hyperventilation or photic stimulation. Only 3 of
21 went on to develop epilepsy and they concluded there is no
clinical utility in using EEG to diagnosis epilepsy in children
with ADHD. Of course children with inattention could be
having absence seizures and a careful history is always warran -
ted (Fig. 16.20).
316 Part III ■ Clinical EEG: General

Figure 16.19 A 17-year-old man with Down syndrome who had an episode of loss of consciousness. Patient attends spe-
cial education and is only mildly retarded. His EEG is normal.

Figure 16.20 A 6-year-old boy with

ADHD has no history of seizures, though
he is said to be a restless sleeper. His EEG
shows right centrotemporal spikes in
sleep.

A higher than expected occurrence of ADD/ADHD has been
described in children with epilepsy. In a group of children with
benign epilepsies and otherwise negative neurologic history,
but with 25 males to 15 females, 70% had ADHD by DSM-IV
criteria with it more common in the females (105).
There have been no specific EEG findings currently noted in
children with dyslexia or any of the learning disabilities. There
is controversy regarding the epileptic or spike burden in chil-
dren with epilepsy, even the “benign” asymptomatic types with
or without treatment who do not perform as well scholastically
or on at least some neuropsychological tests as their siblings or
peer group (Fig. 16.21) (106,107).
REFERENCES
1. McAbee GN, Chan A, Erde EL. Prolonged survival with hydranen-
cephaly: report of two patients and literature review. Pediatr
Neurol. 2000;23(1):80–84.
2. Velasco M, Velasco F, Gardea G, et al. Polygraphic characterization
of the sleep–epilepsy patterns in a hydranencephalic child with
severe generalized seizures of the Lennox–Gastaut syndrome. Arch
Med Res. 1997;28(2):297–302.
3. Ines DF, Markand ON. Epileptic seizures and abnormal electroen-
cephalographic findings in hydrocephalus and their relation to the
shunting procedures. Electroencephalogr Clin Neurophysiol.
1977;42(6):761–768.
4. Bartoshesky LE, Haller J, Scott RM, et al. Seizures in children with
meningomyelocele. Am J Dis Child. 1985;139(4):400–402.
5. Saukkonen AL, Serlo W, von Wendt L. Epilepsy in hydrocephalic
children. Acta Paediatr Scand. 1990;79(2):212–218.
6. Noetzel MJ, Blake JN. Seizures in children with congenital hydro-
cephalus: long-term outcome. Neurology. 1992;42(7):1277–1281.
7. Hack CH, Enrile BG, Donat JF, et al. Seizures in relation to shunt
dysfunction in children with meningomyelocele. J Pediatr.
1990;116(1):57–60.
Chapter 16 ■ The EEG in Congenital Malformations 317
Figure 16.21 A 6-year-old girl with a
recent seizure out of sleep after going to
sleep several hours later than usual. Her
10-year-old sister has the same EEG pat-
tern and clinical history. They are in regu-
lar classes in school doing well and are on
no medications. EEG: abundant independ-
ent left centrotemporal spikes and sharp
waves at times reflected to the right hemi-
sphere activated by sleep.
8. Faillace WJ, Canady AI. Cerebrospinal fluid shunt malfunction
signaled by new or recurrent seizures. Childs Nerv Syst. 1990;6(1):
37–40.
9. Sarnat HB, Flores-Sarnat L. Neuropathologic research strategies in
holoprosencephaly. J Child Neurol. 2001;16(12):918–931.
10. Patterson MC. Holoprosencephaly: the face predicts the brain; the
image predicts its function. Neurology. 2002;59(12):1833–1834.
11. Lewis AJ, Simon EM, Barkovich AJ, et al. Middle interhemispheric
variant of holoprosencephaly: a distinct cliniconeuroradiologic
subtype. Neurology. 2002;59(12):1860–1865.
12. Clegg NJ, Gerace KL, Sparagana SP, et al. Holoprosencephaly: a
review. Am J EEG Technol. 2002;42:59–72.
13. Packard AM, Miller VS, Delgado MR. Schizencephaly: correlations
of clinical and radiologic features. Neurology. 1997;48(5):
1427–1434.
14. Granata T, Battaglia G, D’Incerti L, et al. Schizencephaly: neurora-
diologic and epileptologic findings. Epilepsia. 1996;37(12):
1185–1193.
15. Iannetti P, Nigro G, Spalice A, et al. Cytomegalovirus infection and
schizencephaly: case reports. Ann Neurol. 1998;43(1):123–127.
16. Paladin F, Chiron C, Dulac O, et al. Electroencephalographic
aspects of hemimegalencephaly. Dev Med Child Neurol.
1989;31(3):377–383.
17. Aicardi J, Lefebvre J, Lerique-Koechlin A. A new syndrome: spasms
in flexion, callosal agenesis, ocular abnormalities. Electroencephalogr
Clin Neurophysiol. 1965;19:609–610.
18. Fariello RG, Chun RW, Doro JM, et al. EEG recognition of
Aicardi’s syndrome. Arch Neurol. 1977;34(9):563–566.
19. Erenberg G. Aicardi’s syndrome: report of an autopsied case and
review of the literature. Cleve Clin Q. 1983;50(3):341–345.
20. Lacey DJ. Agenesis of the corpus callosum. Clinical features in 40
children. Am J Dis Child. 1985;139(9):953–955.
21. Kolodny EH. Agenesis of the corpus callosum: a marker for inher-
ited metabolic disease? Neurology. 1989;39(6):847–848.
22. Gastaut H, Regis H, Gastaut JL, et al. Lipomas of the corpus callo-
sum and epilepsy. Neurology. 1980;30(2):132–138.
318 Part III ■ Clinical EEG: General
23. Bale JF Jr, Murph JR. Congenital infections and the nervous sys-
tem. Pediatr Clin North Am. 1992;39(4):669–690.
24. Jones KL, Smith DW. Recognition of the fetal alcohol syndrome in
early infancy. Lancet. 1973;2(7836):999–1001.
25. Kaneko WM, Phillips EL, Riley EP, et al. EEG findings in fetal alco-
hol syndrome and Down syndrome children. Electroencephalogr
Clin Neurophysiol. 1996;98(1):20–28.
26. Chasnoff IJ, Griffith DR, Freier C, et al. Cocaine/polydrug use in
pregnancy: two-year follow-up. Pediatrics. 1992;89(2):284–289.
27. Dominguez R, Aguirre Vila-Coro A, Slopis JM, et al. Brain and
ocular abnormalities in infants with in utero exposure to cocaine
and other street drugs. Am J Dis Child. 1991;145(6):688–695.
28. Lairy GC. The evolution of the EEG from birth to adulthood. In:
Handbook of Electroencephalography and Clinical Neurophysiology.
Vol 6B. Amsterdam: Elsevier; 1973:3–117.
29. Jan JE, Wong PKH. Behaviour of the alpha rhythm in electroen-
cephalograms of visually impaired children. Dev Med Child
Neurol. 1988;30(4):444–450.
30. Kriegseis A, Hennighausen E, Rösler F, et al. Reduced EEG alpha
activity over parieto-occipital brain areas in congenitally blind
adults. Clin Neurophysiol. 2006;117(7):1560–1573.
31. Ho SS, Kuzniecky RI, Gilliam F, et al. Congenital porencephaly and
hippocampal sclerosis. Clinical features and epileptic spectrum.
Neurology. 1997;49(5):1382–1388.
32. Gandy SE, Heier LA. Clinical and magnetic resonance features of
primary intracranial arachnoid cysts. Ann Neurol. 1987;21(4):
342–348.
33. Arroyo S, Santamaria J. What is the relationship between arach-
noid cysts and seizure foci? Epilepsia. 1997;38(10):1098–1102.
34. Barkovich AJ, Kuzniecky RI, Jackson GD, et al. Classification sys-
tem for malformations of cortical development: update 2001.
Neurology. 2001;57(12):2168–2178.
35. Tassi L, Colombo N, Garbelli R, et al. Focal cortical dysplasia: neu-
ropathological subtypes, EEG, neuroimaging and surgical out-
come. Brain. 2002;125(Pt 8):1719–1732.
36. Palmini A, Najm I, Avanzini G, et al. Terminology and classification
of the cortical dysplasias. Neurology. 2004;62(6 suppl 3):S2–S8.
37. Krsek P, Maton B, Korman B, et al. Different features of
histopathological subtypes of pediatric focal cortical dysplasia.
Ann Neurol. 2008;63(6):758–769.
38. Kim DW, Lee SK, Chu K, et al. Predictors of surgical outcome and
pathologic considerations in focal cortical dysplasia. Neurology.
2009;72(3):211–216.
39. d’Orsi G, Tinuper P, Bisulli F, et al. Clinical features and long term
outcome of epilepsy in periventricular nodular heterotopia.
Simple compared with plus forms. J Neurol Neurosurg Psychiatry.
2004;75(6):873–878.
40. Battaglia G, Chiapparini L, Franceschetti S, et al. Periventricular
nodular heterotopia: classification, epileptic history, and genesis of
epileptic discharges. Epilepsia. 2006;47(1):86–97.
41. Sisodiya SM, Free SL, Thom M, et al. Evidence for nodular epilep-
togenicity and gender differences in periventricular nodular het-
erotopia. Neurology. 1999;52(2):336–341.
42. Sicca F, Kelemen A, Genton P, et al. Mosaic mutations of the LIS1
gene cause subcortical band heterotopia. Neurology. 2003;61(8):
1042–1046.
43. Mai R, Tassi L, Cossu M, et al. A neuropathological, stereo-EEG,
and MRI study of subcortical band heterotopia. Neurology.
2003;60(11):1834–1838.
44. Ono J, Mano T, Andermann E, et al. Band heterotopia or double
cortex in a male: bridging structures suggest abnormality of the
radial glial guide system. Neurology. 1997;48(6):1701–1703.
45. Fong J, Wu G, Wyllie E, et al. Interictal hypermetabolic subcortical
band on brain FDG–PET in doublecortin mutation. Neurology.
2008;71(7):535.
46. Dobyns WB, Andermann E, Andermann F, et al. X-linked mal-
formations of neuronal migration. Neurology. 1996;47(2):
331–339.
47. Gastaut H, Pinsard N, Raybaud C, et al. Lissencephaly
(agyria–pachygyria): clinical findings and serial EEG studies. Dev
Med Child Neurol. 1987;29(2):167–180.
48. de Rijk van Andel JF, Arts WF, de Weerd AW. EEG in type I
lissencephaly [letter to the editor]. Dev Med Child Neurol.
1988;30(1):126–127.
49. Sébire G, Goutieres F, Tardieu M, et al. Extensive macrogyri or no
visible gyri: distinct clinical, electroencephalographic, and genetic
features according to different imaging patterns. Neurology.
1995;45(6):1105–1111.
50. Kuzniecky R, Andermann F, Guerrini R. Congenital bilateral peri-
sylvian syndrome: study of 31 patients. The CBPS Multicenter
Collaborative Study. Lancet. 1993;341(8845):608–612.
51. Chang BS, Piao X, Giannini C, et al. Bilateral generalized polymi-
crogyria (BGP): a distinct syndrome of cortical malformation.
Neurology. 2004;62(10):1722–1728.
52. Guerrini R, Genton P, Bureau M, et al. Multilobar polymicrogyria,
intractable drop attack seizures, and sleep-related electrical status
epilepticus. Neurology. 1998;51(2):504–512.
53. Young H, Hyman S, North K. Neurofibromatosis 1: clinical review
and exceptions to the rules. J Child Neurol. 2002;17(8):613–621
[discussion 27–29, 46–51].
54. Korf BR, Carrazana E, Holmes GL. Patterns of seizures observed in
association with neurofibromatosis 1. Epilepsia. 1993;34(4):
616–620.
55. Kulkantrakorn K, Geller TJ. Seizures in neurofibromatosis 1.
Pediatr Neurol. 1998;19(5):347–350.
56. Vivarelli R, Grosso S, Calabrese F, et al. Epilepsy in neurofibro-
matosis 1. J Child Neurol. 2003;18(5):338–342.
57. Crino PB, Nathanson KL, Henske EP. The tuberous sclerosis com-
plex. N Engl J Med. 2006;355(13):1345–1356.
58. Pampiglione G, Moynahan EJ. The tuberous sclerosis syndrome:
clinical and EEG studies in 100 children. J Neurol Neurosurg
Psychiatry. 1976;39(7):666–673.
59. Muzykewicz DA, Costello DJ, Halpern EF, et al. Infantile spasms in
tuberous sclerosis complex: prognostic utility of EEG. Epilepsia.
2008;50(2):290–296.
60. Mackay MT, Becker LE, Chuang SH, et al. Malformations of corti-
cal development with balloon cells: clinical and radiologic corre-
lates. Neurology. 2003;60(4):580–587.
61. Jarrar RG, Buchhalter JR, Raffel C. Long-term outcome of epilepsy
surgery in patients with tuberous sclerosis. Neurology. 2004;62(3):
479–481.
62. Jansen FE, Braams O, Vincken KL, et al. Overlapping neurologic
and cognitive phenotypes in patients with TSC1 or TSC2 muta-
tions. Neurology. 2008;70(12):908–915.
63. Jansen FE, Vincken KL, Algra A, et al. Cognitive impairment in
tuberous sclerosis complex is a multifactorial condition.
Neurology. 2008;70(12):916–923.
64. Major P, Rakowski S, Simon MV, et al. Are cortical tubers epilepto-
genic? Evidence from electrocorticography. Epilepsia. 2009;50(1):
147–154.
65. Thomas-Sohl KA, Vaslow DF, Maria BL. Sturge–Weber syndrome:
a review. Pediatr Neurol. 2004;30(5):303–310.
66. Comi AM. Advances in Sturge–Weber syndrome. Curr Opin
Neurol. 2006;19(2):124–128.

67. Hennel SJ, Ekert PG, Volpe JJ, et al. Insights into the pathogenesis
of cerebral lesions in incontinentia pigmenti. Pediatr Neurol.
2003;29(2):148–150.
68. Placantonakis DG, Ney G, Edgar M, et al. Neurosurgical manage-
ment of medically intractable epilepsy associated with hypome-
lanosis of Ito. Epilepsia. 2005;46(2):329–331.
69. Menascu S, Donner EJ. Linear nevus sebaceous syndrome: case
reports and review of the literature. Pediatr Neurol. 2008;38(3):
207–210.
70. Guillet R, Kwon J. Seizure recurrence and developmental
disabilities after neonatal seizures: outcomes are unrelated to
use of phenobarbital prophylaxis. J Child Neurol. 2007;22(4):
389–395.
71. Sinclair DB, Campbell M, Byrne P, et al. EEG and long-term out-
come of term infants with neonatal hypoxic-ischemic
encephalopathy. Clin Neurophysiol. 1999;110(4):655–659.
72. Mutch L, Alberman E, Hagberg B, et al. Cerebral palsy epidemiol-
ogy: where are we now and where are we going? Dev Med Child
Neurol. 1992;34(6):547–551.
73. Russman BS, Ashwal S. Evaluation of the child with cerebral palsy.
Semin Pediatr Neurol. 2004;11(1):47–57.
74. Gurses C, Gross DW, Andermann F, et al. Periventricular leukoma-
lacia and epilepsy: incidence and seizure pattern. Neurol.
1999;52(2):341–345.
75. Kuchukhidze G, Unterberger I, Dobesberger J, et al. Electroclinical
and imaging findings in ulegyria and epilepsy: a study on 25
patients. J Neurol Neurosurg Psychiatry. 2008;79(5):547–552.
76. Usui N, Mihara T, Baba K, et al. Posterior cortex epilepsy second-
ary to ulegyria: is it a surgically remediable syndrome? Epilepsia.
2008;49(12):1998–2007.
77. American Psychiatric Association. Diagnostic and Statistical
Manual of Mental Disorders. 4th ed. Washington, DC: American
Psychiatric Association; 1994.
78. Rapin I. Autism. N Engl J Med. 1997;337(2):97–104.
79. Kagan-Kushnir T, Roberts SW, Snead OC 3rd. Screening elec-
troencephalograms in autism spectrum disorders: evidence-based
guideline. J Child Neurol. 2005;20(3):197–206.
80. Canitano R, Luchetti A, Zappella M. Epilepsy, electroencephalo-
graphic abnormalities, and regression in children with autism. J
Child Neurol. 2005;20(1):27–31.
81. Steffenburg U, Hagberg G, Kyllerman M. Characteristics of
seizures in a population-based series of mentally retarded children
with active epilepsy. Epilepsia. 1996;37(9):850–856.
82. Hamdan FF, Gauthier J, Spiegelman D, et al. Mutations in SYN-
GAP1 in autosomal nonsyndromic mental retardation. N Engl J
Med. 2009;360(6):599–605.
83. Martin JP, Bell J. A pedigree of mental defect showing sex linkage.
J Neurol Neurosurg Psychiatry. 1943;6:154–156.
84. Lubs HA. A marker X chromosome. Am J Hum Genet. 1969;21(3):
231–244.
85. Musumeci SA, Hagerman RJ, Ferri R, et al. Epilepsy and EEG find-
ings in males with fragile X syndrome. Epilepsia. 1999;40(8):
1092–1099.
86. Berry-Kravis E. Epilepsy in fragile X syndrome. Dev Med Child
Neurol. 2002;44(11):724–728.
Chapter 16 ■ The EEG in Congenital Malformations 319
87. Rett A. On a unusual brain atrophy syndrome in hyperammone-
mia in childhood. Wien Med Wochenschr. 1966;116(37):723–726.
88. Wark J. The electroencephalogram in Rett syndrome: correlation
with hand movements. J Child Neurol. 1993;8(1):101–105.
89. Laan LA, Vein AA. A Rett patient with a typical Angelman EEG.
Epilepsia. 2002;43(12):1590–1592.
90. Angelman H. “Puppet children”. A report of three cases. Dev Med
Child Neurol. 1965;7:681–688.
91. Zori RT, Hendrickson J, Woolven S, et al. Angelman syndrome:
clinical profile. J Child Neurol. 1992;7(3):270–280.
92. Williams CA. Neurological aspects of the Angelman syndrome.
Brain Dev. 2005;27(2):88–94.
93. Minassian BA, DeLorey TM, Olsen RW, et al. Angelman syn-
drome: correlations between epilepsy phenotypes and genotypes.
Ann Neurol. 1998;43(4):485–493.
94. Korff CM, Kelley KR, Nordli DR Jr. Notched delta, phenotype, and
Angelman syndrome. J Clin Neurophysiol. 2005;22(4):238–243.
95. Valente KD, Andrade JQ, Grossmann RM, et al. Angelman syn-
drome: difficulties in EEG pattern recognition and possible mis-
interpretations. Epilepsia. 2003;44(8):1051–1063.
96. Laan LA, Renier WO, Arts WF, et al. Evolution of epilepsy and
EEG findings in Angelman syndrome. Epilepsia. 1997;38(2):
195–199.
97. Guerrini R, De Lorey TM, Bonanni P, et al. Cortical myoclonus
in Angelman syndrome. Ann Neurol. 1996;40(1):39–48.
98. Stafstrom CE, Patxot OF, Gilmore HE, et al. Seizures in children
with Down syndrome: etiology, characteristics and outcome. Dev
Med Child Neurol. 1991;33(3):191–200.
99. Wolf NI, Bast T, Surtees R. Epilepsy in inborn errors of metabo-
lism. Epileptic Disord. 2005;7(2):67–81.
100. Wolraich ML. Attention-deficit hyperactivity disorder. Semin
Pediatr Neurol. 2006;13(4):279–285.
101. Furman L. What is attention-deficit hyperactivity disorder
(ADHD)? J Child Neurol. 2005;20(12):994–1002.
102. Snyder SM, Hall JR. A meta-analysis of quantitative EEG power
associated with attention-deficit hyperactivity disorder. J Clin
Neurophysiol. 2006;23(5):440–455.
103. Hughes JR, DeLeo AJ, Melyn MA. The electroencephalogram in
attention deficit-hyperactivity disorder: emphasis on epilepti-
form discharges. Epilepsy Behav. 2000;1(4):271–277.
104. Richer LP, Shevell MI, Rosenblatt BR. Epileptiform abnormalities
in children with attention-deficit-hyperactivity disorder. Pediatr
Neurol. 2002;26(2):125–129.
105. Zelnik N, Sa’adi L, Silman-Stolar Z, et al. Seizure control and
educational outcome in childhood-onset epilepsy. J Child Neurol.
2001;16(11):820–824.
106. Riva D, Vago C, Franceschetti S, et al. Intellectual and language
findings and their relationship to EEG characteristics in benign
childhood epilepsy with centrotemporal spikes. Epilepsy Behav.
2007;10(2):278–285.
107. Nicolai J, van der Linden I, Arends JB, et al. EEG characteristics
related to educational impairments in children with benign
childhood epilepsy with centrotemporal spikes. Epilepsia.
2007;48(11):2093–2100.

Brain Tumors and Other
Space-Occupying Lesions
ADAM L. HARTMAN AND RONALD P. LESSER
GENERAL OVERVIEW
Brain tumors are a leading cause for new-onset seizures in the
adults and many patients still present with seizures as their first
manifestation of a tumor (Fig. 17.1). Although EEG at one time
was important for diagnosing brain tumors and other space-
occupying lesions, this is now more commonly done using
imaging studies, such as computed tomography (CT) and mag-
netic resonance imaging (MRI) (Fig. 17.2). Clinical neurophys-
iology still is important in managing these patients, including
the use of electrocorticography (ECoG) during testing to iden-
tify eloquent cortex during resection surgeries, application of
evoked potentials in assessing the location of sensorimotor cor-
tex or the extent of tumor involvement, and the application of
magnetoencephalography, for both magnetic source imaging
(e.g., in localizing spike-generating zones) and functional map-
ping. These topics will be discussed in this chapter.
The EEG findings discussed in this chapter are not specific
for tumor type and also can occur with other space-occupying
lesions (hence, their inclusion in this chapter). In a series of
classical experiments Gloor et al. (1) found that focal delta-
range slowing was seen over well-circumscribed lesions in the
white matter (with no apparent change in faster frequency
activity) or after a focal thalamic lesion, while diffuse hemi-
spheric slowing was seen after larger thalamic and/or hypothal-
CHAPTER
17
amic lesions or lesions causing cerebral edema; generalized
slowing was seen after lesions were made in the midbrain
tegmental area. Solid tumors generally are not epileptogenic in
themselves; it is the infiltrated or surrounding nontumor tissue
that causes the seizures (2,3). Notable exceptions include devel-
opmental tumors such as hypothalamic hamartomas (HH; dis-
cussed in the section “Hypothalamic Hamartoma”). EEG
findings in most nondevelopmental tumors likely are caused by
displacement of normal tissue (either by the mass itself or by
altered flow of cerebrospinal fluid), physical disruption of
normal neuronal circuitry, ischemia, or hemosiderin deposition
(4,5), gap-junction proteins (6), and alterations in neurotrans-
mitter pathophysiology, such as vesicular glutamate trans-
porters (7), both ionotropic and metabotropic glutamate
receptor subunits (8), and GABA receptors (9). Data on
patients with brain tumors provide some of the most convinc-
ing arguments for secondary epileptogenesis (wherein an
epileptogenic region gives rise to a secondary epileptogenic
focus via strong connectivity between the two regions) (10).
Epidemiology
The prevalence of seizures depends in part on type of tumor:
for example, anywhere from 66% to 90% of patients with
low-grade astrocytomas have seizures at some point (11).

Figure 17.1 EEG from a 4-year-old male who presented to Pediatric Epilepsy Clinic with a 2-month history of episodes of
staring and fumbling with the fingers that lasted about 30 seconds. The EEG showed a burst of slowing in the left hemi-
sphere (left figure, fourth second) and runs of periodically repeating high-voltage rhythmical sharp activity over the left
hemisphere region during sleep (right figure). A longitudinal bipolar montage is shown.

321
322 Part III ■ Clinical EEG: General

Figure 17.2 MRI of the patient from Figure 17.1.

Axial sections (FLAIR and T2-weighted sequences)
at the level of the temporal lobes show cystic
changes, swelling, and infiltration of normal brain
parenchyma in the left anterior mesial temporal
lobe. The patient had an MRI soon after his clinic
appointment, underwent a near-total resection of a
WHO grade II astrocytoma, and has remained
seizure-free since surgery.

Seizures are more common in patients with low-grade (vs.
high-grade) tumors (12). They are the presenting symptom in
up to 76% of children with benign astrocytic and oligoden-
drocytic tumors of the cerebral hemispheres (13). Seizures
occurred in 40% of children with hemispheric brain tumors
(with the temporal lobe being most commonly involved),
while none with thalamic tumors had seizures (14). There was
a focal component to the majority of the seizures in that
series, although nearly 25% of the patients had only general-
ized seizures. Seizures can occur in children with supratento-
rial or infratentorial tumors, although they are more
commonly the presenting sign in hemispheric tumors (15).
Conversely, only up to 11% of patients with a “possible
seizure” have brain tumors diagnosed (16).
LOCALIZATION-RELATED ISSUES
Hemispheric Tumors
Cortical tumors produce abnormal EEGs 96% of the time,
including focal delta activity (correctly localized) (Figs. 17.1 and
17.2); a minority causes focal dysrhythmia, but EEGs can be
falsely localizing (Figs. 17.3 and 17.4) (20). EEGs are abnormal
in up to 87% of children with benign astrocytic and oligoden-
drocytic tumors of the cerebral hemispheres (13). EEG findings
are generally related to location of the tumor and rapidity of
growth. Most (4,21), but not all (22), studies associate rapidly

Historical Perspectives on the Utility of EEG

Before the advent of modern imaging, EEG was a critical com-
ponent in the evaluation of patients with suspected masses,
along with pneumoencephalography, ventriculography, and
angiography. In fact, serial EEGs were somewhat useful in dis-
tinguishing strokes (where EEG abnormalities resolved over
time) from tumors (where EEG abnormalities became pro-
gressively worse) (4). Postoperative EEGs were used to moni-
tor tumor recurrence, which was heralded by focal arrhythmic
slow activity at the former tumor site; a recent study found
that EEG slowing indicated that postoperative seizure control
was less likely (17). Much of the literature on EEG in tumors
and other space-occupying lesions is older and studies were
performed without the modern standards of statistical rigor.
The classification of tumors also has evolved over time, so we Figure 17.3 EEG from the end of a seizure in a 79-year-old male with
attempted to be as nonspecific as possible in describing tumor a right frontal astrocytoma. Phase reversals are present in the right
type in older studies, while using current tumor nomenclature frontal, temporal, and parietal leads. Also, there is a prominent phase
for more recent studies. One final caveat is that many reversal in the left parietal region (channels 12 and 13). However, with
observers have noted that the routine EEG can be normal or the transverse montage used, the channels with the most prominent
even misleading in terms of lateralization (14,15,18,19), thus activity only indicate where the differences between the recording elec-
limiting the utility of this test in some instances. In fact, nor- trodes are the greatest. These are not necessarily were the actual ampli-
mal frequency activity may be seen ipsilateral to a tumor (18). tudes are greatest.
 Chapter 17 ■ Brain Tumors and Other Space-Occupying Lesions 323

growing tumors with very slow delta-range activity, and more sharp waves occurred in half of these patients. In over 90% of
slowly growing tumors with arrhythmic theta-range activity temporal gliomas, slowing was continuous (27). A more recent
with occasional intermixed epileptiform spikes and sharp series (which included prolonged video–EEG monitoring)
waves. It should be noted that up to one third of patients with noted temporal epileptiform discharges and focal temporal
brain tumors have an epileptogenic focus remote from the slowing as the most common finding with temporal lobe
tumor site, which may give rise to some confusion in localiza- tumors, although a few patients also showed bilateral, general-
tion (10). Meningiomas and other slow-growing or midline ized, or false localization (28).
tumors may be difficult to localize, although hyperventilation
may reveal a focus of slowing not seen otherwise (20,23). Focal Posterior Fossa and “Deep” Tumors
theta-range activity can occur with any type of tumor and one Because most tumors in children over 1 year of age are infraten-
series noted it in 75% of patients with hemispheric tumors torial, it is important to consider posterior fossa and so-called
(18). In that study, focal delta-range activity was most common “deep” (e.g., sellar, chiasmal, and diencephalic) tumors separately.
in patients with glioblastoma multiforme, although it also Although most of these patients have had an imaging study, the
occurred with meningeal tumors. Activity in the delta range presenting symptoms can be so nonspecific (e.g., altered mental
was less frequent with astrocytomas, where focal sharp activity status) that an EEG might be obtained for other reasons in the
was more common. Focal epileptiform activity on the EEG course of a patient’s treatment (e.g., to rule out nonconvulsive
appeared to be correlated generally with clinical seizures in status epilepticus). Therefore, it is important to understand
patients with astrocytomas, but clinical seizures were not as EEG findings in this context so that an appropriate interpreta-
common among patients with meningeal tumors, even though tion of EEG patterns can be made. One found that over three
they also had a fair amount of focal epileptiform activity on fourth of records was abnormal in children with posterior fossa
EEG, seizures were not as common. Newmark et al. found that tumors, with likelihood of abnormality decreasing with age
epileptiform activity was unrelated to focal slowing in patients (from below 10 to over 50 years) (23). However, two other stud-
with gliomas (22). Patients with multiple metastases also may ies found EEGs to be normal in 19% (15) and 24% (19) of chil-
have a good deal of epileptiform activity in their EEGs; over dren. Location within the posterior fossa may be important.
90% of patients with metastatic disease have abnormal EEGs EEGs were abnormal in only 30% of patients with brainstem
including epileptiform activity (4), delta-range activity, and tumors, while over 80% of EEGs were abnormal in those with
dysrhythmias (20).
Recording from the cortical surface shows similar findings.
While focal delta-range activity and/or background attenuation
may be seen with some hemispheric tumors, focal slowing can
be seen from an area literally centimeters away from the actual
tumor (24). Those authors suggested that one potentially help-
ful localizing sign is resistance to afterdischarge occurrence,
even at high current intensities, during cortical stimulation.
Focal delta- and theta-range activity correlates with white
matter involvement of tumors (1,22). This includes tumors
involving both gray and white matter as well as those involving
white matter only. Rhythmic delta activity may indicate
involvement of the thalamus, although involvement of deep
frontal white matter may produce similar findings (22).
Attenuation of background activity is seen in patients with
high-grade gliomas that have thalamic involvement (22) but
also may be seen in regions with extensive peritumoral edema
(25), although this is a nonspecific finding. Irregular delta-
range activity is unrelated to peritumoral edema (22) and the
degree of EEG abnormality does not appear to correlate with
degree of edema (25).
Tumor was present in 18% of series of 282 patients showing
typical periodic lateralized epileptiform discharges (PLEDs) (26).
Frontal tumors tend to produce the slow patterns noted pre-
viously, at times showing frontal intermittent rhythmic dis-
charges (FIRDA). Parietal and occipital tumors may affect the Figure 17.4 EEG from a 56-year-old female with a burst of high-volt-
posterior basic rhythm, but also may show abnormalities in age sharply contoured delta-range activity that was bilateral (longitudi-
more anterior regions, possibly leading to false localization nal bipolar recordings, channels 1 to 8) but maximal over the left
(Fig. 17.4). Temporal tumors can show a unilateral abnormality hemisphere (channels 11 to 16). Her underlying pathology was a left
up to 89% of the time (Fig. 17.1) (27). Focal delta activity or parietal astrocytoma.
324 Part III ■ Clinical EEG: General
cerebellar or fourth ventricle tumors (19). In that series, 27%
had posterior rhythmic delta waves, 32% had generalized bilat-
eral bursts of rhythmic slowing, 51% had posterior arrhythmic
delta waves, and 11% had rhythmic theta or delta waves on ver-
tex or anterior quadrants.
Both Bickford and Martinius et al. noted the suppression of
abnormal rhythmic delta-range activity on arousal in patients
with deep tumors (19,20). Because these patterns may be pro-
duced by tumors distant from where the abnormal EEG pat-
terns are recorded, they are sometimes referred to as
“projected” rhythms. EEG may not lateralize, or may falsely lat-
eralize, posterior fossa tumors (19). EEG abnormalities (poste-
rior slow activity) were more common in patients with
posterior fossa tumors and evidence of increased intracranial
pressure (i.e., evidence of third ventricle dilation) (19).
Arrhythmic delta activity is more common in patients with
more rapid progression of symptoms, possibly due to rapid
expansion of tumor size (19). Nearly half the EEGs of patients
with tumors of the cerebellopontine angle (CPA) were normal
in one series (23).
Sellar Tumors
EEG abnormalities in tumors of the sellar region include tem-
poral lobe abnormalities, unilateral delta-range activity, and
bitemporal dysrhythmia. It is important to remember this in
the differential diagnosis of temporal lobe abnormalities. In
tumors that compressed the third ventricle, generalized slowing
was noted, and degree of compression was the only factor that
correlated with abnormalities in the EEG (29). In that series,
EEG abnormalities did not predict tumor type (29).
Hypothalamic Hamartoma
HH is a developmental malformation that frequently presents
with seizures, including infantile spasms. Other associated con-
ditions may include precocious puberty and behavioral prob-
lems. An excellent recent review of the topic is available (30). In
children, gelastic seizures are most typical at onset, although
this may not be the case in adults, who tend to present with par-
tial-onset seizures (31). The other main cause of gelastic
epilepsy is temporal or frontal lobe epilepsy (32). The EEG in
HH initially may be normal, similar to other deep-seated
masses, although over time it may evolve through the appear-
ance of focal (partial onset), and then generalized seizures, con-
sistent with the clinical semiology (30). There may be associated
autonomic features (33). Focal slowing or epileptiform activity
over frontal and/or temporal head regions may be the initial
appearance on EEG, although this evolves into bilateral spike
wave over time if untreated (30). Eventually, a pattern consist-
ing of generalized slow spike wave, paroxysmal fast, and elec-
trodecremental patterns can occur, reminiscent of the EEG in
Lennox Gastaut syndrome (34). There is a possibility that the
hamartoma in some way generates abnormal activity that prop-
agates through the cortex, leading to the various seizure types
noted in HH (30,35). In four patients with HH presenting
epileptic manifestations and displaying interictal spikes over
the frontal and temporal areas, EEG source analysis, based on
scalp recordings (32 electrodes), was able to estimate that the
epileptiform spikes have deep sources in the neighborhood of
the hamartoma, with later spread to cortical areas (36). A
subsequent study from the same group of a patient with HH
and gelastic epilepsy using simultaneous EEG and fMRI record-
ings of several seizures indicated that the epileptic activity
appeared to originate in the area around the tumor and propa-
gate through the left fornix to the temporal lobe, and later
through the cingulate fasciculus to the left frontal lobe (37).
ELECTROPHYSIOLOGY IN SURGICAL
NEURO-ONCOLOGY
Electrophysiology retains a vital role in the surgical manage-
ment of tumors and other space-occupying lesions. One of the
most challenging surgical decisions is balancing the desire to
resect as much tumor as possible with the desire to leave intact
tissues important to motor, sensory, or language-related func-
tions. Certain masses, particularly tumors with a tendency to
infiltrate otherwise normal tissue (e.g., glioblastoma multi-
forme), pose a real dilemma for the surgeon and oncology
team, requiring decisions that may leave the patient either with
more residual tumor or without functions critical to optimal
daily living. Neurophysiology can offer guidance in defining
two important regions: the peritumoral epileptogenic zone and
functional cortex (i.e., areas that typically are critical to motor,
sensory, or language function and whose resection would lead
to significant deficits). Simple lesionectomies also are appropri-
ate in the setting of certain well-defined lesions (e.g., caver-
nomas) or those near noneloquent cortex. Data from some
series (38,39) suggest lesionectomies may be adequate in some
cases, although selection of candidates for this approach has not
been thoroughly studied; success in these cases suggests that
disruption of an epileptogenic network may suffice to suppress
seizure activity (40).
One application of electrophysiology in the operating room
is in the resection of tumors that are associated with epilepsy,
such as certain astrocytomas, dysembroplastic neuroepithelial
tumors (DNET), gangliogliomas, and gangliocytomas. These
tumors are associated with epileptogenic tissue at their margins,
and although the tumors themselves are not typically aggressive,
there can be epileptogenic tissue adjacent to the tumor (41).
Intraoperative ECoG using standard carbon-tipped electrodes
or electrodes designed for subdural recordings can help define
the extent of the epileptogenic zone, even beyond the tumor
margin (41). In this series, seizure control was achieved in one
patient after epileptogenic cortex was resected, even though not
all the tumor was removed, although this may have been an
exceptional case. Additionally, the authors point out this strategy
may be more useful in children than adults. Given tumor recur-
rence rates in some series (42), some suggest that monitoring be
done routinely in cases involving tissue near motor, sensory, or
language areas. However, extraoperative recordings from
patients with subdural electrodes show that epileptiform activ-
ity can be intermittent, and so might be missed during the lim-
ited recording time available in the operating room.
Cortical mapping of motor, sensory, or language cortex
can be done intraoperatively (43), or outside the operating
 Chapter 17
room (44), depending on the goals of the surgery. Each tech-
nique has its advantages and applications. Intraoperative
mapping may reduce the risks of inserting and maintaining
subdural electrodes (such as infection and bleeding) and is
less expensive, but extraoperative mapping allows for more
time to better define the regions important for specific func-
tions, while minimizing operative time when tissue is
removed and minimizing the chances of a stimulation-
induced intraoperative seizure (40). If seizures are not a sig-
nificant concern, then intraoperative mapping is more
straightforward, since only EEG slowing or suppression need
be sought. If intractable epilepsy is the major concern, we tend
to consider the case operation as an epilepsy (rather than
tumor) evaluation and consider extraoperative EEG recording
and mapping, but others have good results with intraoperative
mapping (44–46). Mapping using evoked potentials is dis-
cussed later. The technical aspects of extraoperative mapping
are discussed elsewhere in this book (Chapter 55).
Invasive recordings from patients with tumors have rein-
forced some findings from scalp recordings but also have
identified a number of additional interesting phenomena.
Patients with bilateral interictal findings can have unilateral
ictal onsets and good postoperative outcomes (28). In one
series, ECoG during resection of gangliogliomas showed high-
voltage slowing in the area of the tumor (47). Another series
using stereo-EEG in patients with low-grade tumors showed
activity less than 10 V in white matter that was infiltrated
with tumor or only affected by edema in some cases, while
high-voltage slowing was noted in some patients with simi-
larly affected gray matter, or in peritumoral gliosis (some
specimens also had normal activity) (2). Solid tumors always
showed electrical depression. Intraoperative mapping has
demonstrated acute changes in local motor maps right after
tumor resection (46).
EVOKED POTENTIALS IN THE TREATMENT
OF BRAIN TUMORS
Electrophysiologic techniques such as evoked potentials were
once used to help diagnose and locate mass lesions. This has
largely been replaced by imaging. However, electrophysiology,
used as a test of function, retains an important role in the man-
agement of mass lesions, particularly in the operating room.
The techniques are not specific to the management of tumors,
but instead have been developed to help insure that surgical
procedures are performed as safely as possible.
Cortical and Subcortical
Preoperative evoked potentials can be used to document
whether sensory pathways are intact prior to, or at the start of,
surgery. During surgery, changes in response can signal to the
surgeon that the surgery might be impinging on the monitored
pathway. The surgeon can then modify the surgical approach
and extent of surgical resection. However, a change in responses
may be irreversible once seen. Moreover, on occasion there can
be postoperative deficits even though intraoperative evoked
■ Brain Tumors and Other Space-Occupying Lesions 325
potentials did not change (48). Also, phase reversal of
somatosensory evoked potentials (SEPs) across the rolandic
sulcus helps the surgeon locate, and therefore avoid, the pri-
mary sensorimotor cortex (49,50). Motor cortex stimulation,
and recording of evoked electromyographic changes peripher-
ally, is another way to locate motor cortex (50). Responses to
motor cortex stimulation are sensitive to anesthesia however,
and this restricts its use in some circumstances.
Posterior Fossa Surgery
Electrophysiologic are no longer a primary method for locating
posterior fossa tumors, but they can help to determine the rela-
tionship of the mass to auditory pathways (51). BAEPs fre-
quently are lost with CPA tumors, due to effects on either the
peripheral or brainstem portions of the auditory pathway
(52,53). Possible changes include decreased amplitude and
increased latency (51–54). As would be expected, BAEPs are not
as sensitive as MRI in patients with small CPA tumors (55), or
tumors that do not impinge on the auditory pathway (56,57).
In 40 patients with acoustic neuromas, Tanaka et al. (58)
reported that 6 patients with normal ABRs using standard test-
ing methods had abnormalities when the clicks were given at
high stimulus rates.
BAEPs can be used alone or together with other modalities
to assess the effects of tumors on the brainstem, or to monitor
during surgery (51). For example, some have used recordings of
the blink or masseter reflex (53,59). During posterior fossa sur-
gery, other cranial nerves, including III, IV, V, VI, VII, IX, X, XI,
and XII, have been monitored, along with muscles innervated
by these nerves, and along with SEPs (51,60). BAEPs along with
EMG monitoring can help insure safe microvascular decom-
pression procedures in the posterior fossa. When SEPs are used
to monitor during posterior fossa surgery, it should be kept in
mind that the usually recorded SEP responses to peripheral
nerves are generated above and below, but not within, the
brainstem (61).
Magnetoencephalography
Nakasato et al. (62) evaluated cortical auditory function in 14
patients with temporal lobe tumors, using an MRI-linked
whole-head MEG system. The authors suggested that the MEG
system could be used to evaluate cortical auditory function
noninvasively before and after surgical treatment of temporal
lobe tumors. An example of the usefulness of MEG in the pre-
operative evaluation of tumors using somatosensory evoked
fields (SEFs) is presented in Chapter 42 and illustrated in
Figure 42.27.
Event-related Potentials
Moore et al. (63) evaluated 33 survivors of childhood cancer
with P300, motor reaction time, and neuropsychiatric testing.
There were abnormalities in P300 and in neuropsychiatric
scores of patients who had received CRT. The authors suggested
that cognitive changes in this group might be due to white mat-
ter damage. As with other evoked potentials, changes in P300
are nonspecific with respect to etiology, but may correlate with
overall cognitive status.
326 Part III ■ Clinical EEG: General
EEG COHERENCE
EEG coherence, a measure of the association between signals in
a specified frequency band from different brain regions, has
been used in both surface and invasive recordings. One study
that used surface recordings found decreased coherence
between the cortical region containing a brain lesion, and other
portions of the same hemisphere (64). Another study examined
the use of coherence to assess the central sulcus and sensorimo-
tor area during tumor resections in 10 patients (65). The central
sulcus was a region of low coherence, whereas high coherence
was noted in the region of the brain tumor. In contrast, coher-
ence measures were not helpful in the temporal lobe. Further
work would be needed to resolve the discrepancies between
these results.
EEG IN BRAIN ABSCESS
Despite the hope expressed in the last edition of this text that
brain abscess would become a rare diagnosis, the surge in
patients with acquired immune deficiency syndrome, increased
numbers of patients on immunomodulators or immunosup-
pression after organ transplants or the diagnosis of autoim-
mune disease, and increased recognition of congenital
immunodeficiencies have led to a resurgence of this condition
(Chapter 15, p. 314 of the 5th edition). More recent case reports
and series of patients with brain abscesses mention both epilep-
tiform activity (66,67) and PLEDs (68). One series showed that
14 of 45 patients had focal EEG abnormalities (not specified
further), while an additional 21 had either regional or hemi-
spheric disturbances; of 3 normal EEGs, 2 patients had cerebel-
lar abscesses (69).
Michel et al. (70) studied 13 patients with brain abscess and
found slowing of background activity in the hemisphere with
the abscess in 6 cases and bilateral generalized slowing in
another 4 (3 were normal), continuous and nonreactive poly-
morphic delta-range activity in 12 patients with frontal inter-
mittent delta-range activity in 5 patients, and pseudoperiodic
sharp waves in 2 patients. The EEG localized the abscess in
61.5% of their cases. Focal arrhythmic delta-range activity was
the most consistent finding in another series of 13 patients (71).
Importantly, although EEG was localizing in most cases, it was
falsely localizing in one patient with a cerebellar abscess, per-
haps of the connections between each cerebellar hemisphere
and the contralateral cerebral hemisphere. Epileptiform activity
occurred in four cases. Asymmetry in distribution of beta-
range activity helped lateralize the hemisphere with the abscess
in seven. In summary, EEG may be useful in localizing cortical
abscesses in many cases, but may be misleading or even normal
in others.
EEG IN SUBDURAL HEMATOMA
Just as in tumors, the typical first diagnostic study obtained in
most patients with intracranial hemorrhages is either a CT
scan or, less commonly, an MRI. Nonetheless, persistent altered
mental status after evacuation or management of a hemor-
rhage may be due to nonconvulsive status epilepticus, or to a
postictal state; in either case, an EEG may be obtained.
Intracranial hemorrhages may include epidural hematomas (a
true neurosurgical emergency), subdural hematomas (both
acute and chronic), intraparenchymal hemorrhages, and intra-
ventricular hemorrhages. As with tumors, EEG findings
depend heavily on the location, extent, and chronicity of the
hemorrhage.
Acute Subdural Hematoma
Slowing (including slow posterior basic rhythm and paroxys-
mal slow activity) and epileptiform discharges have been
noted in patients with acute subdural hematomas (Fig. 17.5)
(72–75). Case reports have mentioned PLEDs as one EEG
finding in acute subdural hematoma (76), although they have
been seen after subdural hematomas have been evacuated, as
well (77). EEG abnormalities may not be related to the size of
the hematoma, among other factors (73). One study suggested
that acute and subacute hematomas are associated with
changes in EEG amplitude, while chronic hematomas were
associated with changes in frequency (i.e., increased slow
activity) (78).
Chronic Subdural Hematoma
In a study of patients with unilateral chronic subdural
hematomas and herniation who underwent xenon/CT (to
assess cerebral blood flow) and electroencephalograms,
increased flow was noted in areas where alpha- and beta-range
activity predominated, while blood flow was decreased in
areas where slower frequencies predominated (79).
Furthermore, electrical activity correlated with thalamic
(rather than cortical) blood flow, leading the authors to con-
clude that thalamus is the source of the abnormal EEG activ-
ity in this population. Patients with chronic subdural
hematomas also can have normal EEGs; this was the case for
eight of nine patients recorded 8 months to nearly 4 years
after the initial insult (80).
Figure 17.5 EEG from a 76-year-old male showing delta-range slowing
over the right hemisphere. A right frontotemporoparietal subdural
hematoma and left temporofrontal hygroma were evacuated at surgery.
 Chapter 17
EEG AND EVOKED POTENTIALS IN NORMAL
PRESSURE HYDROCEPHALUS
The EEG in patients with normal pressure hydrocephalus (NPH,
i.e., idiopathic) can be normal or can show unilateral or bilateral
intermittent slowing (81). In one series of 14 patients with NPH,
half had EEGs with bursts of bilateral monorhythmic slow activ-
ity. This activity became more noticeable in four of five patients
who had serial EEGs, while a similar number had improvements
in the EEG after shunting operations (82). This suggests that
EEG changes may be useful for tracking progression of illness in
some patients with NPH, but another paper suggested that NPH
patients do not have specific patterns on their EEG (83).
EEG CHANGES DUE TO ONCOLOGICAL
AGENTS
CNS toxicity has been associated with many chemotherapy
agents. Electroencephalographic changes include slowing,
epileptiform activity, and voltage suppression. One difficulty in
interpreting these findings is that patients often are receiving
multiple agents, plus radiotherapy. Also, more than one finding
can occur in a given patient. Finally, these are patients that often
are systemically ill both from their primary illness and from
treatments. Therefore, a second difficulty lies in determining
whether a given medication represents a primary or secondary
cause of the EEG changes.
Diffuse slowing has been reported in children with systemic
tumors (including acute lymphoblastic leukemia [ALL], Ewing
sarcoma, non-Hodgkin lymphoma, reticulum cell sarcoma,
malignant chest wall tumor, and ovarian stroma cell carcinoma)
treated with agents including methotrexate, vincristine, leucov-
orin, L-asparaginase, adriamycin, actinomycin, 5-fluorouracil,
cyclophosphamide, cytosine arabinoside, and VP-12-213 (84).
Diffuse or generalized slowing occurred in less than 20% of
children receiving combinations of intrathecal methotrexate
alone, cranial radiation, and triple intrathecal therapy (consist-
ing of methotrexate, cytarabine, and prednisone) for B-lineage
ALL without CNS leukemia (85). Another study concluded that
vincristine and L-asparaginase were most likely to be associated
with EEG background slowing in a group of 13 children receiv-
ing combination therapy for ALL (86). The same agents were
associated with transient EEG slowing in 79 children treated for
ALL (87). The significance of these EEG changes is unclear, as
EEG abnormalities in long-term survivors of the protocol in the
latter study did not correlate with the presence or degree of
abnormalities in their exams or neuroimaging (88).
Other studies report slowing (generalized in distribution
unless otherwise described in the following list) with:
methotrexate (89); intrathecal methotrexate and cytosine ara-
binoside with intravenous cyclophosphamide, vincristine,
methotrexate, ifosfamide, and prednisolone (90); procar-
bazine, CCNU, and vincristine for recurrent gliomas (91);
paclitaxel (92) although one series showed no adverse effects
on the EEG in 28 patients with ovarian carcinoma treated
with a paclitaxel–carboplatin regimen (93);
■ Brain Tumors and Other Space-Occupying Lesions 327
combination therapy using cisplatin, vinblastine, and
bleomycin in a patient with an ovarian germ cell tumor with
posterior leukoencephalopathy associated with slowing in an
area of involvement noted on MRI (94);
combination therapy for a gastrointestinal stromal tumor using
5-fluorouracil, adriamycin, and mitomycin-C (95);
high-dose 5-fluorouracil and leucovorin (associated with
hyperammonemia, and lactic acidosis (96);
busulfan, given with clonazpam for seizure prophylaxis in chil-
dren with various types of tumors, with the EEG slowing
focal slowing (97);
IL-2 administration with bilateral slowing in the frontal
regions (98);
interleukin-2 for various malignancies or HIV infection with
bursts of frontal intermittent rhythmic delta activity (99);
interferon-alpha in adults (100–102) and children (103);
ifosfamide and mesna (104,105).
Epileptiform activity or seizures have been reported with:
intravenous high-dose methotrexate infusion (106);
chlorambucil (diffuse spike-wave activity) in two of nine chil-
dren with nephrotic syndrome (107);
intravenous vincristine infusions (a patient with unilateral
seizures) (108);
infusion of paclitaxel (109);
busulfan and cyclophosphamide conditioning regimen prior to
bone marrow transplant in patients with leukemia
(110,111);
interleukin-2 for various malignancies or HIV infection (with
periodic epileptiform discharges) (99);
interferon-alpha (100,102) and children (103);
ifosfamide and mesna (104,105), including nonconvulsive sta-
tus epilepticus (112).
EEG attenuation has been reported with:
high-dose methotrexate, vincristine in combination with other
cytotoxic agents, and L-asparaginase (84);
inadvertent intrathecal vincristine in a patient with lym-
phoblastic leukemia; this patient had EEG slowing that
evolved into “alpha coma” and voltage attenuation after the
event (113);
interferon-alpha and -gamma (114).
REFERENCES
1. Gloor P, Ball G, Schaul N. Brain lesions that produce delta waves
in the EEG. Neurology. 1977;27:326–333.
2. Munari C, Musolino A, Daumas-Duport C, et al. Correlation
between stereo-EEG, CT-scan and stereotactic biopsy data in
epileptic patients with low-grade gliomas. Appl Neurophysiol.
1985;48:448–453.
3. Aronica E, Boer K, Becker A, et al. Gene expression profile analysis of
epilepsy-associated gangliogliomas. Neuroscience. 2008;151:272–292.
4. Klass DW, Daly DD. Electroencephalography in patients with
brain tumor. Med Clin North Am. 1960;44:1041–1051.
5. Fish DR. How do tumors cause epilepsy? In: Kotagal P, Lüders HO,
eds. The Epilepsies: Etiologies and Prevention. San Diego: Academic
Press; 1999:301–314.
328 Part III ■ Clinical EEG: General
6. Aronica E, Gorter JA, Jansen GH, et al. Expression of connexin 43
and connexin 32 gap-junction proteins in epilepsy-associated brain
tumors and in the perilesional epileptic cortex. Acta Neuropathol.
2001;101:449–459.
7. Alonso-Nanclares L, De FJ. Vesicular glutamate transporter 1
immunostaining in the normal and epileptic human cerebral cor-
tex. Neuroscience. 2005;134:59–68.
8. Aronica E, Yankaya B, Jansen GH, et al. Ionotropic and metabotropic
glutamate receptor protein expression in glioneuronal tumours from
patients with intractable epilepsy. Neuropathol Appl Neurobiol. 2001;
27:223–237.
9. Richardson MP, Hammers A, Brooks DJ, et al. Benzodiazepine-
GABA(A) receptor binding is very low in dysembryoplastic
neuroepithelial tumor: a PET study. Epilepsia. 2001;42:
1327–1334.
10. Morrell F, de Toldedo-Morrell L. Secondary epileptogenesis and
brain tumors. In: Kotagal P, Lüders HO, eds. The Epilepsies:
Etiologies and Prevention. San Diego: Academic Press, 1999:
357–363.
11. Laws ER Jr, Taylor WF, Clifton MB, et al. Neurosurgical manage-
ment of low-grade astrocytoma of the cerebral hemispheres.
J Neurosurg. 1984;61:665–673.
12. Lynam LM, Lyons MK, Drazkowski JF, et al. Frequency of seizures
in patients with newly diagnosed brain tumors: a retrospective
review. Clin Neurol Neurosurg. 2007;109:634–638.
13. Hirsch JF, Sainte RC, Pierre-Kahn A, et al. Benign astrocytic and
oligodendrocytic tumors of the cerebral hemispheres in children.
J Neurosurg. 1989;70:568–572.
14. Low NL, Correll JW, Hammill JF. Tumors of the cerebral hemi-
spheres in children. Arch Neurol. 1965;13:547–554.
15. Backus RE, Millichap JG. The seizure as a manifestation of
intracranial tumor in childhood. Pediatrics. 1962;29:978–984.
16. Forsgren L, Bucht G, Eriksson S, et al. Incidence and clinical char-
acterization of unprovoked seizures in adults: a prospective popu-
lation-based study. Epilepsia. 1996;37:224–229.
17. Khan RB, Hunt DL, Boop FA, et al. Seizures in children with pri-
mary brain tumors: incidence and long-term outcome. Epilepsy
Res. 2005;64:85–91.
18. Kershman J, Conde A, Gibson WC. Electroencephalography in dif-
ferential diagnosis of supratentorial tumors. Arch Neurol
Psychiatry. 1949;62:255–268.
19. Martinius J, Matthes A, Lombroso CT. Electroencephalographic
features in posterior fossa tumors in children. Electroencephalogr
Clin Neurophysiol. 1968;25:128–139.
20. Bickford RG. Electroencephalographic diagnosis of brain tumors.
Am J Surg. 1957;93:946–951.
21. Kirstein L. The occurrence of sharp waves, spikes and fast activity
in supratentorial tumours. Electroencephalogr Clin Neurophysiol.
1953;5:33–40.
22. Newmark ME, Theodore WH, Sato S, et al. EEG, transmission
computed tomography, and positron emission tomography with
fluorodeoxyglucose 18F. Their use in adults with gliomas. Arch
Neurol. 1983;40:607–610.
23. Daly D, Whelan JL, Bickford RG, et al. The electroencephalogram
in cases of tumors of the posterior fossa and third ventricle.
Electroencephalogr Clin Neurophysiol. 1953;5:203–216.
24. Morris HH III, Luders H, Hahn JF, et al. Neurophysiological tech-
niques as an aid to surgical treatment of primary brain tumors.
Ann Neurol. 1986;19:559–567.
25. Gastaut JL, Michel B. The impact of cranial computerized tomog-
raphy on electroencephalography. Electroencephalogr Clin
Neurophysiol Suppl. 1978;34:123–132.
26. Reyes P, Jameson HD. Clinical correlation of periodic lateralized
epileptiform discharges in a general hospital. Electroencephalogr
Clin Neurophysiol. 1979;46:3P.
27. Fischer-Williams M, Dike GL. Brain tumors and other space-
occupying lesions. In: Niedermeyer E, Lopes da Silva F, eds.
Electroencephalography: Basic Principles, Clinical Applications, and
Related Fields. Philadelphia: Lippincott Williams & Wilkins;
2005:305–321.
28. Zaatreh MM, Firlik KS, Spencer DD, et al. Temporal lobe tumoral
epilepsy: characteristics and predictors of surgical outcome.
Neurology. 2003;61:636–641.
29. Nau HE, Bock WJ, Clar HE. Electroencephalographic investiga-
tions in sellar tumours, with special regard to different methods of
operative treatment. Acta Neurochir (Wien). 1978;44:207–214.
30. Harvey AS, Freeman JL. Epilepsy in hypothalamic hamartoma:
clinical and EEG features. Semin Pediatr Neurol. 2007;14:60–64.
31. Mullatti N, Selway R, Nashef L, et al. The clinical spectrum of
epilepsy in children and adults with hypothalamic hamartoma.
Epilepsia. 2003;44:1310–1319.
32. Arroyo S, Lesser RP, Gordon B, et al. Mirth, laughter and gelastic
seizures. Brain. 1993;116(Pt 4):757–780.
33. Cerullo A, Tinuper P, Provini F, et al. Autonomic and hormonal
ictal changes in gelastic seizures from hypothalamic hamartomas.
Electroencephalogr Clin Neurophysiol. 1998;107:317–322.
34. Berkovic SF, Andermann F, Melanson D, et al. Hypothalamic
hamartomas and ictal laughter: evolution of a characteristic
epileptic syndrome and diagnostic value of magnetic resonance
imaging. Ann Neurol. 1988;23:429–439.
35. Kuzniecky R, Guthrie B, Mountz J, et al. Intrinsic epileptogenesis
of hypothalamic hamartomas in gelastic epilepsy. Ann Neurol.
1997;42:60–67.
36. Leal AJ, Passao V, Calado E, et al. Interictal spike EEG source analy-
sis in hypothalamic hamartoma epilepsy. Clin Neurophysiol.
2002;113:1961–1969.
37. Leal AJ, Monteiro JP, Secca MF, et al. Functional brain mapping of
ictal activity in gelastic epilepsy associated with hypothalamic
hamartoma: a case report. Epilepsia. 2009;50:1624–1631.
38. Cascino GD, Kelly PJ, Sharbrough FW, et al. Long-term follow-up
of stereotactic lesionectomy in partial epilepsy: predictive factors
and electroencephalographic results. Epilepsia. 1992;33:639–644.
39. Packer RJ, Sutton LN, Patel KM, et al. Seizure control following
tumor surgery for childhood cortical low-grade gliomas. J
Neurosurg. 1994;80:998–1003.
40. Van Ness PC. Pros and cons of lesionectomy as treatment for par-
tial epilepsy. In: Kotagal P, Lüders HO, eds. The Epilepsies: Etiologies
and Prevention. San Diego: Academic Press, 1999:391–397.
41. Berger MS, Ghatan S, Geyer JR, et al. Seizure outcome in children
with hemispheric tumors and associated intractable epilepsy: the
role of tumor removal combined with seizure foci resection.
Pediatr Neurosurg. 1991;17:185–191.
42. Ebeling U, Fischer M, Kothbauer K. Surgery of astrocytomas in the
motor and premotor cortex under local anesthesia: report of 11
cases. Minim Invasive Neurosurg. 1995;38:51–59.
43. Berger MS, Ojemann GA. Intraoperative brain mapping tech-
niques in neuro-oncology. Stereotact Funct Neurosurg. 1992;58(1–4):
153–161.
44. Goldring S. A method for surgical management of focal epilepsy,
especially as it relates to children. J Neurosurg. 1978;49:344–356.
45. Duffau H, Peggy Gatignol ST, Mandonnet E, et al. Intraoperative
subcortical stimulation mapping of language pathways in a con-
secutive series of 115 patients with grade II glioma in the left dom-
inant hemisphere. J Neurosurg. 2008;109:461–471.
 Chapter 17
46. Duffau H. Acute functional reorganisation of the human motor
cortex during resection of central lesions: a study using intraoper-
ative brain mapping. J Neurol Neurosurg Psychiatry. 2001;70:
506–513.
47. Pilcher WH, Silbergeld DL, Berger MS, et al. Intraoperative
electrocorticography during tumor resection: impact on seizure
outcome in patients with gangliogliomas. J Neurosurg. 1993;78:
891–902.
48. Lesser RP, Raudzens P, Luders H, et al. Postoperative neurological
deficits may occur despite unchanged intraoperative somatosen-
sory evoked potentials. Ann Neurol. 1986;19:22–25.
49. Lueders H, Lesser RP, Hahn J, et al. Cortical somatosensory evoked
potentials in response to hand stimulation. J Neurosurg. 1983;58:
885–894.
50. Cedzich C, Taniguchi M, Schafer S, et al. Somatosensory evoked
potential phase reversal and direct motor cortex stimulation dur-
ing surgery in and around the central region. Neurosurgery. 1996;
38:962–970.
51. Cheek JC. Posterior fossa intraoperative monitoring. J Clin
Neurophysiol. 1993;10:412–424.
52. Nurlu G, Bavbek M, Colak A, et al. Brainstem auditory evoked
potentials and blink reflexes in patients with pontocerebellar angle
tumors. Neurosurg Rev. 1994;17:253–260.
53. Csecsei GI, Klug N, Szekely G Jr, et al. Multimodality electroneu-
rophysiological findings in intra-axial and extra-axial lesions of
the brain stem. Acta Neurochir (Wien). 1995;137:48–53.
54. Stanton SG, Cashman MZ. Auditory brainstem response. A com-
parison of different interpretation strategies for detection of cere-
bellopontine angle tumors. Scand Audiol. 1996;25:109–120.
55. Naessens B, Gordts F, Clement PA, et al. Re-evaluation of the ABR in
the diagnosis of CPA tumors in the MRI-era. Acta Otorhinolaryngol
Belg. 1996;50:99–102.
56. Khatib ZA, Heideman RL, Kovnar EH, et al. Predominance of pilo-
cytic histology in dorsally exophytic brain stem tumors. Pediatr
Neurosurg. 1994;20:2–10.
57. Baran JA, Catherwood KP, Musiek FE. “Negative” ABR findings in
an individual with a large brainstem tumor: hit or miss? J Am Acad
Audiol. 1995;6:211–216.
58. Tanaka H, Komatsuzaki A, Hentona H. Usefulness of auditory
brainstem responses at high stimulus rates in the diagnosis of
acoustic neuroma. ORL J Otorhinolaryngol Relat Spec. 1996;58:
224–228.
59. Broggi G, Scaioli V, Brock S, et al. Neurophysiological monitoring
of cranial nerves during posterior fossa surgery. Acta Neurochir
Suppl. 1995;64:35–39.
60. Eisner W, Schmid UD, Reulen HJ, et al. The mapping and contin-
uous monitoring of the intrinsic motor nuclei during brain stem
surgery. Neurosurgery. 1995;37:255–265.
61. Lesser RP, Lueders H, Hahn J, et al. Early somatosensory potentials
evoked by median nerve stimulation: intraoperative monitoring.
Neurology. 1981;31:1519–1523.
62. Nakasato N, Kumabe T, Kanno A, et al. Neuromagnetic evaluation
of cortical auditory function in patients with temporal lobe tumors.
J Neurosurg. 1997;86:610–618.
63. Moore BD III, Copeland DR, Ried H, et al. Neurophysiological
basis of cognitive deficits in long-term survivors of childhood can-
cer. Arch Neurol. 1992;49:809–817.
64. Harmony T, Marosi E, Fernandez T, et al. EEG coherences in
patients with brain lesions. Int J Neurosci. 1994;74:203–226.
65. Towle VL, Syed I, Berger C, et al. Identification of the sensory/
motor area and pathologic regions using ECoG coherence.
Electroencephalogr Clin Neurophysiol. 1998;106:30–39.
■ Brain Tumors and Other Space-Occupying Lesions 329
66. Aebi C, Kaufmann F, Schaad UB. Brain abscess in childhood—
long-term experiences. Eur J Pediatr. 1991;150:282–286.
67. Ma JS, Chen PY, Lau YJ, et al. Brain abscess caused by Salmonella
enterica subspecies houtenae in a patient with chronic granuloma-
tous disease. J Microbiol Immunol Infect. 2003;36:282–284.
68. Gurer G, Yemisci M, Saygi S, et al. Structural lesions in periodic
lateralized epileptiform discharges (PLEDs). Clin EEG Neurosci.
2004;35:88–93.
69. Beller AJ, Sahar A, Praiss I. Brain abscess. Review of 89 cases
over a period of 30 years. J Neurol Neurosurg Psychiatry. 1973;36:
757–768.
70. Michel B, Gastaut JL, Bianchi L. Electroencephalographic cranial
computerized tomographic correlations in brain abscess.
Electroencephalogr Clin Neurophysiol. 1979;46:256–273.
71. Vignadndra V, Ghee LT, Chawla J. EEG in brain abscess: its value in
localization compared to other diagnostic tests. Electroencephalogr
Clin Neurophysiol. 1975;38:611–622.
72. Gutierrez-Luque AG, Maccarty CS, Klass DW. Head injury with
suspected subdural hematoma. Effect on EEG. Arch Neurol. 1966;
15:437–443.
73. Jaffe R, Librot IE, Bender MB. Serial EEG studies in unoperated
subdural hematoma. Arch Neurol. 1968;19:325–330.
74. Poole EW. Some aspects of electroencephalographic disturbances
following head injury. J Clin Pathol Suppl (R Coll Pathol)
1970;4:187–201.
75. Jones SC, Bamford JM, Heath J, et al. Multiple forms of epileptic
attack secondary to a small chronic subdural haematoma. BMJ.
1989;299:439–441.
76. Chu NS. Acute subdural hematoma and the periodic lateralized
epileptiform discharges. Clin Electroencephalogr. 1979;10:145–150.
77. Westmoreland BF. Periodic lateralized epileptiform discharges
after evacuation of subdural hematomas. J Clin Neurophysiol.
2001;18:20–24.
78. Ibrahim MM, Elian M. A reappraisal of the value of the EEG in
subdural haematoma. A report on 86 patients, with a follow-up
study. J Neurol. 1974;207:117–128.
79. Tanaka A, Kimura M, Yoshinaga S, et al. Quantitative electroen-
cephalographic correlates of cerebral blood flow in patients with
chronic subdural hematomas. Surg Neurol. 1998;50:235–240.
80. Lusins J, Jaffe R, Bender MB. Unoperated subdural hematomas.
Long-term follow-up study by brain scan and electroencephalog-
raphy. J Neurosurg. 1976;44:601–607.
81. Brown DG, Goldensohn ES. The electroencephalogram in normal
pressure hydrocephalus. Arch Neurol. 1973;29:70–71.
82. Hashi K, Nishimura S, Kondo A, et al. The EEG in normal pressure
hydrocephalus. Acta Neurochir (Wien). 1976;33:23–35.
83. Petersen RC, Mokri B, Laws ER Jr. Surgical treatment of idiopathic
hydrocephalus in elderly patients. Neurology. 1985;35:307–311.
84. Schaffler L, Imbach P, Rudeberg A, et al. Conventional and spectral
EEG analysis in children treated with cytotoxic agents. Eur J
Cancer Clin Oncol. 1982;18:827–832.
85. Lo Nigro L, Di CA, Schiliro G. Acute neurotoxicity in children with
B-lineage acute lymphoblastic leukemia (B-ALL) treated with
intermediate risk protocols. Med Pediatr Oncol. 2000;35:449–455.
86. Lichter-Konecki U, Benninger C, Brandeis WE, et al. Changes in
the EEG background activity of children with acute lymphoblastic
leukemia during cytotoxic therapy. Pediatr Hematol Oncol.
1987;4:77–85.
87. Korinthenberg R, Igel B. Prospective neurophysiological study in
children treated for acute lymphoblastic leukaemia: serial EEG
during treatment and long-term follow up with evoked potentials.
Eur J Pediatr. 1990;150:127–131.
330 Part III ■ Clinical EEG: General
88. Ueberall MA, Skirl G, Strassburg HM, et al. Neurophysiological
findings in long-term survivors of acute lymphoblastic leukaemia
in childhood treated with the BFM protocol 81 SR-A/B. Eur J
Pediatr. 1997;156:727–733.
89. Ochs J, Parvey LS, Mulhern R. Prospective study of central nerv-
ous system changes in children with acute lymphoblastic
leukemia receiving two different methods of central nervous sys-
tem prophylaxis. Neurotoxicology. 1986;7:217–226.
90. Geissler RG, Bergmann L, Hacker H, et al. Severe central nervous
adverse effect of intrathecal chemotherapy in a 16-yr-old patient
with Burkitt’s type lymphoma. Eur J Haematol. 1997;58:286–288.
91. Postma TJ, van Groeningen CJ, Witjes RJ, et al. Neurotoxicity of
combination chemotherapy with procarbazine, CCNU and vin-
cristine (PCV) for recurrent glioma. J Neurooncol. 1998;38:69–75.
92. Perry JR, Warner E. Transient encephalopathy after paclitaxel
(taxol) infusion. Neurology. 1996;46:1596–1599.
93. Mayerhofer K, Bodner K, Saletu B, et al. Paclitaxel–carboplatin
chemotherapy does not cause encephalopathy in patients with
ovarian cancer: a prospective EEG mapping study in 28 patients.
Anticancer Res. 2001;21:803–808.
94. Sueblinvong T, Noophun P, Pataradool K, et al. Posterior
leukoencephalopathy following cisplatin, bleomycin and vinblas-
tine therapy for germ cell tumor of the ovary. J Obstet Gynaecol
Res. 2002;28:99–103.
95. Ki SS, Jeong JM, Kim SH, et al. A case of neurotoxicity following
5-fluorouracil-based chemotherapy. Korean J Intern Med. 2002;
17:73–77.
96. Yeh KH, Cheng AL. High-dose 5-fluorouracil infusional therapy
is associated with hyperammonaemia, lactic acidosis and
encephalopathy. Br J Cancer. 1997;75:464–465.
97. Vassal G, Deroussent A, Hartmann O, et al. Dose-dependent neu-
rotoxicity of high-dose busulfan in children: a clinical and phar-
macological study. Cancer Res. 1990;50:6203–6207.
98. Pace A, Pietrangeli A, Bove L, et al. Neurotoxicity of antitumoral
IL-2 therapy: evoked cognitive potentials and brain mapping. Ital
J Neurol Sci. 1994;15:341–346.
99. Karp BI, Yang JC, Khorsand M, et al. Multiple cerebral lesions
complicating therapy with interleukin-2. Neurology. 1996;47:
417–424.
100. Smedley H, Katrak M, Sikora K, et al. Neurological effects of
recombinant human interferon. Br Med J (Clin Res Ed). 1983;
286:262–264.
101. Rohatiner AZ, Prior PF, Burton AC, et al. Central nervous system
toxicity of interferon. Br J Cancer. 1983;47:419–422.
102. Suter CC, Westmoreland BF, Sharbrough FW, et al. Electro-
encephalographic abnormalities in interferon encephalopathy: a
preliminary report. Mayo Clin Proc. 1984;59:847–850.
103. Miller VS, Zwiener RJ, Fielman BA. Interferon-associated refrac-
tory status epilepticus. Pediatrics. 1994;93:511–512.
104. Meanwell CA, Blake AE, Kelly KA, et al. Prediction of ifos-
famide/mesna associated encephalopathy. Eur J Cancer Clin
Oncol. 1986;22:815–819.
105. Pratt CB, Green AA, Horowitz ME, et al. Central nervous system
toxicity following the treatment of pediatric patients with ifos-
famide/mesna. J Clin Oncol. 1986;4:1253–1261.
106. Kubo M, Azuma E, Arai S, et al. Transient encephalopathy
following a single exposure of high-dose methotrexate in a child
with acute lymphoblastic leukemia. Pediatr Hematol Oncol.
1992;9:157–165.
107. Ichida F, Konishi T, Asada R, et al. Chlorambucil central nerv-
ous toxicity: a significant side effect of chlorambucil therapy in
childhood nephrotic syndrome. Eur J Pediatr. 1985;144:
283–286.
108. Dallera F, Gamoletti R, Costa P. Unilateral seizures following vin-
cristine intravenous injection. Tumori. 1984;70:243–244.
109. McGuire WP, Rowinsky EK, Rosenshein NB, et al. Taxol: a
unique antineoplastic agent with significant activity in advanced
ovarian epithelial neoplasms. Ann Intern Med. 1989;111:
273–279.
110. Meloni G, Raucci U, Pinto RM, et al. Pretransplant conditioning
with busulfan and cyclophosphamide in acute leukemia patients:
neurological and electroencephalographic prospective study. Ann
Oncol. 1992;3:145–148.
111. La MC, Mondini S, Guarino M, et al. Busulfan neurotoxicity and
EEG abnormalities: a case report. Neurol Sci. 2004;25:95–97.
112. Wengs WJ, Talwar D, Bernard J. Ifosfamide-induced nonconvul-
sive status epilepticus. Arch Neurol. 1993;50:1104–1105.
113. Bleck TP, Jacobsen J. Prolonged survival following the inadver-
tent intrathecal administration of vincristine: clinical and
electrophysiologic analyses. Clin Neuropharmacol. 1991;
14:457–462.
114. Born J, Spath-Schwalbe E, Pietrowsky R, et al. Neurophysiological
effects of recombinant interferon-gamma and -alpha in man.
Clin Physiol Biochem. 1989;7:119–127.

The EEG in Inflammatory
CNS Conditions
ROBERT L. BEACH, HELEN BARKAN, AND EDGAR DEPERALTA
ACKNOWLEDGMENTS
This chapter addresses the electroencephalogram in infection-
related and immune-mediated CNS inflammatory conditions.
This distinction is somewhat artificial in itself, since most CNS
inflammation is immune-mediated, either primarily or second-
arily. Previous versions of this chapter were written by the sec-
ond author’s most esteemed mentor, Dr. Barbara Westmoreland.
The current version is an update, which reflects significantly her
work and wisdom. Since the last edition, the information on the
etiology of several immune-mediated CNS inflammatory
processes has progressed significantly, and their associated elec-
troencephalographic findings are reviewed here. Some previous
topics are covered in other chapters of this volume.
INTRODUCTION
The range of EEG abnormalities in inflammatory and infectious
disorders was the subject of great interest in the 1950s and
1960s, but has not received as much attention since then. The
main reason for this lack of engagement is the improvement in
diagnostic imaging modalities, and the resultant recognition of
the limitations of the use of EEG in the diagnosis, treatment, and
follow-up of most of those conditions, with a few notable excep-
tions. The EEG of an inflamed brain can be completely normal,
can reveal minor generalized or lateralized abnormalities, or
may show severe focal or global slowing, and may also contain a
variety of potentially epileptogenic abnormalities or frank elec-
trographic seizures of focal, lateralized, or generalized onset.
Hence, many believe that most of those EEG abnormalities are
nonspecific, and of low to moderate diagnostic and prognostic
value, except for a few “signature” EEG patterns corresponding
to specific conditions and even to specific pathogens. More
EEG-based research directed specifically at validating the prog-
nostic and diagnostic value of the EEG in infectious and inflam-
matory CNS conditions is clearly needed. In the digital EEG age,
and with sophisticated analysis software, sensitive feature recog-
nition should allow the revitalization of this pursuit.
The majority of EEG changes in inflammatory and infec-
tious CNS disease processes represent the “final common path-
way” of the electrographic activity of the inflamed cortex in
either primary or secondary inflammation. Nevertheless, it is
important to recognize both the specific and the nonspecific
patterns, and to use them as aid to diagnosis, to gauge the
response to treatment, and for prognostic purposes as well.
For consistency, all EEG tracings in the figures illustrating
the text below were captured at the amplitude of 5 V/mm,
CHAPTER
18
with 1- to 70-Hz bandpass filter, and with the notch filter on,
with about 13 seconds per page.
PART 1: INFECTIOUS CONDITIONS
Meningitis
Infection and inflammation of the meninges can have bacterial,
fungal, parasitic, autoimmune, and malignancy-related patho-
genesis, or it can be a secondary to a brain abscess or empyema
with contiguous spread of infection. In the latter case, lateraliz-
ing EEG findings can be anticipated. In the rest of the condi-
tions, the EEG may range from completely normal to severely
abnormal including focal- or generalized-onset seizures and
potentially epileptogenic discharges. However, in the majority
of these conditions, only diffuse slow-wave abnormalities of
variable severity will be seen. There is objective evidence (1), as
cited in the previous version of this chapter (2), that:
Although the electroencephalographic findings are not essential
for making the specific diagnosis of meningitis, the EEG and
particularly serial recordings are helpful in following the course
of the disease, detecting the development of complications or
relapse, and indicating the presence of sequelae of residual brain
damage.
Purulent Meningitis
Meningococcal, streptococcal, and staphylococcal meningitis
are usually community-acquired, with pneumococcal meningi-
tis being an infrequent complication of pneumonia, otitis
media, mastoiditis, severe alcoholism, and CSF leakage after
ENT procedures. Staphylococcal meningitis may complicate
postoperative course of cranial surgery.
Diffuse moderate to very severe and nonspecific slow-wave
abnormalities are typical for the EEG in those conditions (Fig.
18.1) (3). The slowing is frequently prominent bifrontally, resem-
bling FIRDA (frontal intermittent rhythmic delta activity) possi-
bly associated with increased intracranial pressure (Fig. 18.2).
Studies of initial EEG evaluation of patients with purulent
bacterial meningitis suggest that about one half of the patients
have slow-wave abnormalities or electrographic seizures (3).
If only community-acquired disease is considered, about
20% of the patients will have clinical seizures, and a small num-
ber of those have clinical and electrographic status epilepticus
(4). The patients with seizures have mortality of almost three
times higher than those without the seizures. Focal radi-
ographic lesions are found in one third of patients with menin-
gitis and in-hospital seizures.
331
332 Part III ■ Clinical EEG: General

Figure 18.1 Awake EEG of a 70-year-old man with community-acquired staphylococcal meningitis showing diffuse mode -
rately severe delta slowing.

Figure 18.2 EEG in an obtunded 63-year-old man with cervical MRSA abscess and secondary meningitis, showing rhyth-
mic bifrontal slowing consistent with FIRDA.
 Chapter 18
Other EEG abnormalities in bacterial meningitis also have
been described, such as triphasic waves and FIRDA, although
both are more common for meningitis of other etiologies.
An attempt to predict the bacterial or viral nature of the
pathogen from the EEG abnormalities suggests that moderately
severe slowing in the delta range is much more likely in bacter-
ial meningitis (5).
In neonates with meningitis, with group B streptococcus and
E. coli being the most common pathogens, the presence of EEG
abnormalities and/or electrographic and clinical seizures was
directly correlated with prognosis, and associated with poor
outcome (Fig. 18.3). Neonatal convulsive disorders of multiple
etiologies are covered in a separate chapter of this volume.
Brain abscess and subdural empyema can be either cause or
sequelae of bacterial meningitis. The pathogens are often mixed
mouth flora, including anaerobes, staphylococci, and gram-
negative commensal organisms. The EEG findings in such cases
often reflect a focal lesion—with moderately severe slowing,
potentially epileptogenic abnormalities, and clinical and/or
electrographic seizures. The focal findings may manifest as
polymorphic delta slowing, suppression of activity over the area
of suppuration, or, infrequently, lateralized polyphasic poten-
tially epileptogenic abnormalities such as sharp waves or peri-
odic lateralized epileptiform discharges (PLEDs).
Basal Meningitis
Basal meningitis due to mycobacteria, Listeria, or spirochetes
(syphilis and borreliosis) overlaps with the entity called “granu-
lomatous meningitis,” or “chronic meningitis,” and also with
meningoencephalidites described for the same pathogens. The
separate use of those terms is imprecise, since biopsy is obtained
only in rare cases, and the degree of chronicity varies widely.
EEG in basal meningitis ranges from completely normal to
severely abnormal, and in addition to varied degrees of general-
ized slowing, hemispheric asymmetries, and sharp paroxysmal
activity that is either focal or generalized, triphasic waves are not
uncommon (6), as well as FIRDA similar to that in Figure 18.2
(7), which may reflect the underlying increase in intracranial
pressure. Because of the association of those pathogens with
cerebral vessel vasculitis, it is rare to see normal EEG in those
patients. In one series of 19 patients, 11 had EEG abnormalities.
Tuberculous meningitis in another series was associated with
similar EEG abnormalities in 24 out of 32 patients.
Focal epileptic discharges and electrographic seizures are
also common in tuberculous meningitis (8), particularly in
children, and reflect the patchy focal manner of meningeal
inflammation.
Fungal and Parasitic Meningitis
These pathogens are relatively uncommon in a healthy host.
Immunosuppressed patients with cryptococcal, cystericercal,
candidial, or other meningitis—with the inflammation that can
be isolated to the meninges, but usually also involving the brain
parenchyma—are typically very ill, and their EEG pattern is
that of a severe encephalopathy, with polymorphic slowing.
Focal seizures and status epilepticus are common in those
patients as well as in those with basal meningitis.
■ The EEG in Inflammatory CNS Conditions 333
Meningitis due to cystocercosis, aspergillosis, nocardiosis,
echinococcosis, mycormycosis, toxoplasmosis, and candidiasis
is associated with focal slow-wave abnormalities (2), again
reflecting the patchy nature of meningeal pathology due to
infection by those organisms. No organism-specific patterns are
evident, however.
Viral Meningitis
Viral, otherwise known as “aseptic,” meningitis may not result
in any EEG abnormalities, or may result in mild to moderate
generalized slowing. In one series of 62 patients, 52% had
abnormal EEGs with slow-wave abnormalities (9). Other inves-
tigators found that only 28% out of 82 patients (10), or in yet
other studies, one third of patients with CSF-confirmed aseptic
meningitis had EEG abnormalities. Thus, a normal EEG with
inflammatory CSF findings strongly suggests viral nature of the
pathogen—or early stage of CNS involvement.
Practically any infectious agent can cause mild to moderate
transient meningoencephalitis with variable EEG changes.
Parenchymal involvement is assumed, since biopsy is rarely
available, and thus will be discussed in the section “Encephalitis
with Nonspecific EEG Patterns.” Rarely a purely encephalitic
process, such as HSV encephalitis, can cause focal meningeal
irritation and a clinical and laboratory picture consistent with
aseptic or chemical meningitis.
Mollaret Meningitis
The agent of this recurrent, lymphocytic meningitis with rela-
tively benign CSF findings and clinical course was thought to
be herpes simplex type II virus, which has been recovered from
the CSF in some cases. However, some suggest the term should
be restricted to idiopathic cases, as several other infectious and
autoimmune etiologies as well as CSF leaks and shunts have
been implicated in cases of recurrent meningitis (11–14). EEG
in Mollaret meningitis is usually normal; however, diffuse
moderately severe slowing can be seen, correlated with the
clinical symptoms of headache and confusion. Seizures are
unusual.
Carcinomatous Meningitis
Involvement of the meninges in the neoplastic process is not an
uncommon terminal manifestation of systemic cancer. The
EEG in these patients is practically never normal, with moder-
ate or very severe slowing that correlates with the clinical sever-
ity of the encephalopathic state. Focal abnormalities such as
PLEDs, sharp waves, and spikes are also common. Generalized
triphasic periodic waves have also been described.
Meningoencephalitis
Meningoencephalitis is a general term, describing the inflam-
mation of both the meninges and the brain parenchyma due to
toxic, metabolic, infectious, or autoimmune etiologies, with
many of them overlapping on many occasions. Demyelinating
disease, connective tissue diseases, and malignancy-related
encephalitides are discussed later in this chapter. This section
will concentrate on infectious and autoimmune etiologies
exclusive of the above-mentioned.
334 Part III ■ Clinical EEG: General

Figure 18.3 EEG in a 7-week-old female born at 32 weeks, who has group B streptococcus meningitis, reveals interictal
diffuse suppression of activity with intermittent focal sharp waves (A), and independent left- and right-sided PLEDs

Similar to the EEG in meningitis, the encephalitic brain can
produce a normal or an almost normal study, or display pro-
found abnormalities. The same groups of pathogens that cause
meningitis, bacterial, viral, parasitic, and fungal, also com-
monly cause meningoencephalitis. Dissimilar to viral meningi-
tis, the patients with viral encephalitis of any etiology rarely
have a normal EEG. The EEG abnormalities range from diffuse
slowing to electrographic seizures, and are of the two
varieties—disease-specific and diagnostically vague. There are
three of the former variety and hundreds of the latter.
Nevertheless, the EEG is of value in both cases, since disease-
specific patterns represent common conditions, and the
 Chapter 18
Figure 18.3 (continued) (B and C) with no clinical accompaniment; electrographic seizures with the subtle correlate of
contralateral eye deviation were also seen.
absence of a specific pattern may in itself be diagnostic. The
evolution of encephalitic EEG from the prodromal to acute
phase of the disease and to recovery is similar to that in menin-
gitis, and may provide useful information for prognosis, com-
plications, and potential sequelae.
Encephalitis with Disease-Specific EEG Patterns
Only three neurologic infections have a corresponding pathog-
nomic EEG signature: herpes simplex encephalitis, most common
variants of prion disease, and subacute sclerosing panencephalitis
(SSPE), although the latter is a very rare diagnostic entity in the
Western world, where measles vaccination is a norm.
Herpes Simplex (HSV-I) Encephalitis This common cata-
strophic CNS illness strikes healthy people of all ages, but is
more common in patients over 50 years. HSV-1 has a predilec-
tion for the temporal lobes and for the frontotemporal areas, as
compared to its cousin HSV-2, the presumed causative agent of
Mollaret meningitis, which does not usually affect the
parenchyma. The classical presentation of intermittent fever,
confusion, and seizures with CSF lymphocytosis and positive
PCR for the virus is atypical in clinical practice: the diagnosis is
best made based on any combination of clinical, radiographic,
laboratory, and EEG findings compatible with this condition.
Since this is one of the few treatable encephalitides, initial over-
diagnosis is preferable in the case of HSV-1 encephalitis.
The initial EEG abnormalities (1,15) consist of polymorphic
delta slowing, often more prominent over the most involved
temporal lobe, with focal sharp complexes appearing and
■ The EEG in Inflammatory CNS Conditions 335
rapidly evolving into pseudoperiodic to periodic PLEDs, 0.3 to
1 Hz in frequency, and di- or triphasic broad or tight giant
sharp waves. These may be bilateral (biPEDs), bifrontal, bitem-
poral, or even generalized (GPEDs). On occasion, PLEDs are
focal with a small field in a specific location, such as the ante-
rior temporal region (Fig. 18.4).
Temporally, PLEDs and/or focal sharp complexes reveal
themselves within days to a week after the onset of symptoms,
but they may persist for several weeks into the illness, and may
be correlated with the response to treatment.
In addition, seizure discharges are common over the most
affected areas, consisting of rhythmic activity or polyspike
bursts. Clinical correlate may or may not occur, but confusion
and dysphasia are likely if the dominant hemisphere is affected.
Obtundation and coma are not uncommon, especially when
EEG abnormalities are severe and widespread. When clinical
motor seizures do occur, the frequency of clonic limb activity
may not match the frequency of PLEDs, which may be signifi-
cantly slower than the clonus.
In refractory or fatal cases, a burst suppression pattern may
be seen, with generalized EEG attenuation following a burst of
PLEDs or polyspikes, which indicates a grim prognosis. In non-
fatal HSV encephalitis, the strikingly abnormal EEG recovers
slowly to almost normal on serial examinations. The PLEDs
become infrequent, intermittent, and are replaced by lateralized
slow-wave abnormalities. Clinical improvement often precedes
electrographic normalization.
Possibly because of the associated severe encephalopathy,
other, seemingly atypical patterns are frequently seen in HSV
336 Part III ■ Clinical EEG: General

Figure 18.4 EEG from a 62-year-old man with CSF-positive HSV encephalitis and MRI abnormalities maximal in the right
frontotemporal head region reveals initially right temporal slowing and isolated sharp waves and spikes (A), and with symp-
tomatic progression of the disease, pseudoperiodic PLEDs followed by suppression over the entire right hemisphere (B).

encephalitis, including generalized triphasic waves or atypical
triphasics, with posterior-to-anterior phase shift, multifocal
sharp waves, and severe polymorphic delta slowing.
Because the most common reason for PLEDs is a vascular
insult, dual etiology needs to be considered as part of the differ-
ential diagnosis. The coexistence of HSV-1 encephalitis and
stroke or a hemorrhage is common, and radiologic correlation
is crucial.
Prion Diseases Creutzfeldt–Jakob disease (CJD), variant CJD or
“mad cow disease,” the exceedingly rare fatal familial insomnia
(a.k.a. Gerstmann–Straussler–Scheinker disease), and kuru are
 Chapter 18
the four well-described prion disorders, which can be familial,
sporadic, or infectious in etiology. The pathogen is an abnor-
mally folding isoform of a native CNS protein, known as prion
protein. The most common variant of classic CJD is sporadic
rather than genetic, although many inherited polymorphisms
predispose to the disease. Presenting with the clinical manifes-
tations of a rapidly progressive dementia in a person over 60
years, it is often misdiagnosed at early stages as depression, psy-
chosis, or another psychiatric illness. Spontaneous myoclonus
and stimulus-related myoclonus come in later stages of the dis-
ease, as well as severe generalized spasticity, rhizus sardonicus,
obtundation, and death. There are no known cases of recovery.
The initial EEG abnormalities are nonspecific and mild, usu-
ally consisting of generalized slowing and disorganization of
the background (16,17). However, subclinical, that is, noncon-
vulsive status epilepticus (NCSE) or epilepsia partialis continua
(EPC) have been repeatedly described as initial presenting find-
ings in CJD (18–23). Over weeks to months, focal di- or tripha-
sic sharp discharges appear, with an after-coming slow
component. These are often generalized or bifrontal, occasion-
ally more prominent over one region or hemisphere, without a
clinical correlate. As the disease progresses, these become more
persistent, severe, and eventually relentless (17).
The pathognomic pattern of CJD in the EEG consists of con-
tinuous generalized, bisynchronous periodic stereotypic sharp
waves, occurring in intervals of 0.5 to 1 second with duration of
200 to 400 msec. Initially those abnormalities are lateralized;
however, they become generalized as disease progresses.
Atypical patterns may be seen, with bifrontal sharp waves that
are pseudoperiodic, or even with multifocal spike–wave dis-
charges. CSF 14-3-3 protein and MRI abnormalities on DWI
and later T2 images in cortex and basal ganglia are usually pos-
itive in both typical and atypical cases (Fig. 18.5) (24).
Myoclonic jerks maximal in the upper body may occur in
close temporal association with generalized sharp wave com-
plexes, or they may have no EEG correlate. Startle myoclonus or
provoked myoclonus is seen on occasion. Occasionally, external
stimuli can drive the rate of both the myoclonus and the corre-
sponding sharp wave discharges.
In the advanced stages of the disease, a burst suppression
and then a diffuse suppression pattern is evident, with eventual
isoelectric EEG heralding death.
Atypical prion disease variants in younger populations or
related to animal consumption have been described, where the
pathognomic CJD pattern on the EEG is absent. Insufficient
information exists to comment on the EEG abnormalities in
kuru, “mad cow,” and a rare FFI variant of prion disease.
Heidenhain variant of CJD has been described with the pre-
dominant occipital abnormalities of the EEG (25).
Generalized sharp wave discharges and multifocal or
bifrontal pseudoperiodic or periodic polyphasic discharges
with a clinical correlate of a rapidly progressive dementia
should alert the practitioner to the possibility of a prion disease;
however, other infectious or vascular or toxic–metabolic etiolo-
gies should always be considered. Typical diffusion-weighted
MRI changes may be of help, but the diagnosis is made on a
combination of longitudinal EEG and clinical findings.
■ The EEG in Inflammatory CNS Conditions 337
Subacute Sclerosing Panencephalitis SSPE is almost an irrele-
vant condition in the Western world, where measles vaccina-
tions are routine. However, given an active antivaccination
minority movement in the developed countries, and the fact
that measles is endemic in some underdeveloped countries, it is
important to include SSPE in this discussion.
SSPE occurs as a late complication in children who survived
a measles infection, and clinically resembles CJD, with rapid
intellectual decline, anorexia, spasticity, immobility, myoclonus,
choreiform movements, and seizures.
Typical periodic giant complexes are present in the EEG,
with the frequent clinical accompaniment of jerks or whole-
body spasms.
The pattern, initially described by Radermecker and Poser
(26), consists of high-voltage repetitive polyphasic and sharp-
and slow-wave complexes 0.5 to 2 seconds in duration, occur-
ring multiple times in 1 minute.
Multiple typical and atypical electrographic presentations of
SSPE have been described (27–31). The typical pattern described
by Radermecker, correlated with historical and clinical findings,
is definitive for the diagnosis, while atypical EEG patterns of sta-
tus epilepticus, bifrontal spikes, or other generalized abnormali-
ties as described above may also be useful to the diagnosis.
Encephalitis with Nonspecific EEG Patterns
A multitude of infectious and autoimmune inflammatory
processes, with the two coexisting, affect the brain parenchyma,
causing disease clinically ranging from mild to catastrophic.
Common infections in pediatric and adult world, if investigated
with a CSF and EEG, reveal abnormalities in both, with diffuse
slowing in the EEG roughly correlated with the onset and the
resolution of symptoms. There are hundreds of reports of mild-
to-moderate-to-fatal encephalitis associated with common
pathogens such as coxsackie virus, adenovirus, mycoplasma,
rickettsia, mumps, rubella, measles, influenza A and B, and
many others. None of those have an EEG pattern of significant
diagnostic value. Almost all of these pathogens described as
capable of causing meningitis actually cause meningoen-
cephalitis. Bacterial, viral, tuberculous, spirochetal, parasitic, or
fungal encephalitis typically has generalized severe polymor-
phic slowing in the delta range on the EEG, but focal slowing is
also possible in granulomatous or parasitic disease.
Viral encephalitis caused by enteroviridae, arboviridae, and
echoviridae, with vectors ranging from ticks to mosquitoes,
with frequently undetermined etiology, results in EEG abnor-
malities ranging from mild to severe slowing to electrocerebral
silence, and from lateralized potentially epileptogenic dis-
charges to seizures and status epilepticus with burst suppres-
sion pattern or with constant seizures (Fig. 18.6). The acute
EEG is rarely highly prognostic unless it contains a pattern of
burst suppression, which is always ominous, and the long-term
prognosis, even with fair for survival, puts the patient at lifelong
risk of intractable multifocal seizures. Bilateral mesial temporal
sclerosis is not an uncommon complication of viral encephali-
tis with or without an identified pathogen (Fig. 18.7).
Borrelia burgdorferi-related encephalitis has gotten some
recent attention (32,33), although it is not different than other
338 Part III ■ Clinical EEG: General

Figure 18.5 EEG of a 61-year-old woman with tremor and confusion, with MRI findings suggesting CJD (abnormal T2
signal in the caudate and lentiform nuclei), reveals a progression over 2 weeks from disorganization and slowing with occa-
sional pseudoperiodic sharp waves (A) to continuous periodic sharp wave discharges that are lateralized with a wide field
(B), without a clinical correlate.

secondary autoimmune CNS inflammatory processes, and is
usually mild and self-limited clinically, with nonspecific
encephalopathic changes in the EEG infrequently noted during
the course of the disease.
Hepatitis, particularly hepatitis A, may cause CNS inflamma-
tion and an encephalitic presentation with EEG features of
diffuse or focal slowing. West Nile encephalitis generally presents
with widespread electrographic abnormalities similar to other
viral meningoencephalitis (34), though anterior or temporal pre-
dominant slowing has been reported (35). The range of EEG
abnormalities is wide, from none to status epilepticus and/or
burst suppression pattern that is always ominous (36–38).
 Chapter 18 ■ The EEG in Inflammatory CNS Conditions 339

Figure 18.6 Typical nonspecific EEG pattern shown in A, in a 25-year-old woman with encephalitis NOS without a known
pathogen, contracted during a summer vacation. It shows generalized severe slowing and multifocal spikes, some with a
narrow field, and some nearly generalized. Another typical nonspecific EEG pattern, B, is from a 17-year-old male with
CSF-negative, presumably tick-borne encephalitis with a devastating initial course of coma, choreiform movements, status
myoclonus, and wasting. The EEG during clinically comatose state demonstrates generalized moderately severe delta slow-
ing and suppression.
340 Part III ■ Clinical EEG: General
Figure 18.7 This awake-state EEG is from a 60-year-old man with neuroborreliosis, and demonstrates mostly generalized
suppression.
Other encephalitides of viral or presumed viral etiologies
also offer no specific EEG patterns of value for diagnosis
(39–41).
To summarize, specific electroencephalographic diagnosis
for most viral or bacterial or fungal or parasitic etiologies of
encephalitis does not exist; however, serial EEG recordings pro-
vide a good basis for prognostication (42). The amazing thing
about viral encephalitis is that a severely abnormal EEG may
recover in days or weeks, and the EEG recovery is often corre-
lated with the clinical outcome.
HIV-Related Encephalitis
Clinically or electrographically, AIDS-related encephalitis is not
different from other viral encephalitides; however, there is evi-
dence of early CNS involvement and poor prognosis in some
studies (43,44). HIV, “the great imitator,” may present with
meningoencephalitis at early, middle, or late stages of the dis-
ease, without the evidence for secondary opportunistic CNS
infection (45). No disease-related or diagnostic EEG pattern
specific for HIV meningoencephalitis has been reported, though
multiple studies confirmed that a variety of generalized and
focal abnormalities in the EEG may be seen, which are different
for the acute and chronic phases of the disease (43,44). In the
early acute phase, the EEG may be normal or may show the non-
specific pattern of diffuse slowing, while in the late stages, severe
delta slowing and even suppression can be seen. Encephalitides
caused by opportunistic infections in AIDS, such as Toxoplasma
gondii, Cytomegalovirus, and Candida, are discussed above, and
may show focal EEG abnormalities corresponding to the under-
lying lesions, as well as generalized slowing (45).
PART 2: IMMUNE-MEDIATED
INFLAMMATORY CNS CONDITIONS
In the conditions discussed in Part 1, the brain is the primary
or secondary target of attack by an infection or a presumed
infection. In immune-mediated conditions, no CNS infection
can be identified, and other organ systems may be the princi-
pal site of the inflammation. In cerebral demyelinating disor-
ders and Rasmussen encephalitis the principal site of the
inflammation appears to be the brain. In immune-mediated
conditions the EEG abnormalities may be diffuse or lesion-
specific. In lupus and antiphospholipid antibody diseases,
cerebral inflammation is partly secondary to systemic activa-
tion of pro-inflammatory cytokines, but the effects on vascula-
ture may also result in ischemia that then accounts for the focal
nature of some EEG abnormalities observed clinically. In para-
neoplastic CNS injury, although the tumor may be the primary
site of immune-mediated attack, the brain is the site of the
most damage. Some of the same immunogenic CNS antigens
also may be attacked in the absence of a neoplasm, and result
in similar damage.
Idiopathic Inflammatory Demyelinating Disorders
Acute disseminated encephalomyelopathy (ADEM) is an
umbrella diagnosis for a generally monophasic cerebral
demyelination event, with a range of manifestations from mild
to very severe. A prodromal event such an upper respiratory
infection, recent vaccination, or a flu-like illness is frequently
identified, as are recent trauma and surgery. The Marburg vari-
ant of fulminant multiple sclerosis (MS), acute hemorrhagic
 Chapter 18
leukoencephalitis (Hurst disease), tumefactive MS (Balo
disease), and Schilder disease in children are related conditions.
Although white matter is primarily affected in those condi-
tions, the inflammation may affect both normal cortical func-
tion and neuronal excitability, and result in EEG abnormalities
(46,47) that are varied and nonspecific, with both diffuse and
focal delta slowing typical for deep white matter lesions. There
may also be focal potentially epileptogenic discharges in tume-
factive MS and ADEM (47). With recovery the EEG usually nor-
malizes, somewhat proportional to the extent of recovery from
the initial event.
In chronic and late-stage MS, EEG is usually abnormal, with
diffuse slowing reflecting the degree of encephalopathy, how-
ever, without a clear correlation to radiographic plaque load
(48). Although overall only 6.5% of MS patients in a recent
study had epilepsy, the percentage was significantly higher in
progressive MS patients. The interictal EEG of MS patients with
epilepsy may reveal epileptiform abnormalities in up to two
third of these patients (49). MS plaques effectively disconnect
cortical areas, potentially leading to decreased inhibitory trans-
mission and ictogenesis.
Rasmussen Encephalitis
This progressive epileptic encephalopathy of uncertain etiology
and pathogenesis, involving one hemisphere, could be dis-
cussed with the rest of the epileptic encephalopathies. These are
childhood disorders characterized by cognitive, speech, hemi-
motor, and functional deterioration, and with progressive EEG
changes. However, traditionally Rasmussen disease has been
discussed in this section. It is likely that immune-mediated CNS
inflammation is part of the etiopathogenesis of other epileptic
encephalopathies. In the case of Rasmussen encephalitis, a pri-
mary role for the immune attack was proposed (50), but fails to
explain the disease adequately (51) as common CNS antigenic
responses are seen in many other CNS conditions.
Autoantibodies against glutamate receptors, once thought to
be specific for Rasmussen encephalitis, are found in other focal
epilepsies, ischemic infarcts, transient ischemic attacks (TIAs),
spinocerebellar atrophy, systemic lupus erythematosus (SLE),
and paraneoplastic encephalopathies, and are thus not useful in
the diagnosis of Rasmussen encephalitis. They are a biomarker
for brain injury particular to many encephalopathies, and may
be of importance in neurologic dysfunction in various condi-
tions. Immune suppression as a therapy in Rasmussen
encephalitis may slow the progression of cerebral degeneration
and functional decline, but has limited, and transient, effects on
the clinical or electrographic seizures, or on the progression and
evolvement of EEG abnormalities (52).
EEG abnormalities in both glutamate receptor antibody pos-
itive and negative diseases initially consist of lateralized slow-
ing, followed in months to years by the emergence of potentially
epileptogenic spikes and sharp wave discharges, often associ-
ated with clinical and electrographic seizures evolving into
the syndrome of EPC. Focal electrical status epilepticus
undetectable on standard scalp EEG has been demonstrated in
Rasmussen encephalitis (53). An early hemispherectomy may
■ The EEG in Inflammatory CNS Conditions 341
prevent or arrest the devastating course of the disease. Most
patients are hemiparetic by early adolescence, have uncontrol-
lable hemiconvulsions and severely reduced function by the late
teens, and require full-time or inpatient care by their twenties.
Late EEG findings reveal frequent unilateral seizures with a
wide field, and bilaterally abnormal background (Fig. 18.8).
Hashimoto Encephalopathy
Hashimoto encephalopathy (or encephalitis) (HE) or steroid-
responsive encephalopathy associated with autoimmune thy-
roiditis (SREAT) is a serious but treatable condition that is
probably underdiagnosed, partly because it has a broad range of
clinical presentations. The encephalopathy usually presents with
the subacute or abrupt onset of confusional state or alteration of
the level of consciousness, and is frequently accompanied by the
development of focal or generalized seizures. The most frequent
clinical symptoms in a recent meta-analysis (54) were seizures
(80%), confusion (52%), headache (40%), hallucinations
(32%), and ataxia (36%). Tremor, myoclonus, stroke-like events
(including transient aphasia), and rapid or gradually progressive
cognitive and behavioral decline are also seen (55,56).
Several thyroid-related autoantibodies have been found in
serum and CSF of patients with HT, reacting against thyroglob-
ulin, microsomal antigens, thyroid peroxidase, colloid-related
cell surface antigen, the thyroid-stimulating hormone (TSH)
receptor, as well as thyroxin and triiodothyroxine (57). In addi-
tion, patients have cell-mediated immunity directed against
thyroid antigens (58). High titers of antithyroid antibodies in
blood and high CSF protein without an accompanying pleocy-
tosis are often seen (59), but not reliably. A recent study found
that CSF IgG from HE patients specifically recognized two iso-
forms of dimethylargininase-I (DDAHI), and aldehyde reduc-
tase-I (AKRIAI) (60).
A report of a patient with HE who died of an unrelated con-
dition suggests a peculiar brain stem-dominated vasculitis
characterized by T-lymphocytic infiltrates of leptomeningeal
venules (58). Similar findings were seen in a case in our practice
(RL Beach, personal observations). In another case autopsy
showed dense lymphocytic infiltrates of the wall of the
parenchymal arterioles and venules (61).
The Mayo group reported mild to severe generalized slowing
on the EEG corresponding to the severity of the clinical
encephalopathy, as well as triphasic waves, epileptiform
abnormalities, photomyogenic, and photoparoxysmal
responses (62). Others have also found focal or lateralized slow-
ing and epileptiform discharges (Fig. 18.9), and seizures, in
addition to generalized slowing with high-voltage, delta, bipha-
sic, and triphasic waves (55,63). When follow-up EEGs have
been reported, the EEG findings improved or disappeared with
steroid treatment.
Electroencephalography, brain MRI, cerebrospinal fluid
examination, and serological tests are useful diagnostic tools,
but none is yet diagnostic in itself. The encephalopathy usually
responds well to corticosteroid therapy, although additional
immunosuppressive agents are sometimes used (64). The long-
term prognosis generally appears to be favorable.
342 Part III ■ Clinical EEG: General

Figure 18.8 These EEG tracings from a 24-year-old man with late sequelae of Rasmussen encephalitis and epilepsia par-
tialis continua (EPC) illustrate the variety of interictal and ictal abnormalities observed at any time of sampling the EEG.
Interictally (A), only a moderately severe slowing and a lack of normal background is noted: PLED-like pattern without a
clinical correlate is also common (B).

Limbic Encephalitis relies on the clinical picture combined with the demonstration of
The classic syndrome of limbic encephalitis includes subacute cog- MRI and EEG abnormalities in the temporal lobes, and presence of
nitive impairment with behavioral changes, hallucination, sleep inflammatory changes: in the cerebrospinal fluid (66), on imaging
disturbance, and temporal lobe seizures (56,65). The diagnosis (67,68), and rarely pathologically (69–71). Once considered an
 Chapter 18 ■ The EEG in Inflammatory CNS Conditions 343

Figure 18.8 (continued) Epilepsia partialis continua in Rasmussen syndrome (EPC) may present as morphologically differ-
ent seizure patterns that may have no clinical accompaniment (C); or may be accompanied by rhythmic movements (D).

extremely rare disorder, almost always related to cancer, and refrac-
tory to treatment, limbic encephalitis is now regarded as a relatively
frequent disorder, often unrelated to cancer, associated with
immune reaction to neuronal surface antigens, that may respond
to immunotherapy (65,72).
Regardless whether limbic encephalitis is paraneoplastic or
an autoimmune response to normal antigens, the clinical
and radiologic findings at presentation of these patients are
indistinguishable (56). The electroencephalographic findings
are nonspecific, may be similar to those seen in a variety of
344 Part III ■ Clinical EEG: General
Figure 18.9 This EEG shows interictal slowing and sharp waves in a patient with Hashimoto thyroiditis and subacute
encephalopathy. She had C lymphocytosis and increased protein, as well as dramatically elevated thyroid peroxidase and
thyroglobulin antibodies. She improved dramatically with a short course of high-dose solumedrol. Several years later she
relapsed and had similar findings and course.
encephalopathies or associated with other etiologic factors that
produce temporal lobe seizures. These EEG findings vary from
normal to severely abnormal.
The background activity varies from normal to variable
slowing, which may be focal or generalized (66,68,73). Focal
slowing occurs mostly in the temporal region (66–68) and may
be more prominent on the left (67,71). It may also be frontal or
unilateral hemispheric. Patients with generalized slowing gen-
erally had clinical and MRI evidence of more widespread cen-
tral nervous system disease, than patients with focal slowing
(55). In addition to generalized slowing, triphasic waves and
FIRDA have been described (68). In one study, electroen-
cephalography was reported to be the most sensitive investiga-
tion, with generalized or focal slowing as the most commonly
observed abnormality (66).
Epileptiform abnormalities that occur mostly in the tem-
poral region, where they usually are unilateral, bilateral,
bilateral independent, or multifocal independent, but they
may also be extratemporal, multifocal, or generalized
(66,68,75,76). The spikes or sharp waves are always most
prominent over the anteriorly situated temporal scalp elec-
trodes. Focal epileptiform activity is strongly associated with
clinical seizure and with MRI signal abnormalities in the
medial temporal regions (66). PLEDs or bilateral epilepti-
form discharges (BIPLEDs), predominantly involving the
temporal region, but also parietal and frontal regions, have
been described (66,73,77). Epileptiform EEG abnormalities
in temporal lobe were seen in 50% to 60% of patients with
paraneoplastic limbic encephalitis.
Electrographic seizures have onset over the temporal region
(Fig. 18.9) and may secondarily generalize (66,78). The clinical
deterioration of patient’s mental status correlates with worsen-
ing of EEG findings. Epileptogenic potential may start focally,
and then evolve into a more diffuse process, eventually involv-
ing the entire brain as the patient’s condition worsens. This may
become equivalent to EPC. This may be associated with focal or
diffuse pathologic abnormalities (Fig. 18.10) (73,77).
A very rapidly progressive nonconvulsive status epilepticus,
starting from focal pseudoperiodic spike and wave discharges,
extending to the contralateral hemisphere within a week and
becoming continuous sharp waves a day later, despite
antiepileptic medications was reported in one patient, without
any imaging abnormalities (79). Seizures in limbic encephalitis
are frequently refractory to antiepileptic medications and may
require surgical intervention (74,76,80–82).
Antiphospholipid Antibody Syndrome
The antiphospholipid syndrome (APS) is defined by the pres-
ence of laboratory markers such as antiphospholipid antibody
syndrome (APLA) and circulating anticoagulant (CAC) and
clinical manifestations such as recurrent thrombosis and
abortions (83).
 Chapter 18 ■ The EEG in Inflammatory CNS Conditions 345

Figure 18.10 This EEG shows a focal-onset, left temporal lobe seizure in a patient with antibodies to voltage potassium
channels in the absence of any neoplasm. He presented with intermittent and subacute progressive confusion and MRI
showed left greater than right temporal lobe increased signal on FLAIR. Interictal epileptiform discharges and subtle slow-
ing were also seen.

Among patients with neuropsychiatric symptoms, EEG
abnormalities are common and correlate with the presence of
APLA even in the absence of brain abnormalities on the MRI.
The main EEG abnormalities are bursts of bitemporal irregular
slowing, consisting of a mixture of theta and delta waves, often
with accentuation on the left during wakefulness. Epileptogenic
potentials such as temporal sharp waves and generalized epilep-
tiform abnormalities can also be seen (84). Among patients with
dementia or memory problems, the EEG may show diffuse or
bitemporal slowing and rarely a focal abnormality (77,83,84).
There is a relationship between antiphospholipid antibodies
and patients with epilepsy (83,85–87). The presence of
antiphospholipid antibodies is associated with recurrent
seizures in patients with refractory focal epilepsy, such as tem-
poral lobe epilepsy and extratemporal lobe epilepsy including
multifocal epilepsy (87).
Behcet Disease
Behcet disease (BD) is a multisystem disease of unknown
etiopathogenesis characterized by recurrent oral and genital
aphthous ulcers and uveitis with various clinical manifestations
due to central nervous involvement. Neurologic involvement
has been classified into two major forms: the intra-axial neuro-
Behcet syndrome (NBS), which is the inflammation of the
small veins of the parenchyma in the central nervous system,
and the extra-axial NBS that is the thrombosis of the cerebral
venous sinus (88).
The electroencephalographic findings in BD are nonspecific
and vary from normal to severely abnormal. Diffuse and focal
slowing in the delta and theta range occurs in patients with BD
(89–91). Focal slowing is seen over frontal, temporal, or fron-
totemporal regions (89,92). They are usually associated with
pathologic abnormalities on imaging studies (89,92). Abnormal
EEG findings are found in up to 80% of patients with neuro-
logic involvement (93), and up to one third of patients without
any neurologic symptoms (93,94).
EEG findings ranging from normal to severely abnormal
may be seen in patients with subclinical neurologic involve-
ment. Abnormal findings include focal slowing in the temporal
or parieto-occipital regions, spike waves in the temporal, occip-
ital, or frontal region, polyspikes, multifocal spikes, generalized
spikes, polyspike–wave complexes, secondary bilateral syn-
chrony, slow spike waves, 6-Hz spike waves, polyspike bursts,
photoparoxysmal response, photomyogenic response, and theta
and delta waves (94).
Neurologic involvement in BD varies and represents one of
the most severe BD manifestations (92). The reported fre-
quency of seizures varies from 0% to 27% (89,90,95–97).
Epileptic seizures occur as an initial neurologic presentation
(28,33) or as a sole neurologic manifestation of BD (92). They
also occur during exacerbations and chronic course of BD (89).
Seizures have been reported to be provoked by fever, medica-
tions (INH, penicillin, and interferon), or withdrawal of
antiepileptic medications (89).
346 Part III ■ Clinical EEG: General
Seizures may be simple partial, complex partial, myoclonic
jerks, generalized tonic–clonic seizures epilepsialis partialis con-
tinua (EPC) (85), or convulsive status epilepticus (89–92,97,98).
Focal epileptic activity over temporal and frontal temporal
regions has been reported (89). In one study, generalized
tonic–clonic convulsions accompanied by focal motor seizures
were the predominant type, and seizures were associated with
high mortality rate (89). Motor seizures, including epilepsia par-
tialis continua, are associated with a poor prognosis (82). Other
EEG findings reported include PLEDs (99,100).
There are conflicting findings regarding the abnormalities in
the CSF and the seizure frequency and severity (89,101). EEG
abnormalities, however, are correlated with CSF pleocytosis
(92), and the presence of (but not reliably the nature or location
of) clinical symptoms and progression of NBD (102).
Primary Angiitis of the Central Nervous System
Primary angiitis of the central nervous system (PACNS) is rare
multisystem vasculitis that primarily involves the central nerv-
ous system. Electroencephalography is a highly sensitive test in
the diagnostic of PACNS (103,104). The EEG abnormalities
tend to present early in the course of the disease, often being the
sole laboratory abnormality in the initial evaluation (104). The
lack of specificity compromises EEG’s usefulness in confirming
the diagnosis. The most common abnormality is diffuse slow-
ing, but at times it may be focal (103–105).
Focal symptoms occur in 50% or more of patients, and they
reliably occur in the setting of diffuse cortical dysfunction as
evidenced by encephalopathy and/or diffuse EEG changes
(104). Seizures have been noted in 10% to 25% of cases of
PACNS (103,104). Clear electrographic seizures are uncommon
(96). Partial motor seizures related to focal cerebral vasculitis in
the temporal and parietal lobes, progressing to generalized
tonic–clonic seizures that responded to steroid treatment, have
been reported (106,107).
Rheumatoid Arthritis
Cerebral involvement associated with rheumatoid arthritis
(RA) is rare but well documented. This includes vasculitis, cere-
bral pachymeningitis, leptomeningitis caused by fibrinoid infil-
tration and rheumatoid nodule, choroid plexus infiltration, or
cerebral complications induced by hyperviscosity associated
with high titers of rheumatoid factors presenting as
encephalopathy or seizures and responding to immunomodu-
latory treatment (108–111).
The morning symptoms of pain, weakness, and fatigue in
RA may relate to a nonrestorative sleep disorder associated with
prominent alpha rhythm EEG anomaly, a presumed arousal
state occurring within sleep, during NREM sleep stages 2, 3, and
4 (112).
In juvenile rheumatoid arthritis (JRA), various EEG abnor-
malities are seen including slowing of activity that may be con-
tinual or episodic, abundant beta activity, focal-asymmetrical
disturbances that could be either paroxysmal or nonparoxysmal,
and bilateral paroxysmal disturbances including photic-induced
paroxysm. Patients with positive antinuclear antibodies or vis-
ceral symptoms like pericarditis and/or myocarditis have fewer
pathologic EEG changes than those suffering from a form of the
disease with symptoms only in the joints. There is no clear cor-
relation between EEG changes and the duration of the disease or
age at onset (113). The pathologic EEG abnormalities are pre-
sumed to be caused by primary cerebral process connected with
the disease itself, probably vasculitis (113,114).
Sjogren Syndrome
Sjogren syndrome is an autoimmune disease that may be pri-
mary or secondary to another connective tissue disease charac-
terized by mononuclear infiltration and destruction of salivary
and lachrymal glands leading to xerostomia and xerophthalmia.
CNS involvement occurs in up to 68%, cognitive involvement
in 11%, and seizures in 8.5% of patients with Sjogren syndrome
(115). Various types of seizures occur in Sjogren syndrome
including partial, temporal mostly and parietal, and generalized
seizures (115).
Systemic Lupus Erythematosus
Abnormal electrographic findings have been reported in up to
87% of SLE patients (116). These are mostly nonspecific gener-
alized slowing in theta and delta ranges (117–119). In one
study, paroxysmal focal changes and abnormalities in the back-
ground were the most common findings (120). Patients with-
out neurologic deficits have abnormalities in the EEG,
suggesting the presence of subclinical involvement of central
nervous system (120). However, there is no significant associa-
tion between EEG abnormalities and cerebral atrophy, and the
EEG abnormalities are not correlated with functional deficits in
the neuropsychologic tests. Patients with EEG abnormalities
have significantly lower complement factor C4, more pro-
nounced thrombocytopenia, and significantly higher titer of
anticardiolipin IgG than those with normal EEG (119).
Unilateral and bilateral, focal and more widespread abnor-
malities including delta and theta slowing and sharp waves have
been reported; the left hemisphere, particularly the left tempo-
ral leads, is the most common abnormality (116,117,119).
Routine and quantitative EEG analyses also show theta and
delta slowing predominantly affecting the left hemisphere in
patients with neuropsychiatric manifestations (121).
Among SLE patients with headache, symptomatic headache
is associated with frequent paroxysmal changes and idiopathic
headache is associated with background changes. Idiopathic
headache shows more cerebral pathology on imaging study
including multiple perfusion deficits on SPECT and hyperin-
tense white matter lesions in the bilateral spheres on MRI (120).
Epileptic seizures occur in 8% to 37% of cases (116,123–125).
Seizures occur mostly after the onset of SLE, and acute sympto-
matic seizures are related to stroke and antiphospholipid anti-
bodies. Most patients with a single epileptic seizure have normal
EEG findings. Those with recurrent epileptic seizures generally
have abnormal EEGs, with interictal epileptiform abnormalities
predominant in the temporal lobes (125).
Generalized tonic–clonic and complex partial seizures are
the most common type of seizures. Other seizure types include
simple partial and secondary generalized (126). Complex par-
tial status epilepticus has been reported with a prolonged
 Chapter 18
period of confusion and EEG revealed continuous rhythmic
slow waves and triphasic waves (127); resolution of the EEG
abnormalities and improvement of clinical findings were noted
after a benzodiazepine injection.
Sydenham Chorea
Sydenham chorea is an immune-mediated movement disorder
associated with acute rheumatic fever, seen mainly in children
and adolescents. Three types of EEG abnormalities were
described in 17 of 20 children 10 to 16 years of age, with rheu-
matic chorea (128). Seven patients had increased amount of
posterior slowing, five patients had disruption of the alpha
rhythm and diffuse, monomorphic slowing, and five patients
had bursts of 2- to 4-Hz posterior delta. No correlation between
the EEG abnormalities and severity of the chorea, or between
localization of the EEG abnormalities and the distribution of
movements, was found. Others have found EEG slowing pro-
portional to the severity of the movement disorder (129,130),
and rarely epileptiform abnormalities have been observed
(129). Lateralized suppression of sleep spindles without
improvement with the disease regression has also been reported
(130). Others report the EEG may improve with the clinical
state, or show a delayed response (131).
Other Autoimmune Encephalitides
Kawasaki disease, postinfectious and postimmunization
encephalitis that does not have the imaging findings of ADEM,
as well as Vogt–Koyanagi–Harada disease—all may present with
transient meningoencephalitis of varied clinical severity. The
EEG pattern in those is nonspecific, reflecting the diffuse
encephalopathic state of the patient. Seizures or focal abnor-
malities are possible but uncommon.
SUMMARY
Quoting Dr. Westmoreland from the previous edition of this
chapter (2):
Although the EEG shows a variety of patterns with various
inflammatory processes, the EEG can help to (a) confirm and
indicate the degree and extent of CNS involvement by the
inflammatory process; (b) make the diagnosis of herpes simplex
encephalitis, Jacob–Creutzfeldt disease, and SSPE when the
characteristic electroencephalographic pattern of these diseases
is present; (c) indicate the presence of a focal lesion; (d) moni-
tor the course of the disease process; and (e) detect the develop-
ment of complications and sequelae.
In addition to her succinct summary of the usefulness of the
EEG in infectious and inflammatory disorders, we want to
emphasize the adjuvant diagnostic value of the EEG. Despite
the great range of possible EEG patterns at presentation and
their significant overlap, the initial diagnostic value of the EEG
should not be overlooked because:
• Many of the above-described inflammatory and infectious
conditions can present with NCSE, which is a life- threatening
neurologic emergency requiring immediate aggressive treat-
ment, and EEG remains the one modality to diagnose NCSE.
■ The EEG in Inflammatory CNS Conditions 347
• Some more specific patterns—PLEDs or FIRDA—may be
associated with treatable conditions that are otherwise not
suspected, for example, early HSV encephalitis without fever
or CSF abnormalities, or basal meningitis with developing
hydrocephalus.
• Gross hemispheric asymmetries uncovered by the EEG in
cases where imaging is delayed or equivocal may be diagnos-
tic for a lateralized process such as a brain abscess or a sub-
dural empyema, that is a neurosurgical emergency.
Of no less importance than the diagnosis is the information
that the EEG offers about the severity and extent of CNS disease,
which is crucial for monitoring the response to treatment.
Developing a library of EEG markers and patterns that may
accurately predict prognosis in inflammatory and infectious
CNS disease is a project for the future. With the maturation of
quantitative and digital EEG techniques there should be a
renewed interest in pursuing this. Despite the nonspecificity of
the brain response to inflammation, certain features may emerge
that are helpful in predicting the residual damage to the brain.
In conclusion, the EEG in inflammatory and infectious dis-
orders has significant diagnostic and prognostic value, and the
pathognomic as well as nonspecific pathologic EEG patterns
need to be recognized by the electroencephalographer.
REFERENCES
1. Radermecker FJ. Infections and Inflammatory Reactions, Allergy
and Allergic Reactions; Degenerative Diseases. Amsterdam: Elsevier
Scientific Publishing Co; 1997.
2. Westmoreland BF. The EEG in cerebral inflammatory processes.
In: Niedermeyer E, Lopes da Silva F, eds. Niedermeyer’s Textbook of
Electroencephalography: Basic Principles, Clinical Applications, and
Related Fields. 5th ed. Philadelphia: Lippincott, Williams and
Wilkins; 2005:323–338.
3. Kepa L, Oczko-Grzesik B, Warakomska I, et al. Badania EEG u
chorych z ropnymi, bakteryjnymi zapaleniami opon i mozgu-
przeglad publikacji i obserwacje wlasne [EEG examination in
patients with purulent, bacterial meningoencephalitis—
literature review and own observations]. Wiad Lek. 2006;
59(7–8):458–462.
4. Zoons E, Weisfelt M, de Gans J, et al. Seizures in adults with bacte-
rial meningitis. Neurology. 2008;70(22 pt 2):2109–2115.
5. Bartel P, Schutte CM, Becker P, et al. Discrimination between viral
and nonviral meningitis by visually analyzed and quantitative elec-
troencephalography. Clin Electroencephalogr. 1999;30(2):35–38.
6. Konno S, Sugimoto H, Nemoto H, et al. Triphasic waves in a
patient with tuberculous meningitis. J Neurol Sci. 2010;
291:114–117.
7. Kalita J, Misra UK, Das BK. SPECT changes and their correlation
with EEG changes in tuberculous meningitis. Electromyogr Clin
Neurophysiol. 2002;42(1):39–44.
8. Patwari AK, Aneja S, Ravi RN, et al. Convulsions in tuberculous
meningitis. J Trop Pediatr. 1996;42(2):91–97.
9. Wang RJ, Wang DX, Wang JW, et al. Analysis of 62 adult patients
with viral meningitis. Zhonghua Shi Yan He Lin Chuang Bing Du
Xue Za Zhi. 2009;23(3):218–220.
10. Pollak L, Klein C, Schiffer J, et al. Electroencephalographic abnor-
malities in aseptic meningitis and noninfectious headache. A com-
parative study. Headache. 2001;41(1):79–83.
348 Part III ■ Clinical EEG: General
11. Capouya JD, Berman DM, Dumois JA. Mollaret’s meningitis due
to human herpesvirus 6 in an adolescent. Clin Pediatr.
2006;45(9):861–863.
12. Gao B, Yang J, Zhuang S, et al. Mollaret meningitis associated with
an intraspinal epidermoid cyst. Pediatrics. 2007;120(1):e220–e224.
13. Mikdashi J, Kennedy S, Krumholz A. Recurrent benign lympho-
cytic (Mollaret) meningitis in systemic lupus erythematosus.
Neurologist. 2008;14(1):43–45.
14. Prandota J. Mollaret meningitis may be caused by reactivation of
latent cerebral toxoplasmosis. Int J Neurosci. 2009;119(10):
1655–1692.
15. Cobb W. The periodic events of subacute sclerosing leucoen-
cephalitis. Electroencephalogr Clin Neurophysiol. 1966;21(3):
278–294.
16. Wieser HG, Schwarz U, Blattler T, et al. Serial EEG findings in spo-
radic and iatrogenic Creutzfeldt–Jakob disease. Clin Neurophysiol.
2004;115(11):2467–2478.
17. Wieser HG, Schindler K, Zumsteg D. EEG in Creutzfeldt–Jakob
disease [see comment]. Clin Neurophysiol. 2006;117(5):935–951.
18. Cohen D, Kutluay E, Edwards J, et al. Sporadic Creutzfeldt–Jakob
disease presenting with nonconvulsive status epilepticus. Epilepsy
Behav. 2004;5(5):792–796.
19. Fernandez-Torre JL. Creutzfeldt–Jakob disease presenting as
refractory nonconvulsive status epilepticus. J Intensive Care Med.
2005;20(3):180–181.
20. Fernandez-Torre JL. Nonconvulsive status epilepticus in
Creutzfeldt–Jakob disease. Clin Neurophysiol. 2006;117(8):
1879–1880.
21. Parry J, Tuch P, Knezevic W, et al. Creutzfeldt–Jakob syndrome
presenting as epilepsia partialis continua. J Clin Neurosci.
2001;8(3):266–268.
22. Rossetti AO, Dunand M. Creutzfeldt–Jakob disease: evolution
from nonconvulsive status epilepticus, through SIRPIDs, to gener-
alized periodic discharges. Clin Neurophysiol. 2007;118(11):
2533–2536.
23. Shapiro JM, Shujaat A, Wang J, et al. Creutzfeldt–Jakob disease
presenting as refractory nonconvulsive status epilepticus.
J Intensive Care Med. 2004;19(6):345–348.
24. Lanska DJ. Diagnosis of Creutzfeldt–Jakob disease: effect of clini-
cal criteria on incidence estimates. Analysis of EEG and CSF
14-3-3 proteins as aids to the diagnosis of Creutzfeldt–Jakob dis-
ease. Neurology. 2001;56(10):1422–1423.
25. Keyrouz SG, Labib BT, Sethi R. MRI and EEG findings in
Heidenhain variant of Creutzfeldt–Jakob disease. Neurology.
2006;67(2):333.
26. Radermecker J, Poser CM. The significance of repetitive parox-
ysmal electroencephalographic patterns. Their specificity in
subacute sclerosing leukoencephalitis. World Neurol. 1960;1:
422–433.
27. Caselli RJ, Scheithauer BW, O’Duffy JD, et al. Chronic inflamma-
tory meningoencephalitis should not be mistaken for Alzheimer’s
disease. Mayo Clin Proc. 1993;68(9):846–853.
28. Kister I, Inglese M, Laxer RM, et al. Neurologic manifestations of
localized scleroderma: a case report and literature review.
Neurology. 2008;71(19):1538–1545.
29. Schott JM. Ischemic bilateral hippocampal dysfunction during
transient global amnesia. Neurology. 2008;70(20):1940–1941.
30. Smith JB, Westmoreland BF, Reagan TJ, et al. A distinctive clinical
EEG profile in herpes simplex encephalitis. Mayo Clin Proc.
1975;50(8):469–474.
31. Westmoreland BF, Sharbrough FW, Donat JR. Stimulus-induced
EEG complexes and motor spasms in subacute sclerosing panen-
cephalitis. Neurology. 1979;29(8):1154–1157.
32. Eriksson B, Wictor L. EEG with triphasic waves in Borrelia
burgdorferi meningoencephalitis. Acta Neurol Scand. 2007;116(2):
133–136.
33. Kerling F, Blumcke I, Stefan H. Pitfalls in diagnosing limbic
encephalitis—a case report. Acta Neurol Scand. 2008;118(5):
339–342.
34. Klein C, Kimiagar I, Pollak L, et al. Neurological features of West
Nile virus infection during the 2000 outbreak in a regional hospi-
tal in Israel. J Neurol Sci. 2002;200(1–2):63–66.
35. Gandelman-Marton R, Kimiagar I, Itzhaki A, et al.
Electroencephalography findings in adult patients with West Nile
virus-associated meningitis and meningoencephalitis. Clin Infect
Dis. 2003;37(11):1573–1578.
36. Bagic A, Boudreau EA, Greenfield J, et al. Electro-clinical evolution
of refractory non-convulsive status epilepticus caused by West Nile
virus encephalitis. Epileptic Disord 2007;9(1):98–103.
37. DellaBadia J Jr, Jaffe SL, Singh J, et al. An occipital lobe epilepto-
genic focus in a patient with West Nile encephalitis. Eur J Neurol.
2004;11(2):111–113.
38. Rodriguez AJ, Westmoreland BF. Electroencephalographic charac-
teristics of patients infected with West Nile virus. J Clin
Neurophysiol. 2007;24(5):386–389.
39. Misra UK, Kalita J, Syam UK, et al. Neurological manifestations of
dengue virus infection. J Neurol Sci. 2006;244(1–2):117–122.
40. Nishiya K, Taira K, Kuroki S. A case of secondary meningoen-
cephalitis associated with mycoplasma pneumoniae infection.
Kansenshogaku Zasshi. 2001;75(3):209–212.
41. Okamoto K, Fukuda M, Shigemi R, et al. Two cases of acute
encephalitis/encephalopathy associated with adenovirus type 3
infection. No to Hattatsu. 2004;36(6):487–491.
42. Misra UK, Kalita J. Seizures in encephalitis: predictors and out-
come. Seizure. 2009;18(8):583–587.
43. Bernad PG. The neurological and electroencephalographic
changes in AIDS. Clin Electroencephalogr. 1991;22(2):65–70.
44. Tinuper P, de Carolis P, Galeotti M, et al. Electroencephalogram
and HIV infection: a prospective study in 100 patients. Clin
Electroencephalogr. 1990;21(3):145–150.
45. Villar Del Saz S, Sued O, Falco V, et al. Acute meningoencephalitis due
to human immunodeficiency virus type 1 infection in 13 patients:
clinical description and follow-up. J Neurovirol. 2008;26:1–6.
46. Chabolla DR, Moore JL, Westmoreland BF. Periodic lateralized
epileptiform discharges in multiple sclerosis. Electroencephalogr
Clin Neurophysiol. 1996;98(1):5–8.
47. Levic ZM. Electroencephalographic studies in multiple sclerosis.
Specific changes in benign multiple sclerosis. Electroencephalogr
Clin Neurophysiol. 1978;44(4):471–478.
48. Striano P, Orefice G, Brescia Morra V, et al. Epileptic seizures in
multiple sclerosis: clinical and EEG correlations. Neurol Sci.
2003;24(5):322–328.
49. Martinez-Juarez IE, Lopez-Meza E, Gonzalez-Aragon Mdel C,
et al. Epilepsy and multiple sclerosis: increased risk among pro-
gressive forms. Epilepsy Res. 2009;84(2–3):250–253.
50. Rogers SW, Andrews PI, Gahring LC, et al. Autoantibodies to glu-
tamate receptor GluR3 in Rasmussen’s encephalitis. Science.
1994;265(5172):648–651.
51. Pleasure D. Diagnostic and pathogenic significance of glutamate
receptor autoantibodies. Arch Neurol. 2008;65(5):589–592.
 Chapter 18
52. Bien CG, Schramm J. Treatment of Rasmussen encephalitis half a
century after its initial description: promising prospects and a
dilemma. Epilepsy Res. 2009;86(2–3):101–112.
53. Fattouch J, Di Bonaventura C, Di Gennaro G, et al. Electrical sta-
tus epilepticus “invisible” to surface EEG in late-onset Rasmussen
encephalitis. Epileptic Disord. 2008;10(3):219–222.
54. Alink J, de Vries TW. Unexplained seizures, confusion or halluci-
nations: think Hashimoto encephalopathy. Acta Paediatr.
2008;97(4):451–453.
55. Rodriguez AJ, Jicha GA, Steeves TD, et al. EEG changes in a patient
with steroid-responsive encephalopathy associated with antibod-
ies to thyroperoxidase (SREAT, Hashimoto’s encephalopathy).
J Clin Neurophysiol. 2006;23(4):371–373.
56. Vernino S, Geschwind M, Boeve B. Autoimmune encephalo-
pathies. Neurologist. 2007;13(3):140–147.
57. Ferracci F, Moretto G, Candeago RM, et al. Antithyroid antibodies
in the CSF: their role in the pathogenesis of Hashimoto’s
encephalopathy. Neurology. 2003;60(4):712–714.
58. Nolte KW, Unbehaun A, Sieker H, et al. Hashimoto encephalopa-
thy: a brainstem vasculitis? Neurology. 2000;54(3):769–770.
59. Mocellin R, Walterfang M, Velakoulis D. Hashimoto’s encephalo-
pathy: epidemiology, pathogenesis and management. CNS Drugs.
2007;21(10):799–811.
60. Gini B, Lovato L, Cianti R, et al. Novel autoantigens recognized by
CSF IgG from Hashimoto’s encephalitis revealed by a proteomic
approach. J Neuroimmunol. 2008;196(1–2):153–158.
61. Striano P, Pagliuca M, Andreone V, et al. Unfavourable outcome of
Hashimoto encephalopathy due to status epilepticus. One autopsy
case. J Neurol. 2006;253(2):248–249.
62. Schauble B, Castillo PR, Boeve BF, et al. EEG findings in steroid-
responsive encephalopathy associated with autoimmune thyroidi-
tis. Clin Neurophysiol. 2003;114(1):32–37.
63. Sporis D, Habek M, Mubrin Z, et al. Psychosis and EEG abnormal-
ities as manifestations of Hashimoto encephalopathy. Cogn Behav
Neurol. 2007;20(2):138–140.
64. Marshall GA, Doyle JJ. Long-term treatment of Hashimoto’s
encephalopathy. J Neuropsychiatry Clin Neurosci. 2006;18(1):14–20.
65. Tuzun E, Dalmau J. Limbic encephalitis and variants: classifica-
tion, diagnosis and treatment. Neurologist. 2007;13(5):261–271.
66. Lawn ND, Westmorel BF, Kiely MJ, et al. Clinical, magnetic reso-
nance imaging, and electroencephalographic findings in paraneo-
plastic limbic encephalitis. Mayo Clinic Proc. 2003;78(11):
1363–1368.
67. Dirr LY, Elsted AD, Donofrio PD, et al. Evolution of brain MRI
abnormalities in limbic encephalitis. Neurology. 1990;40(8):
1304–1306.
68. Tan KM, Lennon VA, Klein CJ, et al. Clinical spectrum of voltage-
gated potassium channel autoimmunity. Neurology. 2008;70(20):
1883–1890.
69. Corsellis JA, Goldberg GJ, Norton AR. “Limbic Encephalitis” and
its association with carcinoma. Brain. 1968;91(3):481–496.
70. Bakheit AM, KennedyPG, Behan PO. Paraneoplastic limbic
encephalitis: clinico-pathological correlations. J Neurol Neurosurg
Psychiatry. 1990;53(12):1084–1088.
71. Dalmau J, Graus F, Rosenblum MK, et al. Anti-Hu--associated
paraneoplastic encephalomyelitis/sensory neuronopathy. A
Clinical Study of 71 Patients. Medicine. 1992;71(2):59–72.
72. Graus F, Saiz A, Lai M, et al. Neuronal surface antigen antibodies
in limbic encephalitis: clinical-immunologic associations.
Neurology. 2008;71(12):930–936.
■ The EEG in Inflammatory CNS Conditions 349
73. Gultekin SH, Rosenfeld MR, Voltz R, et al. Paraneoplastic limbic
encephalitis: neurological symptoms, immunological findings and
tumour association in 50 patients. Brain. 2000;123(pt 7):1481–1494.
74. Fadul CE, Stommel EW, Dragnev KH, et al. Focal paraneoplastic
limbic encephalitis presenting as orgasmic epilepsy. J Neurooncol.
2005;72(2):195–198.
75. Alamowitch S, Graus F, Uchuya M, et al. Limbic encephalitis and
small cell lung cancer. Clinical and immunological features. Brain.
1997;120(pt 6):923–928.
76. Jacobs DA, Fung KM, Cook NM, et al. Complex partial status
epilepticus associated with anti-Hu paraneoplastic syndrome.
J Neurol Sci. 2003;213(1–2):77–82.
77. Shavit YB, Graus F, Probst A, et al. Epilepsia partialis continua:
a new manifestation of anti-Hu-associated paraneoplastic
encephalomyelitis. Ann Neurol. 1999;45(2):255–258.
78. Thieben MJ, Lennon VA, Boeve BF, et al. Potentially reversible
autoimmune limbic encephalitis with neuronal potassium channel
antibody. Neurology. 2004;62(7):1177–1182.
79. Espay AJ, Kumar V, Sarpel G. Anti-Hu-associated paraneoplastic
limbic encephalitis presenting as rapidly progressive non-convul-
sive status epilepticus. J Neurol. Sci. 2006;246(1–2):149–152.
80. Mut M, Schiff D, Dalmau J. Paraneoplastic recurrent multifocal
encephalitis presenting with epilepsia partialis continua.
J Neurooncol. 2005;72(1):63–66.
81. Nahab F, Heller A, Laroche SM. Focal cortical resection for com-
plex partial status epilepticus due to a paraneoplastic encephalitis.
Neurologist. 2008;14(1):56–59.
82. Porta-Etessam J, Ruiz-Morales J, Millan JM, et al. Epilepsia par-
tialis continua and frontal features as a debut of anti-Hu paraneo-
plastic encephalomyelitis with focal frontal encephalitis. Eur JNeurol.
2001;8(4):359–360.
83. Chapman J, Abu-Katash M, Inzelberg R, et al. Prevalence and clin-
ical features of dementia associated with the antiphospholipid
syndrome and circulating anticoagulants. J Neurol Sci. 2002;
203–204:81–84.
84. Lampropoulos CE, Koutroumanidis M, Reynolds PP, et al.
Electroencephalography in the assessment of neuropsychiatric
manifestations in antiphospholipid syndrome and systemic lupus
erythematosus. Arthritis Rheum. 2005;52(3):841–846.
85. Cimaz R, Romeo A, Scarano A, et al. Prevalence of anti-
cardiolipin, anti-beta2 glycoprotein I, and anti-prothrombin anti-
bodies in young patients with epilepsy. Epilepsia. 2002;43(1):
52–59.
86. Verrot D, San-Marco M, Dravet C, et al. Prevalence and significa-
tion of antinuclear and anticardiolipin antibodies in patients with
epilepsy. Am J Med. 1997;103(1):33–37.
87. Liimatainen S, Peltola M, Fallah M, et al. The high prevalence of
antiphospholipid antibodies in refractory focal epilepsy is related
to recurrent seizures. Eur J Neurol. 2009;16(1):134–141.
88. Siva A, Altintas A, Saip S. Behcet’s syndrome and the nervous
system. Curr Opin Neurol. 2004;17(3):347–357.
89. Aykutlu E, Baykan B, Serdaroglu P, et al. Epileptic seizures in
Behcet disease. Epilepsia. 2002;43(8):832–835.
90. Joseph FG, Scolding NJ. Neuro-Behcet’s disease in caucasians:
a study of 22 patients. Eur J Neurol. 2007;14(2):174–180.
91. Aktekin B, Dogan EA, Oguz Y, et al. Epilepsia partialis continua in
a patient with Behcet’s disease. Clin Neurol Neurosurg. 2006;
108(4):392–395.
92. Chroni E, Monastirli A, Polychronopoulos P, et al. Epileptic
seizures as the sole manifestation of neuro-Behcet’s disease:
350 Part III ■ Clinical EEG: General
complete control under interferon-alpha treatment. Seizure.
2008;17(8):744–747.
93. Midorikawa T. Electroencephalographic studies on Behcet’s
disease—with special reference to the electroencephalogram of
neuro-Behcet’s disease. Rinsho Shinkeigaku. 1975;15(11):
860–869.
94. Tunc T, Ortapamuk H, Naldoken S, et al. Subclinical neurological
involvement in Behcet’s disease. Neurol India. 2006;54(4):408–411.
95. Lannuzel A, Lamaury I, Charpentier D, et al. Neurological mani-
festations of Behcet’s disease in a Caribbean population: clinical
and imaging findings. J Neurol. 2002;249(4):410–418.
96. Al-Araji A, Sharquie K, Al-Rawi Z. Prevalence and patterns of
neurological involvement in Behcet’s disease: a prospective study
from Iraq. J Neurol Neurosurg Psychiatry. 2003;74(5):608–613.
97. Akman-Demir G, Baykan-Kurt B, Serdaroglu P, et al. Seven-year
follow-up of neurologic involvement in Behcet syndrome. Arch
Neurol. 1996;53(7):691–694.
98. Mead S, Kidd D, Good C, et al. Behcet’s syndrome may present
with partial seizures. J Neurol Neurosurg Psychiatry. 2000;68(3):
392–393.
99. Pourmand R, Markand ON, Cook JA. Periodic lateralized EEG
abnormality in a case of Neuro-Behcet syndrome. Clin
Electroencephalogr. 1984;15(2):122–124.
100. Kihara M, Takahashi M, Mitsui Y, et al. A case of neuro-Behcet’s
encephalitis with pleds as distinct from herpes simplex
encephalitis: a differential diagnosis. Funct Neurol. 1996;11(2–3):
99–103.
101. Campanile S, Amore M, Cerisoli M, et al. Neurological complica-
tions in Behcet’s disease. Riv Patol Nerv Ment. 1984;105(5):229–237.
102. Matsumoto K. Correlation between EEG and clinicopathological
change in neuro-Behcet’s syndrome. Folia Psychiatr Neurol Jpn.
1984;38(1):65–79.
103. Calabrese LH, Mallek JA. Primary angiitis of the central nervous
system. Report of 8 new cases, review of the literature, and proposal
for diagnostic criteria. Medicine (Baltimore). 1988;67(1):20–39.
104. Vollmer TL, Guarnaccia J, Harrington W, et al. Idiopathic granu-
lomatous angiitis of the central nervous system. Diagnostic chal-
lenges. Arch Neurol. 1993;50(9):925–930.
105. Carandang CG, Grant AL. Delirium and isolated angiitis of the
central nervous system: a case report and review. CNS Spectr.
2008;13(3):209–213.
106. Berger JR, Romano J, Menkin M, et al. Benign focal cerebral vas-
culitis: case report. Neurology. 1995;45(9):1731–1734.
107. Ilhan A, Budak F. Primary angiitis of the central nervous system:
unusual clinical presentation. Int J Angiol. 2000;9(1):23–26.
108. Gobernado JM, Leiva C, Rabano J, et al. Recovery from rheuma-
toid cerebral vasculitis. J Neurol Neurosurg Psychiatry. 1984;
47(4):410–413.
109. Kim RC, Collins G. The neuropathology of rheumatoid disease.
Hum Pathol. 1981;12(1):5–15.
110. Neamtu L, Belmont M, Miller DC, et al. Rheumatoid disease of
the CNS with meningeal vasculitis presenting with a seizure.
Neurology. 2001;56(6):814–815.
111. Zolcinski M, Bazan-Socha S, Zwolinska G, et al. Central nervous
system involvement as a major manifestation of rheumatoid
arthritis. Rheumatol Int. 2008;28(3):281–283.
112. Moldofsky H, Lue FA, Smythe HA. Alpha EEG sleep and morn-
ing symptoms in rheumatoid arthritis. J Rheumatol. 1983;
10(3):373–379.
113. Lang H, Anttila R, Svekus A, et al. EEG findings in juvenile
rheumatoid arthritis and other connective tissue diseases in chil-
dren. Acta Paediatr Scand. 1974;63(3):373–380.
114. Sillanpaa M, Lang AH, Kalimo H. Natural history of juvenile
rheumatoid arthritis. A follow-up study of a case with special ref-
erence to clinical, electroencephalographic and neuropathologi-
cal findings. Acta Paediatr Scand. 1978;67(4):537–541.
115. Delalande S, de Seze J, Fauchais AL, et al. Neurologic manifesta-
tions in primary Sjogren syndrome: a study of 82 patients.
Medicine (Baltimore). 2004;83(5):280–291.
116. Glanz BI, Schur PH, Khoshbin S. EEG abnormalities in systemic
lupus erythematosus. Clin Electroencephalogr. 1998;29(3):128–131.
117. Gibson T, Myers AR. Nervous system involvement in systemic
lupus erythematosus. Ann Rheum Dis. 1975;35(5):398–406.
118. Abel T, Gladman DD, Urowitz MB. Neuropsychiatric lupus.
J Rheumatol. 1980;7(3):325–333.
119. Waterloo K, Omdal R, Jacobsen EA, et al. Cerebral computed
tomography and electroencephalography compared with neu-
ropsychological findings in systemic lupus erythematosus.
J Neurol. 1999;246(8):706–711.
120. Gora MK, Dobrzynska L, Mierzejewska E, et al. EEG examination
in patients with systemic lupus erythematosus. Wiad Lek. 2003;
56(5–6):220–226.
121. Glanz BI, Laoprasert P, Schur PH, et al. Lateralized EEG findings
in patients with neuropsychiatric manifestations of systemic
lupus erythematosus. Clin Electroencephalogr. 2001;32(1):14–19.
122. Gora M, Dobrzynska L, Koseda-Dragan M. Headache in patients
with systemic lupus erythematosus: analysis of clinical EEG and
neuroimaging studies. Pol Arch Med Wewn. 2003;109(5):
497–502.
123. Jennekens FG, Kater L. The central nervous system in systemic
lupus erythematosus. Part 1. Clinical syndromes: a literature
investigation. Rheumatology (Oxford). 2002;41(6):605–618.
124. Ainiala H, Loukkola J, Peltola J, et al. The prevalence of neuropsy-
chiatric syndromes in systemic lupus erythematosus. Neurology.
2001;57(3):496–500.
125. Appenzeller S, Cendes F, Costallat LT. Epileptic seizures in sys-
temic lupus erythematosus. Neurology. 2004;63(10):1808–1812.
126. Gonzalez-Duarte A, Cantu-Brito CG, Ruano-Calderon L, et al.
Clinical description of seizures in patients with systemic lupus
erythematosus. Eur Neurol. 2008;59(6):320–323.
127. Tsuji M, Tanaka H, Yamakawa M, et al. A case of systemic lupus
erythematosus with complex partial status epilepticus. Epileptic
Disord. 2005;7(3):249–251.
128. Terzano MG, Camillo Manzoni G, Mancia D, et al. Valutazione
degli aspetti E.E.G. in rapporto con i parametri clinico-evolutivi
nella corea reumatica [Evaluation of the EEG aspects of rheu-
matic chorea as related to the clinico-evolutive parameters of the
disease]. Riv Neurol. 1979;49(6):451–470.
129. Johnson DA, Klass DW, Millichap JG. Electroencephalogram in
Sydenham’s chorea. Arch Neurol. 1964;10:21–27.
130. Usher SJ, Jasper HH. The etiology of Sydenham’s chorea.
Electroencephalographic studies. CMAJ. 1941;44:365–371.
131. Ganji S, Duncan MC, Frazier E. Sydenham’s chorea: clinical, EEG,
CT scan, and evoked potential studies. Clin Electroencephalogr.
1988;19(3):114–122.
132. Lavy S, Lavy R, Brand A. Neurological and electroencephalo-
graphic abnormalities in rheumatic fever. Acta Neurol Scand.
1964;40:76–88.

Cerebrovascular Diseases and EEG
BARBARA TETTENBORN, ERNST NIEDERMEYER, AND DONALD L. SCHOMER
ROLE OF EEG IN DIAGNOSIS OF
CEREBROVASCULAR DISEASE
Computed tomography (CT) and magnetic resonance imaging
(MRI) are the major diagnostic tools in cerebrovascular disease
considering the morphology is a vascular lesion. The major role
of electroencephalography (EEG) is the evaluation of functional
disturbances caused by cerebrovascular disease, mainly the detec-
tion of epileptogenic foci. Cerebrovascular events are acute, often
catastrophic events that naturally have strong influences on the
EEG. The tempo of evolution is an important factor. Acute vas-
cular slow foci may be much more impressive than focal slowing
caused by a slowly growing neoplasm. On the other hand, small
deep vascular lesions are often too far distant from the cortex and
may escape EEG detection, whereas deep vascular lesions in the
brainstem and cerebellum reveal themselves only as secondary
phenomena in the cerebral hemispheres. In an acute stroke, a
massive and highly impressive EEG focus may be present before
a CT scan can demonstrate the lesion (1). MRI has been found to
be more precise and more informative than CT in the morpho-
logic evaluation of stroke. The EEG is an indicator of abnormal
physiologic function; the function of the involved central nerv-
ous system tissue breaks down before the structure shows its suf-
fering. It has been shown experimentally (2) as well as clinically
with the use of CT scan (3) that edema alone does not account
for delta activity in EEG. In the modern view, the factor of
ischemic tolerance of the brain has been critically analyzed (4)
along with the demonstration of a large number of biomolecular
substances presumed to be involved in this complex process.
The role of EEG as an indicator of disturbed function, locally
as well as diffusely, in the event of a stroke has not always been
duly appreciated. Decades ago, Hachinski and Norris (5) pointed
out in their study on acute stroke that “electroencephalography is
not much help in the diagnosis of stroke, especially when CT
scanning is available.” But apposed to that, even in our days,
including MRI techniques, it has to be stated that the evaluation
of the degree and extension of cerebral dysfunction may be just
as important as the demonstration of local or diffuse morpho-
logic changes. An experienced electroencephalographer might be
able to provide valuable information about the regional and gen-
eral functional or dysfunctional state in a stroke. An imaginative
approach to the EEG evaluation of strokes has been proposed by
Velho-Groneberg (6). Accordingly, the EEG reflects dynamic
processes that occur in the wake of a stroke.
In addition to local or more diffuse slow activity, the EEG
also shows a more or less serious decline of physiologic patterns
such as posterior alpha rhythm and intermixed slower or faster
CHAPTER
19
frequencies. Considerable localizing value has been ascribed to
local depression of the background activity, whereas more exten-
sive depression is regarded as somewhat less informative (7–9).
These authors believe that preservation of faster background
activities, especially in subacute forms of middle cerebral artery
(MCA) ischemia, is indicative of considerable neuronal survival
in the infarcted zone and hence indicative of a good prognosis.
Recent studies by Hilker et al. (10) could also indicate the absence
of delta activity and presence of theta and fast beta frequencies
within the EEG focus in stroke patients as predictor of a benign
course, whereas diffuse slowing and focal slow delta frequency
may point toward a more malignant course.
Even considering new MRI techniques, including functional
MRI, EEG still has its place in diagnostic workup of stroke
patients. Not only is MRI more expensive, but it also can be some-
what difficult to perform in the acutely ill stroke patient, whereas
EEG can easily be done and repeated in practically every patient.
HYPERVENTILATION
Hyperventilation as a routine procedure in clinical EEG can
activate otherwise latent EEG abnormalities in patients with
cerebrovascular disorders. It is probably the hypoxic effect of
hyperventilation that produces these changes. This activation
can give rise to focal EEG slowing in cases of moyamoya disease
(11). Kraaier et al. (12) investigated hyperventilation-induced
changes with quantitative EEG and considered this method a
model for transient ischemia in normal subjects.
RELEVANCE OF ACTIVATION
METHODS TO THE ASSESSMENT
OF CEREBROVASCULAR DISORDERS
At present, the role of these tests is rather limited. The fact
remains, however, that these tests have given new insights into
hemodynamic and neurophysiologic mechanisms underlying
the EEG activity. An interesting overview of the cerebral hemo-
dynamic evaluation has been presented by Carney et al. (13)
with the EEG aspects of this work being aptly described by
Anderson et al. (14).
ACUTE STROKE DETECTION
In the emergency evaluation of acute stroke, the progression
from transient ischemia to infarction is a dynamic process, which
is difficult to evaluate based solely on clinical grounds and
anatomic imaging. Advances in CT and MRI scanning, including
351
352 Part III ■ Clinical EEG: General
diffusion-weighted MRI (DWI), are providing new insights into
the pathophysiology of cerebral ischemia. Perfusion measures,
such as perfusion-weighted imaging, may be able to estimate
infarcted tissue with information about blood flow levels, but
these measures may be less sensitive to dynamic changes in tissue
viability than the EEG. Within seconds, the EEG reflects changes
in cerebral blood flow and metabolism. Both human and animal
evidence indicates that the ischemic cascade is rapid, with
marked changes occurring over seconds and minutes, and irre-
versible changes occurring within a few hours. DWI and EEG
detect pathophysiologic responses to reduced blood flow around
the same range, between 30 and 40 mL/100 g/min (15), suggest-
ing that they are comparable in their sensitivity to the stage of
ischemic gradient. The accuracy of EEG scalp recordings, how-
ever, may be limited by adequate sampling of the electrical field.
Biophysical simulations have indicated for many years that the
spatial frequency of the scalp EEG requires higher sampling
densities to avoid artificial smoothing of the scalp potential
field caused by undersampling (16,17). Luu et al. (18), in their
study on acute stroke patients, found within 8 to 36 hours after
symptom onset using a 128-channel sensor montage, that an
accurate description of stroke-related topographic EEG changes
is dependent on adequate spatial sampling density. Accurate
description of the spatial distribution of the stroke-related EEG
was achieved only with the 64- and 128-channel EEG. As the
recording density decreases to 32 channels, the distribution of the
scalp EEG spectra is distorted, potentially resulting in mislocal-
ization of the affected region. Their results demonstrated that the
EEG acquired with the conventional 19-channel montage is clin-
ically useful, if the data are reviewed by an expert, but localizing
information is limited by the available data. Nevertheless, for
daily clinical routine 128-channel EEG recording for every stroke
patient is not realistic. Therefore, studies on the value of conven-
tional 19-channel montage using the international 10-20 system
are currently being done. Conventional EEG recordings could be
useful in predicting the subgroup of acute stroke patients who are
predisposed to develop epileptic seizures in the follow-up.
In patients with large MCA infarction, space-occupying brain
edema may lead to a malignant course with up to 80% mortal-
ity using conservative treatment. As interventional treatment
strategies must be started before the deterioration occurs, pre-
dictors of a malignant course are necessary. Hilker et al. (10)
investigated the value of EEG within 24 hours after the onset of
stroke in 25 patients suffering a large MCA infarction. Their
findings indicated that the absence of delta activity and the pres-
ence of theta and fast beta frequencies within the focus predict a
benign course, whereas diffuse generalized slowing and slow
delta activity in the ischemic hemisphere may point toward a
malignant course. Therefore, EEG can deliver useful informa-
tion to select those patients who develop malignant edema.
DETERIORATING STROKES
Various complications may interfere with recovery from a stroke.
The term “deteriorating stroke” has been used by Hachinski and
Norris (5) and includes progressively evolving strokes, recurrent
strokes, and also spatially extending strokes. Their causes may be
thrombotic, embolic, or hemorrhagic. Important pathogenic
mechanisms include cerebral edema and the syndrome of inap-
propriate antidiuretic hormone, which leads to reduction of cere-
bral sodium, potassium, and chloride content, altered brain
energy metabolism, and neurotransmitter amino acid uptake (5).
Cardiac factors, poststroke depression, and drug effects also may
lead to poststroke deterioration. In evolving and deteriorating
strokes, the ischemic penumbra plays a major role (19). In dete-
riorating strokes, the EEG slowing tends to become more diffuse.
It also may reflect drug effects (sedatives), but it is apt to remain
notably unaltered in the case of poststroke depression.
EEG IN HEMISPHERIC STROKE
The most common cause of a hemispheric ischemic lesion is
atherosclerotic disease of the internal carotid artery or MCA.
Size of infarct and clinical symptoms relate to localization and to
collateral circulation. An infarct may encompass the entire terri-
tory supplied by carotid artery, followed in a few days by severe
edema, raised intracranial pressure, transtentorial herniation,
and even death. At the other extreme, occlusion of carotid artery
may occur without obvious clinical deficits, because of adequate
collateral circulation. In acute stages, EEG is usually the most
sensitive laboratory index. Massive hemispheric infarct due to
carotid occlusion causes marked EEG abnormalities (20–22):
• Widespread arrhythmic delta activity occurs over the
involved hemisphere, most prominently in temporal or fron-
totemporal regions (Fig. 19.1). In patients with smaller hemi-
spheric infarcts, EEG from uninvolved hemisphere usually
remains normal.
• Ipsilateral alpha rhythm may be absent or of lower amplitude
and slower frequency. Reactivity of alpha rhythm may dimin-
ish or disappear (Fig. 19.2). Rolandic beta activity and sleep
spindles may also disappear.
• Extensive infarcts may markedly decrease or “suppress” elec-
trocerebral activity, normal and abnormal, over the affected
hemisphere during the first few days. In patients with smaller
hemispheric infarcts, EEG from uninvolved hemisphere usu-
ally remains normal. With extensive infarcts causing cerebral
edema and displacement of midline structures, variable
degrees of slow activity appear in contralateral hemicranium,
often with bilateral intermittent rhythmic 2 to 3 per second
delta activity (IRDA) (Fig. 19.3), which may be marked in the
frontal region (FIRDA, frontal intermittent rhythmic delta
activity) (Fig. 19.4). In children, this rhythmic delta activity
may be most marked occipitally (OIRDA, occipital intermit-
tent rhythmic delta activity). Marked diffuse EEG changes
may make lateralization difficult; also, massive infarcts caus-
ing ipsilateral voltage depression and prominent contralateral
slow activity can lead to erroneous lateralization.
• With acute hemispheric infarctions, particularly embolic,
epileptiform activity occurs commonly and tends to be peri-
odic (periodic lateralized epileptiform discharges [PLEDs])
(Fig. 19.5). Such patients are usually obtunded and have focal
seizures with corresponding neurologic deficits. PLEDs and
clinical manifestations usually resolve in 1 to 2 weeks.

50 l V
Figure 19.1 Age 57 years. Acute cerebrovascular event on the day
before this record was obtained. There is marked left frontotemporal
polymorphic delta activity. Also note alpha depression and loss of detail
over left posterior quadrant. The right hemisphere and especially the
right frontal area show some degree of delta activity.
During the first days after infarction, EEG abnormalities are
most prominent, with extent and voltage of slow activity progres-
sively declining. Few prospective studies have evaluated EEGs seri-
ally in a carefully defined population. Kayser-Gatchalian and
Neuendörfer (20) did EEGs on 79 patients with infarcts in carotid
distribution. Initial EEGs were done within 48 hours after infarc-
tion and again on days 5, 10, and 22. They visually classified EEG
abnormalities into “background slowing” and “focal changes,”
with each group being divided by defined criteria into the follow-
ing categories: “none,” “mild,” “moderate,” or “severe.” In the initial
EEGs, “background slowing” reliably predicted state of conscious-
ness: only 1 of 14 patients with severe or moderate slowing was
alert, whereas 33 of 44 patients without slowing were alert. Focal
changes paralleled motor deficits: 39 of 51 patients with moderate
or severe focal changes had severe paresis or hemiplegia, whereas
only 7 of 17 patients without focal changes had severe weakness.
Background slowing in the initial EEG predicted outcome on day
22: only 2 of 14 patients with moderate or severe slowing
improved, whereas 38 of 44 patients without slowing improved.
Persistent focal changes paralleled persistent severe hemiparesis or
hemiplegia. In a smaller series of 15 patients, Sainio et al. (23)
compared visual with spectral analysis. Results of visual analysis
equaled spectral analysis, except for a lower incidence of focal
Chapter 19 ■ Cerebrovascular Diseases and EEG 353
50 l V
Figure 19.2 Recent acute cerebrovascular ischemia (3 days earlier)
due to right middle cerebral artery thrombosis and good CT scan evi-
dence of infarction in the corresponding territory. Acute left hemiplegia.
Age 64 years. Patient awake; right-sided alpha diminished and a large
zone of mixed 3 to 6 per second activity involving most of the right
hemisphere.
abnormalities; in general, their findings paralleled the findings of
Kayser-Gatchalian and Neuendörfer (20). Although unusual in
the acute phase of the infarction, epileptiform discharges can
evolve with time, possibly heralding postinfarction seizures.
In patients with large MCA infarction, space-occupying brain
edema may lead to a malignant course with up to 80% mortal-
ity under conservative treatment. As interventional treatment
strategies must be started before the deterioration occurs, pre-
dictors of a malignant course are necessary. Burghaus et al. (24)
reported the results of a study on early EEG within 24 hours
after the onset of stroke in 25 patients suffering a large MCA
infarct (12 patients with a malignant and 13 with a nonmalig-
nant course). EEG analysis was performed according to well-
established indicators for focal as well as global changes. Their
results indicated that the absence of delta activity and the pres-
ence of theta and fast beta frequencies within the focus predict a
benign course (P 0.05), whereas diffuse generalized slowing
and slow delta activity in the ischemic hemisphere may point to
a malignant course.
Infarction due to occlusion of the anterior cerebral artery is
rare. EEG findings consist of (i) ipsilateral frontal delta activity
(either arrhythmic or rhythmic, or both) and (ii) depression of
anterior beta activity.
354 Part III ■ Clinical EEG: General

Figure 19.3 A 79-year-old male patient presented with acute onset of aphasia, right-sided hemiparesis, and gait instability.
CT scan 3 hours after onset of symptoms without evidence of acute infarction. EEG showed posterior 9 per second alpha
rhythm on the right, suppressed on the left. Generalized intermittent temporoparietal polymorphic theta/ delta slowing (IRDA).

Occlusion of the posterior cerebral artery may result in (i)
infarction of the mesial temporal lobe with arrhythmic delta
activity over the posterior head region and (ii) marked disor-
ganization or absence of alpha rhythm.
PLEDS: CLINICAL CORRELATES
PLEDs are a relatively uncommon electroencephalographic
pattern characterized by lateralized or focal periodic or near-
periodic spike, spike-wave, or sharp wave complexes present
throughout most or all of the recording (Fig. 19.5B). Chatrian
et al. introduced the term PLEDs in 1964 (25), though the phe-
nomenon was first described in 1952 by Echlin et al. (26). PLEDs
occur in all age groups, from infants to adults. They are usually
seen transiently in the setting of an acute destructive cerebral
lesion and occur early in the course of patient illness. They can
be seen less commonly with systemic disturbances and a remote
cerebral lesion. Rare chronic PLEDs, persisting for a period of
3 months to more than 20 years, have been reported in patients
with chronic brain lesions and associated partial seizure disor-
ders. Bilateral, independently occurring PLEDs (BiPLEDs) were
recognized by Chatrian in 1964 (25) and formally characterized
in 1981 by de la Paz and Brenner (27). They are seen in the
setting of multifocal or diffuse cerebral injury, such as anoxia,
and herald a less favorable prognosis with higher mortality.
Approximately 80% to 90% of patients with PLEDs experience
clinical seizure activity, primarily focal motor seizures. In 1991,
Reiher et al. (28) described the entity “PLEDs plus,” character-
ized by PLEDs admixed with high-frequency, low-voltage
“polyspike” rhythms. These have an even stronger correlation
with clinical seizures and status epilepticus. PLEDs are generally
not considered an ictal pattern, though this has been reported,
and remains a matter of ongoing debate (29–31).
EEG IN VERTEBROBASILAR ISCHEMIC
DISEASE
Ischemia in the vertebrobasilar territory includes infarction of
brainstem, diencephalon, or cerebellum with varying neurologic
deficits depending on specific artery involved and extent and
location of the infarct. Disparity between extensive neurologic
deficits and EEG changes characterize brainstem and also most
cerebellar infarcts (Figs. 19.6 and 19.7). Less than one third of
patients with limited infarcts to the brainstem show even slightly
altered EEGs (32). Basilar artery occlusion infarcting ventral pons
but sparing tegmentum causes a distinctive clinical picture
 Chapter 19 ■ Cerebrovascular Diseases and EEG 355

Figure 19.4 A 76-year-old female patient presented with fluctuating aphasia, hemianopia, and right hemiataxia.
Symptoms lasted only a few hours. MRI revealed no acute infarction but signs of a chronic microangiopathic encephalopa-
thy with multiple subcortical infarctions. EEG showed intermittent theta slowing in the left temporal region and intermit-
tent frontal rhythmic delta activity (FIRDA).
Figure 19.5 A,B: A 71-year-old female with acute onset left-sided hemiparesis. MRI revealed paraventricular right-sided
ischemia. EEG showed regional slowing on the right side, predominantly temporal. Here also periodic lateralized epilepti-
form discharges (PLEDS) with maximum over T4 are noticed.
 Chapter 19 ■ Cerebrovascular Diseases and EEG 357

Figure 19.6 A 59-year-old male with sudden onset of vertigo, headache, and general weakness. MRI revealed extended
cerebellar infarction. Etiology was cardioembolic with known atrial fibrillation under insufficient anticoagulation. EEG is
inconspicuous with normal posterior alpha activity.
358 Part III ■ Clinical EEG: General

Figure 19.7 A 69-year-old male with sudden onset of nausea, dizziness, and weakness of right arm. MRI shows infarc-
tion in the territory of the superior cerebellar artery. EEG is normal.

known as “deefferented” or “locked-in” syndrome (33). Patients
are alert but are mute and tetraplegic, with the only preserved
motor acts being blinks and vertical eye movements. In these
patients, EEG is normal (34,35); the alpha rhythm reacts to vari-
ous stimuli, including intermittent photic stimulation that pro-
duces photic responses. Patients with bilateral pontine infarcts
involving tegmentum show striking dissociation between clinical
and EEG findings (25,36). Despite coma and extensive bilateral
motor deficits, EEGs may show activity of alpha frequency.
However, this alpha activity fails to react to various stimuli such
as intermittent photic stimulation; hence, it is abnormal. Because
of the association of coma and alpha activity, the term “alpha
coma” has been applied to such patients (37). A similar associa-
tion also occurs in patients rendered comatose by cardiopul-
monary arrest (38) or drug overdose (39).
At more rostral level, infarcts of midbrain and diencephalon
cause not only coma but also EEG alterations. Bilaterally syn-
chronous and symmetrical theta–delta activity, sometimes maxi-
mal over frontal regions, replaces normal rhythms. Nevertheless,
these relatively mild changes, compared with the depth of coma
and extent of neurologic deficits, should alert EEGers to an
intrinsic brainstem lesion. Thrombosis of rostral portion of basi-
lar artery can also cause infarcts of occipital lobes. EEGs show
posterior arrhythmic delta activity with loss of alpha rhythm,
unilaterally or bilaterally.
Jouvet (40) has correlated sites of brainstem lesions with
alterations in consciousness and changes in EEG activity. He
has concluded the following:
• Lesions of medulla oblongata and caudal pons alter usually
neither consciousness nor EEG.
• Extensive lesions of upper pons and lower mesencephalon
involving more than one half of tegmentum usually cause
coma. EEGs may show alpha activity that does not react to
external stimuli. Although vital in maintaining conscious-
ness, mesopontine regions seemingly contribute little to gen-
esis of EEG activity.
• Lesions of rostral brainstem and caudal diencephalon cause
not only deep coma but also diffuse, nonreactive theta and
delta activity.
In bilateral paramedian thalamic infarction, triphasic waves
can appear on electroencephalogram at acute stage (41).
Triphasic waves are nonspecific EEG findings occurring in both
metabolic and nonmetabolic conditions. The triphasic waves in
bilateral paramedian thalamic infarction might be related to
level of consciousness and not necessarily predictive of poor
prognosis.
EEG IN SUBCORTICAL ARTERIOSCLEROTIC
ENCEPHALOPATHY
The incidence of epileptic seizures in subcortical vascular
encephalopathy is more related to a higher rate of lacunae within
the subcortical white matter rather than in the basal ganglia and
brainstem. The higher incidence of epilepsy in this group can be
explained by diaschisis resulting in suppression of the inhibitory
cortical activity (42). Positron emission tomographic examination
Chapter 19 ■ Cerebrovascular Diseases and EEG 359
of the brain has confirmed that patients with leukoaraiosis and
seizures have a more severe decline of regional cerebral blood
flow and metabolic rate of oxygen in the whole cortex compared
to those without seizures (43) and to those with late-onset symp-
tomatic seizures (44). Patients with extended chronic subcortical
infarctions can show generalized intermittent delta activity on
EEG recording (Fig. 19.8).
EEG IN CEREBRAL ISCHEMIC
DISEASE DUE TO EMBOLISM
All forms of cerebral emboli can cause a variety of clinical and
EEG changes. Already decades ago, Lhermitte and his coworkers
demonstrated that a surprisingly large number of strokes
believed to be thrombotic are due to emboli of cardiac origin
(45–47). Because of new imaging techniques and more insight
into pathophysiology of cerebral ischemia, it is well known that
about 40% of all cerebral ischemic events are embolic in origin.
Emboli can either arise from thrombotic material within the
larger brain supplying arteries (common carotid or internal
carotid arteries, vertebral arteries) or they can be cardiac in ori-
gin. Aortic arch plaque material may also cause cerebral ischemia.
Embolism of thrombotic material and gas and fat emboli may
give rise to clinically and electroencephalographically similar pic-
tures (Fig. 19.9A,B). In embolism, however, the extremely acute
shock-like effect may trigger a chain of vasomotor responses in
adjacent vascular territories. These responses either hasten the
catastrophic events or resolve quickly so that recovery may evolve
with unexpected speed.
Rasheva (48) and Rasheva et al. (49) noted clinical epileptic
seizures in 10.1% and paroxysmal EEG discharges (spikes and
sharp waves) in 57.1% of a large population of 328 patients
with embolic stroke.
The importance of cerebral gas embolism during and after
cardiac surgery is widely known. According to Naquet and
Vigouroux (50), gas emboli may produce EEG changes, similar
to those seen in watershed infarction, to be discussed in the fol-
lowing section. In fat embolism, Müller and Klingler (51)
reported marked delta–theta activity of frontotemporal accen-
tuation. Similar findings were reported by Scherzer (52).
Experimental cerebral embolism lends itself to the study of
immediate EEG effects. These are dramatic and instantaneous.
In the rabbit, Hegedüs et al. (53) demonstrated abrupt flatten-
ing of the affected hemisphere after embolization of the MCA;
there was even some voltage decrease contralaterally (Fig. 19.3).
This clearly shows that the effects of a major embolism extend
beyond the involved arterial territory. But in general, the EEG
cannot effectively assist in the differentiation of thrombotic and
embolic cerebral ischemia.
WATERSHED ISCHEMIA
Ischemic states usually pertain to the territory of the involved
artery, but another mechanism has become clinically impor-
tant. This is the concept of “extraterritorial ischemia” that gives
rise to “watershed infarctions” (“boundary infarctions”) along
the borders of the major arterial territories, between middle,
360 Part III ■ Clinical EEG: General

Figure 19.8 A 79-year-old female patient presented with a first primary focal, secondary generalized epileptic seizure to
the emergency unit. MRI revealed extended chronic subcortical infarctions without evidence of an acute infarction. EEG
showed marked left polymorphic delta activity and generalized intermittent rhythmic delta activity (IRDA).

posterior, and anterior cerebral arteries. This mechanism is
based on a decline in the systemic circulation. This is a drop in
blood pressure (BP) or cardiac output, usually combined with
a diseased and plaque-ridden cerebral vasculature. It is a logical
and convincing concept when it is assumed that boundary areas
between the major cerebral arterial territories are most likely to
suffer from ischemia. The work of Zülch and Behrend (54)
clearly demonstrates the distinction between the watershed
type and the classical obstructive ischemic infarction.
Acute watershed-type ischemia is most commonly found in
elderly patients with a history of cerebral arteriosclerosis and
chronic cardiovascular problems. A variety of illnesses may trig-
ger the acute cerebral ischemia. Acute infection, dehydration,
uncontrolled diabetes mellitus, cardiovascular decompensation,
severe anemia, recent craniocerebral trauma, extracerebral
embolism, liver disease with portocaval encephalopathy, and
severe arterial hypertension are events that can facilitate the
occurrence of watershed ischemia (50). Clinically, a contralateral
hemiparetic deficit is usually present, but it may not be as pro-
nounced as in primary capsular ischemia or hemorrhage.
The EEG is slow with a mixed delta–theta range frequency
and severely disorganized. These changes are essentially diffuse
with consistent lateralization to the affected hemisphere. Against
this slow and disorganized background of rather moderate volt-
age, there are consistently repetitive large sharp waves or spikes,
often compounded and consisting of several subcomponents
and phases, usually over the posterotemporal–occipital–parietal
region of the hemisphere that harbors the lesion. They are called
PLEDS. This periodic spike discharge is rhythmically or quasi-
rhythmically repetitive at a rate of approximately 1 per second.
The discharge may become quite widespread, even reaching into
the other hemisphere. Occasionally, the periodic spikes show a
superior frontal maximum.
At this stage of repetitive spike activity, contralateral focal
motor twitching (fingers or hand, face, abdomen, leg, foot, or
toes) is the common accompaniment of the periodic spikes, but
such PLEDs often occur without accompanying motor activity.
Further work on these discharges was presented by Alajouanine
et al. (55), Barolin et al. (56), Chatrian et al. (25), Markand and
Daly (57), and Pohlmann-Eden et al. (29,30). These latter authors
feel that PLEDs indicate focal hyperexcitability in the penumbra
zone of an ischemic stroke. Multifocal epileptic seizures of partial
character may also evolve in watershed ischemia, along with
independent EEG foci (58).
Figure 19.9 A 22-year-old male. Skiing accident with closed tibial and fibular fracture. Twenty-four hours later, the patient
complained about total loss of vision. MRI showed multifocal tiny high intensity areas on diffusion-weighted images in con-
cordance with the diagnosis of fat embolism. A: EEG shows generalized delta activity. B: 4 days later, a posterior alpha rhythm
in the 8 to 10 per second range has been restored; 3 to 6 per second theta–delta activity centro-parieto-temporo-occipital, more
marked on the right side.
362 Part III ■ Clinical EEG: General
EEG IN TRANSIENT ISCHEMIC ATTACKS
Transient ischemic attacks (TIAs) are defined as a temporary,
clinically completely reversible event of focal, nonconvulsive,
neurologic dysfunction secondary to a failure of sufficient
blood supply to a distinct area of the brain, retina, or other
nervous tissue. Most TIAs last less than 15 minutes, but the old
purely clinical standard definition of TIA was that symptoms
and signs have to be resolved within 24 hours (59,60).
According to this definition DWI demonstrated acute perma-
nent ischemic lesions in approximately 30% of TIA patients,
climbing to 50% when DWI is combined with perfusion-
weighted imaging (61). A new proposed definition by the TIA
Working Group is that TIA is a brief episode of neurologic dys-
function caused by focal brain or retinal ischemia with clinical
symptoms typically lasting less than 1 hour, most often some
minutes, without evidence of acute infarction on modern neu-
roimaging (61,62).
In some patients, TIA presents with the symptomatology
of limb shaking with brief, arrhythmic flailing or jerking
movements of the arm or leg, which can be easily misdiag-
nosed as a seizure disorder. But early diagnosis of a cerebrovas-
cular disorder is important for immediate initiation of
appropriate secondary stroke prevention including medical
and interventional or surgical procedures. EEG testing is often
ordered in a patient with limb shaking TIA because of suspi-
cion of seizures. Taghavy and Hamer (60) reviewed the records
of patients with TIAs over a period of 6 years. In 79 patients
with clinical TIA, an EEG was performed, in 44.3% demon-
strating focal abnormal electrical activity after recovery from
TIA corresponding to the affected hemisphere. In only 17.7%
of these patients, there was also an appropriate hypodensity in
cranial CT scan. Failure to consider cerebral hypoperfusion as
a potential cause of focal delta EEG slowing in the presence of
normal structural imaging studies may contribute to delay in
correct diagnosis.
The above mentioned new TIA definition makes it even
more difficult to distinguish TIAs from seizures, especially from
inhibitory seizures in elderly patients (63). Recognition of
modern diagnostic clues can prompt expedited magnetic reso-
nance angiography (MRA), carotid duplex ultrasound, or
computed tomography angiography (CTA) studies of the neck
and intracranial blood vessels to identify arterial stenoses or
occlusions (64).
De Reuck and Van Maele (63) performed a study on TIAs and
inhibitory seizures in elderly patients and included 25 patients
with inhibitory seizures and 252 patients with TIA over a 7-year
study period taking into account the old and new TIA definition.
The clinical characteristics and the EEG findings were compared.
Temporary speech disturbance, associated with some partial
amnesia for the event, was the most common clinical presenta-
tion of an inhibitory seizure. Additionally, specific and nonspe-
cific postictal EEG abnormalities were observed in the majority
of inhibitory seizure patients, whereas the EEG was normal in
more than 90% of TIA patients. Diffuse slowing and intermittent
rhythmic delta activities were only observed in patients with
seizures, not in TIA patients. They concluded that it should
always be checked whether the patient is able to recall the episode
of focal disturbance completely, as more than two thirds of
patients with an inhibitory seizure will have difficulties to
describe their temporary deficit clearly.
Early EEG should always be performed in patients with a
transient episode of focal neurologic disturbance, as abnormal-
ities can indicate the possibility of a seizure, whereas in TIA a
normal EEG should be expected (63–66).
EEG IN INTRACRANIAL VENOUS
THROMBOSIS
In the era of antibiotic therapy, episodes of infectious intracra-
nial venous thrombosis and thrombophlebitis have become
rare. Complicated ear infections (transverse sinus) or facial car-
buncles (cavernous sinus) occasionally cause such a progression
of infectious material into the venous system of the cranial cav-
ity. Complications of abortion are another cause; widespread
cortical venous thrombosis may develop intrapartum. Reviews
of the field have been presented by Garcin and Pestel (67),
Dubois (68), Huhn (69,70), and Carter (71).
EEG changes in cortical venous thrombosis intrapartum can
be very pronounced, with widespread prominent delta activity;
these patients are not necessarily in a state of impaired con-
sciousness but show marked changes of higher cortical func-
tions, such as aphasia, mutism, or apraxia, and the course of
ensuing recovery may be very slow.
Literature on the EEG in cerebrovenous pathology is scanty.
The work of Huhn (69), Lemmi and Little (72), Franco (73),
and Hubach and Struck (74) has been particularly informative.
Van der Drift and Kok (9) demonstrated the underlying patho-
physiologic mechanisms as well as the correlations of EEG and
venous pathology. Focal slowing and ictal epileptic activity have
been reported (75).
ROLE OF EEG IN PROGNOSIS
OF EPILEPTIC SEIZURES AFTER
ISCHEMIC STROKE
The incidence rates for epilepsies have two peaks regarding age:
one in the first decade of life and a second at higher age with a
steep increase between age 70 and 80 (76). Cerebral ischemia is
the single most common cause of first seizures and epilepsy in
later life above the age of 60 years, ahead of degenerative disor-
ders, brain tumors and head trauma, accounting for up to one
third of newly diagnosed seizures in this population. The lead-
ing type of stroke-related seizures is focal with a high rate of
secondary generalization (77–86).
The frequency of seizures after stroke varies from 5% to
approximately 40% (81,84–91), but only a minority will develop
epilepsy. A number of risk factors such as cortical involvement,
infarct size, and stroke severity have been identified as predictors
of seizure development (92). Large cortical infarctions in the
anterior circulation have a high risk of developing seizures (93).
Status epilepticus can be a presenting symptom of acute stroke
and can lead to increased mortality.

Based on differences in their presumed pathophysiology,
seizures after stroke are usually divided into early- and late-onset
seizures, in a paradigm comparable to posttraumatic epilepsy.
The division in early- and late-onset postischemic seizures is
important because the expected good prognosis of early seizures
(77,84,94–96) could not be approved by other authors
(85,97,98). Usually, early seizures are defined as those occur-
ring within 7 days of the stroke; late-onset seizures are usually
defined as occurring after day 7 within the first year after the
stroke. Results of recent studies revealed early seizures as inde-
pendent risk factor for late seizures and development of epilepsy
(78,99,100).
There are no clear guidelines for the treatment of seizures in
stroke and hence the treatment needs to be initiated in the con-
text of the individual patient. The presence of comorbid condi-
tions and the use of other drugs also complicate antiepileptic
therapy, and the risk of drug interactions is a particular hazard
in elderly patients on multiple comedications. The optimal tim-
ing and type of anticonvulsive treatment remain debated (91).
Several findings suggest that the majority of first-generation
anticonvulsive drugs are not the best choice in stroke patients.
Despite the importance of the problem, there are few data on
the natural history of stroke-related seizures and no good guide-
posts to suggest when to initiate anticonvulsant therapy after
stroke. The exact statistical risk of further seizures after a first
poststroke seizure is not known; therefore, it has to be a case-
by-case decision when to start medication. After a single early
seizure, long-term anticonvulsant therapy is usually not recom-
mended even though recent studies could not reproduce the pre-
viously thought good prognosis of early seizures. There is also
still debate about treatment after a first late seizure. After a sec-
ond seizure long-term anticonvulsant therapy is usually recom-
mended.
Poststroke seizures are in most patients well controlled with
a single anticonvulsant drug. The choice of drug is given by the
general recommendations of anticonvulsant medication in
patients with focal seizures. In these mainly elderly patients,
interactions with other medications and cognitive side effects
have to be considered especially. Whether the new anticonvul-
sants have advantages as compared with the standard medica-
tion with carbamazepine, oxcarbazepine, or valproic acid is still
open and under investigation.
The presence of seizures after stroke constitutes a challenge for
patients, families, and physicians. Limited information is available
regarding the risk factors of developing poststroke seizures and
the impact of seizures on the clinical outcome including mortal-
ity, disability, quality of life, and cognition. No guidelines exist for
the treatment of poststroke seizures, particularly the point of time
at which treatment should be started, and it is not clear whether
treatment influences course of illness and/or long-term prognosis
of stroke patients, quality of life, and development of mood
changes. New therapeutic approaches and guidelines can only be
developed when the mechanisms and risk factors of poststroke
seizures are better described and understood.
Several, mostly retrospective, studies dealing with this topic
have methodologic drawbacks. Most of them have small case
numbers and mix up ischemic and hemorrhagic strokes. The
Chapter 19 ■ Cerebrovascular Diseases and EEG 363
definition of early- and late-onset seizures according to the
guidelines of the “International League against Epilepsy” is often
neglected. Infarct etiology is not taken into account in most of
the studies because of the lack of duplex sonography and
echocardiography. The quality of neuroimaging techniques with
CT (used in most of the older studies) is not powerful enough to
detect clearly infarct morphology. Some studies describing the
predictive value of the EEG have been published in the past. As
an example, Carrera et al. (99) used a continuous EEG detection
to describe the incidence of epileptic electric activity, which has
no predictive power regarding the risk for poststroke seizures.
This method is not practicable in the acute phase of stroke. Some
authors point out that the occurrence of PLEDs has a predictive
value regarding development of poststroke seizures, but this is
discussed controversially (29,30,81,88).
EEG AND BRAIN MAPPING TECHNIQUES
FOR EVALUATING POSTSTROKE RECOVERY
AND REHABILITATION
Brain mapping techniques have proven to be vital in under-
standing the molecular, cellular, and functional mechanisms of
recovery after stroke (101). Magnetoencephalography (MEG)
and EEG techniques work in a completely different manner than
the image of blood flow–based brain mapping methods. MEG
and EEG detect signals that arise predominantly from the den-
dritic fields of cortical pyramidal neurons. They detect these sig-
nals from a large area on the order of several square centimeters.
With EEG, the tissue between the scalp electrode and the cortex
attenuates the strength of the signal and the signal strength
decreases with the square of the distance between them. Because
tissues do not affect the MEG signal, MEG typically enables
better signal source localization than EEG, but the magnetic
fluctuations drop off as the cube of the distance between the
source and the detector increases. EEG and MEG techniques
may use an event-based approach to identify the characteristics
of brain activity in response to specific behaviors. Strens et al.
(102) found that the degree of signal synchrony was greater
between medial and lateral motor areas in stroke patients than
in healthy subjects but reduced with recovery. In another study,
the degree of reduction in delta-wave signals from perilesional
brain tissue correlated with the amount of language recovery
following rehabilitation (103). The drawback of the EEG
method clearly is the lack of anatomic information.
A number of recent studies report that quantitative elec-
troencephalography (QEEG) measures of cerebral pathophysi-
ology in acute or subacute stroke might augment future
prognoses regarding patient outcomes (104–107). Finnigan et
al. (105) reported that QEEG delta power measures in acute
stroke and delta/alpha power ratio measures in subacute stroke
are both highly correlated with ischemic stroke patients’ out-
comes assessed via the National Institute of Health Stroke Scale
(NIHSS). QEEG measures were averaged over all scalp elec-
trodes and further a significant correlation between subacute
delta/alpha power ratio and outcome measures was obtained
when the former was computed from a standard, 19-channel
364 Part III ■ Clinical EEG: General
array. Moreover, multinomial logistic regression analyses
revealed that this QEEG per se enhanced the sensitivity and
specificity of outcome prediction beyond that afforded by sub-
acute lesion volume defined via perfusion-weighted MRI. These
outcomes are consistent with QEEG and observations by other
authors acquired from acute ischemic stroke patients, including
those who had deceased in the ensuing days. There seems to be
a prognostic value of poststroke shift of scalp delta power max-
ima from the ipsilateral to the contralateral hemisphere indicat-
ing substantial worsening of cerebral pathophysiology (107). A
standard 19-electrode array appears adequate for the detection
of this marker, and routine usage of a standard electrode assay
with fewer electrodes renders bedside EEG monitoring more
feasible than does a high-density array. The potential limita-
tions of such EEG monitoring are possible contamination of
QEEG indices by artifacts such as those due to patient move-
ment and/or by sleep- or drowsiness-related EEG activity.
QEEG measures, computed promptly, perhaps automatically,
might help guide future clinical therapeutic decisions. In
patients with large MCA infarctions, where decompressive sur-
gery is being contemplated, such QEEG observations, in con-
cert with other investigations like MRI, may assist in more
timely decisions to operate.
Recovery from stroke can vary greatly among patients even
though they may have identical clinical symptoms. Several
attempts have been made to determine neuroplastic changes on
the basis of electroencephalograms, which are readily available
in clinical settings. In many studies, movement-related poten-
tials (MRPs) were studied in connection with stroke rehabilita-
tion. MRPs constitute a waveform that is obtained by taking the
“average” EEG, in terms of onset of movement or myoelectrical
activity of the prime mover (108,109). Many studies demon-
strated topographical alterations of some MRP components
during the recovery period after a stroke (110–112). The analy-
sis of MRP, however, presents methodologic difficulties because
of its low signal to noise ratio. Adequate recording of MRP
requires careful repetition of many trials in which the elec-
tromyographic envelope of the investigated prime mover is
similar, and subjects need to be trained to perform similar
movements. These practical issues led Eder and colleagues
(113) to consider more stable EEG phenomena as criteria to
mark stroke recovery. Some prior findings indicate that event-
related desynchronization (ERD) might achieve this purpose
(114). The ERD of the central mu-rhythm was studied exten-
sively in normal subjects (115). Subsequent normative studies
have outlined that the symmetry of the mu-rhythm amplitude
had, in comparison with ERD, a greater discriminative power
between normal subjects and stroke patients (116). Results of
more recent studies point to the existence of two relatively inde-
pendent ERD generating systems within the central region of
each hemisphere with beta rebound following movement, also
referred to as postmovement beta event–related synchroniza-
tion (PMBS) (117). These beta oscillations are dominant on the
contralateral side of movement and are found in the first sec-
ond after termination of movement. Several studies showed
that lateralization of PMBS depended on factors such as hand-
edness, age, and pathologic conditions, for example, Parkinson
disease (118,119). Eder et al. (113) introduced a paradigm for
detecting and mapping PMBS for loosely defined movements
of the upper extremities on a drawing board, which was tested
in eight acute stroke patients with mild hemiparesis and eight
normal subjects. A follow-up testing was conducted 3 months
after the initial recordings. Their results confirmed repro-
ducibility of PMBS patterns in normal subjects since the repli-
cation of the recording has only a negligible effect on the
lateralization. The side of hemiparesis in acute stroke patients
could be distinguished on the basis of quantitative measure of
lateralization. The follow-up testing in three recovered stroke
patients revealed a trend of changes in the lateralization toward
the contralateral side of movement, an indication that this tech-
nique is an option for evaluation of therapy-induced changes in
cortical movement organization (113).
EEG IN POSTSTROKE HYPERSOMNIA
The incidence of stroke is about 300 to 600 per 100,000 per year
(Sudlow and Warlow, 1997). The frequency of hypersomnia
after stroke is essentially unknown but may be as high as 20%
to 40% (120,121). In a consecutive series of 100 patients with
ischemic stroke, 22 patients reported at 1 month after stroke
excessive daytime sleepiness and/or increase in sleep needs as
compared with prestroke situation (122). Patients with and
without poststroke hypersomnia were similar in age, gender,
Epworth sleepiness score, and estimated sleep need hours
before stroke. Patients with poststroke hypersomnia had a more
severe stroke and worse short-term outcome. The less well-
defined symptom “fatigue” may be found even in a higher per-
centage of poststroke patients (123).
The semiologic spectrum of poststroke hypersomnia is wide
and varies according to stroke topography (Bassetti, 2005; 122).
In deep (subcortical) hemispheric and particularly paramedian
thalamic lesions, hypersomnia may correspond to a so-called
presleep behavior, during which patients yawn, stretch, close their
eyes, curl up, and assume a normal sleeping posture, while com-
plaining of a constant sleep urge. Some of these patients are able
to control this behavior when stimulated or given explicit, active
tasks to perform. In some patients with deep frontal, thalamic, or
midbrain lesions, hypersomnia evolves to extreme apathy with
lack of spontaneity and initiative, poverty of movement, a condi-
tion for which the term “akinetic mutism” was coined. A contin-
uum exists between hypersomnia, athymormia, akinetic mutism,
and fatigue. In some patients, episodes of hypersomnia, mutism,
and akinesia alternate with episodes of insomnia, psychomotor
agitation, or confusional state (120–122).
The most severe and persistent forms of hypersomnia are
observed in association with lesions in the paramedian thala-
mus, midbrain, or upper pons. The paramedian thalamic stroke
syndrome represents the most striking form of hypersomnia
following focal brain damage. Hypersomnia can also be seen
with unilateral lesions but rarely persists, in these cases, for
more than a few months.
Less commonly hypersomnia may complicate lesions in the
caudate, striatum, lower pons, medial medulla, and cerebral
hemispheres (with or without mass effect) (121,122). The degree

and duration of hypersomnia over time in these patients are less
severe than in patients with paramedian thalamic stroke.
Hypersomnia following focal brain damage is usually con-
sidered to reflect a decreased arousal secondary to the dysfunc-
tion of the wakefulness maintaining neuronal systems
(ascending reticular activating system, ARAS). The occurrence
of hypersomnia following cortical or striatal strokes without
mass effect supports the assumption of a role of these struc-
tures in the maintenance of arousal and more generally in
sleep–wake regulation (124).
The most striking example of increased sleep is seen in
patients with bilateral paramedian thalamic stroke. Arpa et al.
(125) reported a 44-year-old man with right lateral-tegmental
pontine hematoma and severe hypersomnia, in whom long-
term EEG monitoring showed increased amounts of sleep rang-
ing from 11 to 15 hours per day during the first 3 months after
stroke. The relative amounts of slow wave sleep (4% to 11% of
total sleep time) and rapid eye movement (REM) sleep (8% to
10%) were slightly above normal values. Bastiju et al. (126)
described a patient with severe hypersomnia due to bilateral
thalamomesencephalic stroke with an initial sleep behavior
over 18 hours per day. Eight months after stroke, hypersomnia
had regressed clinically to about 12 hours per day. By EEG cri-
teria, sleep was similarly present over about 12 hours per day,
with an increase in both slow wave sleep (30% of total sleep)
and REM sleep (22%).
SLEEP EEG AFTER HEMISPHERIC STROKE
Changes in sleep EEG patterns after hemispheric stroke have
been described in studies for the last three decades (127–129).
Müller et al. (128) found a lower sleep efficacy, as well as lower
amounts of slow wave sleep and REM sleep in patients with
acute hemispheric stroke. Gottselig et al. (129) documented a
reduced spindle activity on the side of the lesion in patients with
extrathalamic stroke. Sleep efficiency, preserved spindle activity,
and amount of REM sleep in the acute phase of stroke have been
shown to be associated with a favorable outcome (130,131).
Siccoli et al. (2008) investigated the hypothesis of a link between
sleep and cognitive functions, particularly memory and atten-
tion, after stroke in 11 patients with first-ever hemispheric
ischemic stroke within 8 days after symptoms onset and in 9 of
them at least 3 months after stroke. In the acute stroke phase, a
correlation between attention and amounts of slow wave sleep,
REM sleep, and sleep efficiency was found.
HYPERTENSIVE ENCEPHALOPATHY
In this condition, severe arterial hypertension gives rise to cere-
bral edematous changes and hence to signs of intracranial
hypertension. Headache may reach an excruciating degree,
especially in the morning, augmented by coughing and strain-
ing and sometimes accompanied by vomiting. Transient attacks
of blindness may occur, and generalized as well as focal epilep-
tic seizures may materialize. Papilledema with retinal arteriolar
spasms and hemorrhages is common. The condition may
progress into somnolence, delirium, stupor, and coma (132).
Chapter 19 ■ Cerebrovascular Diseases and EEG 365
The EEG may show surprisingly little changes in this serious
cerebral disorder. There is reason to presume that the neuronal
oxygenation does not reach critically low values in spite of the
fact that the cerebral blood flow is diminished in this condition
(133). In the case of epileptic convulsions, EEG changes are
usually limited to the period immediately before, during, and
after the attacks. This is akin to the acute encephalopathy of
eclampsia gravidarum, where EEG abnormalities are barely
detectable in the preeclamptic state and are practically limited
to the grand mal convulsions themselves.
However, there have been reports of very pronounced EEG
abnormalities in hypertensive encephalopathy. Aguglia et al.
(134) described severe parieto-occipital changes with promi-
nent paroxysmal features. After acute onset of hypertensive
encephalopathy (BP of 300/140 and age 70), Benna et al. (135)
observed generalized spikes and polyspikes followed by a slow
wave. With gradual improvement of hypertension, the paroxys-
mal pattern lingered on. Figure 19.10 shows marked diffuse
slowing in a patient with hypertensive encephalopathy.
EEG IN INTRACEREBRAL BLEEDING
This form of stroke occurs mainly in hypertensive individuals;
high BP in combination with vascular disease (hyalinosis) may
lead to a massive rupture. Occasionally, intracerebral hemor-
rhages are caused by aneurysms, arteriovenous malformations,
or alteration of vessels inside a cerebral neoplasm. The most fre-
quently involved vessel is the lenticulostriate artery, which is a
small deep branch of the MCA. Its nearness to the internal cap-
sule explains the devastating effects on the motor outflow.
It is likely that subclinical seizures are common after
intracranial hemorrhage. Subclinical seizure activity has gained
attention from an increased understanding of the potential
harmful effects of clinical seizure activity (136). Investigations
in the most extreme case of seizure activity, status epilepticus,
demonstrate that prolonged seizure activity can cause neuronal
death. Animal experiments of convulsive status epilepticus and
some models of nonconvulsive status epilepticus show that
prolonged seizure activity causes immediate necrotic cell death
and delayed apoptotic cell death, primarily from activation of
NMDA-mediated glutamatergic excitotoxic pathways. It is rea-
sonable to consider that briefer periods of seizure activity cause
cell death and this is true in some animal models of limbic
seizures. However, there are few data demonstrating that dis-
crete seizures cause cell death in humans so caution should be
exercised in directly extrapolating this to subclinical seizures.
The utility of prolonged EEG monitoring in the intensive care
unit has been investigated for various conditions. A decrease in
relative alpha power may predict vasospasm and delayed cerebral
ischemia after subarachnoid hemorrhage (137,138).
Claassen et al. (138) performed a retrospective study on 102
consecutive patients with intracerebral hemorrhage who under-
went continuous electroencephalographic monitoring.
Convulsive seizures occurred in 19%, another 13% of patients had
subclinical, purely electroencephalographic seizures, and 5% had
both electroencephalographic and clinical seizures. They found an
increased risk of seizures if the hemorrhage reached the cortex or
Figure 19.10 A 71-year-old female with acute loss of consciousness. CT scan revealed a right-sided thalamic hemorrhage.
A: EEG showed right hemispheric rhythmic bi- and triphasic waves also showing on the left, but not synchronously.
B: After 4 days only slight improvement of EEG changes.

if it expanded by 30% in the first 24 hours. Periodic epileptiform
discharges were more frequently seen in hemorrhages closer to the
cortex. Midline shift and poor outcome were not more frequent in
patients with seizures when compared with those without
seizures. By contrast, periodic discharges were independently
associated with poor outcome. There was a trend toward worse
outcome for patients with electroencephalographic seizures, but
seizures did not remain a predictor in multivariate analysis.
Among those who had seizures during monitoring, 94% of
seizures were detected within the first 72 hours. This implies that
routinely providing continuous EEG monitoring for 3 days would
frequently detect subclinical seizures. But the retrospective nature
of this study limited it to those 13% of all intracranial hemorrhage
admissions for whom continuous EEG was requested (136).
There may have been a selection bias toward those who had some
reason to suspect seizure activity. In another prospective study,
seizures occurred in 28% of patients after intracerebral hemor-
rhage, suggesting that they are frequent (139). The true incidence
of subclinical seizures after intracerebral hemorrhage awaits a
prospective consecutive case series of continuous EEG monitor-
ing. It could be clinically important to find subclinical seizures if
they increase mortality or cause neuronal injury. But, Claassen et
al. (138) have not found seizures to be associated with poor out-
come. Obvious seizures do not present a problem; they are usually
treated with anticonvulsants. It seems reasonable to treat subclin-
ical seizures due to intracranial hemorrhage using routine meth-
ods, that is, oral anticonvulsants, but caution should be exercised
in treating aggressively, such as with continuous intravenous
propofol or midazolam. Aggressive therapies can cause severe
iatrogenic complications, whereas damage from seizures is likely
to be mild. This is unlike status epilepticus where the damage is
significant and risk of mortality is high. It is likely that subclinical
seizures are common after intracranial hemorrhage but paroxys-
mal EEG activity suspicious for seizures has unknown clinical sig-
nificance. Prolonged EEG monitoring is expensive and presents
logistical problems. At present, the significance of subclinical
seizures after intracerebral hemorrhage and the need of suppress-
ing them is not known (136).
THALAMIC HEMORRHAGE
Thalamic bleeding usually causes temporary loss of conscious-
ness. Larger hemorrhages may give rise to ipsilateral delta activ-
ity (Fig. 19.10A,B). According to Jasper and Van Buren (140),
there is a reduction of alpha rhythm in the case of anteroventral
thalamic damage. Alpha enhancement may be noted in patients
with posterior thalamic lesions. Lack of sleep spindles has been
reported (141). Delayed P3 responses were reported by Onofrj
et al. (142).
MIDBRAIN HEMORRHAGE
Mesencephalic bleeding without extension into the pons is
most likely to give rise to Parinaud syndrome; it is seldom fatal.
The EEG often shows diffuse activity in the upper theta range
(7 per second, according to Ref. 9).
Chapter 19 ■ Cerebrovascular Diseases and EEG 367
LOWER BRAINSTEM HEMORRHAGE
These uncommon strokes almost always give rise to coma and
respiratory anomalies (33), sometimes progressing to acute
respiratory failure (143). The pupils are usually very small.
Quadriplegia is very common; bouts of decerebrate posturing
also occur. The EEG may show the striking finding of a well-
preserved posterior alpha rhythm that cannot be blocked by
various modalities of stimuli (144–146), but in exceptional
cases, the reactivity of the alpha rhythm may persist (147).
Other patients show diffuse low-voltage tracings as seen in
cases of lower brainstem infarction (see under “Basilar Artery
Thrombosis”).
CEREBELLAR HEMORRHAGE
Acute cerebellar hemorrhage rapidly leads to coma after a few
minutes of headaches, dizziness, or vertigo. Cerebellar edema
rapidly produces a pressure cone (tonsillar herniation) with
fatal outcome. The rapidity of the evolution usually precludes
EEG studies. A remarkable study of 22 cases was carried out by
Rasheva et al. (49); it demonstrated high-voltage delta activity,
more over the contralateral cerebral hemisphere. In patients
with cerebellar stroke, movement-related cortical potentials
were found to be depressed (148).
SUBARACHNOID HEMORRHAGE
(INTRACRANIAL ANEURYSMS,
ARTERIOVENOUS MALFORMATIONS,
AND OTHER CAUSES)
The extremely dramatic symptomatology of a subarachnoid
hemorrhage with acute meningeal signs does not need special
discussion. Cranial nerve deficits are quite common. In more
severe cases, there is evidence of cerebral involvement with
impairment of consciousness and, in about 10% to 20% of the
cases, epileptic seizures.
The EEG shows diffuse changes, disorganization, disruption
of the posterior alpha rhythm, and excessive slow activity.
Occasionally, lateralization of slow activity may be indicative of
the primarily involved hemisphere (149–151). With total
dependence on arteriographic demonstration of the causative
lesion and partial dependence on refined CT scan methods,
there is no practical need for an EEG in the search for the cor-
rect localization. The EEG, however, remains a valuable indica-
tor of the general state of cerebral functioning (152,153).
The patient’s general condition seriously deteriorates when
subarachnoid bleeding is followed by vasospasm. The patho-
genesis of such widespread vasospasm is poorly understood
(154–156). According to Kiloh et al. (157), marked delta foci
occur in areas of massive vasospasm. Such secondary lesions are
sometimes associated with sharp activity (9). Prolonged
vasospasm and a decrease of cerebral blood flow may lead to
dementia and diffuse cerebral atrophy. The area of massive
vasospasm correlates with the area of local EEG slowing (158).
368 Part III ■ Clinical EEG: General
Ruptured bifurcation aneurysms (circle of Willis and vicin-
ity) are the most common cause of subarachnoid hemor-
rhage. Margerison et al. (159) carried out extensive EEG
studies in 70 patients with ruptured aneurysms in a search for
valid localizing signs. Massive bleeding from a ruptured
aneurysm into the cerebral parenchyma is prognostically omi-
nous and can be fatal within less than 1 hour. The EEG pat-
tern of intracerebral hematomas resulting from such bleeding
was studied by Van der Drift (8). Regional polymorphic delta
activity was found to be more pronounced than in acute MCA
ischemia.
In survivors of subarachnoid hemorrhages from aneurysms,
epileptic seizures have been found in 12.5% (160); marked EEG
changes with spikes are also found (157). Scott and Cabral
(161) stressed the frequent occurrence of epileptic seizures fol-
lowing intracranial aneurysm surgery (clipping, wrapping).
Survivors of subarachnoid bleeding from an aneurysm of
the anterior communicating artery sometimes develop severe
personality changes and dementia. Their state of apathy and
loss of drive is not associated with typical EEG changes. Severe
frontal and hypothalamic (deep perforating branches of ante-
rior cerebral artery) lesions might account for this lamentable
clinical picture.
Arteriovenous malformations are the cause of subarachnoid
hemorrhages that are much less severe and life threatening than
is bleeding from a ruptured aneurysm. These malformations
may also give rise to focal or generalized epileptic seizures that
usually start between ages 10 and 35. The EEG of patients with
arteriovenous malformations has been described as abnormal
in all observed cases (162). The interpretation of the abnormal-
ities is beset with difficulties, because a sizable number of cases
may show an “erroneous” lateralization with slowing or sharp
activity over the “wrong” hemisphere (163–166). As impractical
as such findings may be, their pathophysiologic substratum is
not devoid of clinical interest, since the misleading lateraliza-
tion is determined by intracerebral steal phenomena. In the
writer’s personal experience, even a very large arteriovenous
malformation may be found without any EEG abnormality.
This evidently indicates the lack of consistent hypoxic impair-
ment of the cortical function.
Subarachnoid bleeding may occur without demonstrable
cause. In some cases, the source of bleeding consists of aneurys-
matic widening of a bacterially infected artery (arteritis leading
to a “mycotic” aneurysm). Bacterial endocarditis is usually the
cause of this type of vascular pathology. During treatment with
endovascular embolization, EEG monitoring may detect clini-
cal hazards (167).
CEREBROVASCULAR INVOLVEMENT
IN CONNECTIVE TISSUE DISEASES
Systemic lupus erythematosus gives rise to neurologic signs and
symptoms in 25% to 35% of the patients (168). This is due to
small patchy vessel involvement. Convulsions and psychotic
episodes are the most common CNS changes. About 2% to 5% of
the patients with lupus erythematosus develop cerebrovascular
syndromes with occlusion and focal cerebral deficit (168). The
EEG in CNS complication of lupus erythematosus is deter-
mined by the severity of clinical neurologic deficits and may
range from normal to markedly abnormal findings of either
focal or diffuse character.
EEG studies of systemic lupus erythematosus and other
collagen or connective tissue diseases such as periarteritis
nodosa, scleroderma, giant cell arteritis, or rheumatoid arthri-
tis have been scanty (169–171; see also Ref. 172). According to
Finn et al. (173), the EEG is useful in the early detection of
this disease. Kogeorgos and Scott (174) studied 74 eases; EEG
abnormalities were present in 79%. Gottwald and Sturm
(175) reported abnormal records in 77% of 40 patients. The
EEG abnormalities consist of focal or diffuse slowing, spikes
or sharp waves of focal or multifocal character, and “dysrhyth-
mic phenomena.”
In Wegener’s granulomatosis, CNS involvement is less com-
mon than in lupus erythematosus. EEG abnormalities may
occur; metabolic (renal) causes have to be ruled out (176). In
morphea (localized scleroderma), epileptic seizures are fairly
common (25%), and both epileptic and nonepileptic EEG
abnormalities are three times as common as in age-matched
control population (177).
In Sjögren syndrome, the CNS may be affected by an aseptic
meningoencephalitis (178). Cognitive disturbances and
dementia may also be caused by vasculitis that is treatable with
corticosteroids in high dosages (179). These authors also found
diffuse EEG abnormalities in a patient with vasculitis.
OTHER TYPES OF VASCULAR DISEASE
In thromboangiitis obliterans cerebri, temporal focal delta activ-
ity may be found over areas of recent cortical damage, but these
changes disappear quickly; the EEG is probably of little diag-
nostic value in this disease (180). EEG findings are also
regarded as noncontributory in patients with moyamoya disease
(181). According to Aoki et al. (11), a reappearance of the
hyperventilation-induced buildup of slow activity approxi-
mately 20 to 60 seconds after the activation is a typical finding
in moyamoya disease. This phenomenon has been studied more
extensively by Ohyama et al. (182). Granulomatous angiitis
(183,184) is a progressive and fatal disease. Prominent diffuse
or local slowing of the EEG is a common finding.
In sickle cell anemia, neurologic complications may occur
occasionally. Impairment of consciousness, convulsions, and
neurologic deficits may prevail for several days or weeks.
Marked EEG changes are noted (172,185). Focal slowing was
found to be more impressive than the corresponding neuro-
logic symptomatology, with considerable attenuation of the
EEG changes during sleep (186).
Neurologic and electroencephalographic changes due to
blood diseases will be discussed under the heading of metabolic
encephalopathies and brain tumors in neurooncology. Briefly,
intracerebral hematomas are not uncommon in hemophilia,
thrombocytopenia, and leukemia. In such cases, marked focal
or diffuse slowing is found in the EEG.

CEREBRAL MANIFESTATIONS
OF CARDIAC DISEASES
Basic Considerations
Cardiac diseases and dysfunctions can impair the brain in vari-
ous ways with a wide variety of pathologic changes. In other
words, only a part of the CNS lesions and disturbances caused
by cardiac problems falls into the category of cerebrovascular
disorder. Cerebrovascular phenomena such as arterial
embolism are inseparably associated with bacterial infections
such as cerebritis and brain abscess. In other cases, one is deal-
ing with hypoxic–anoxic cerebral pathology, especially in com-
plications of open heart surgery.
Congenital Heart Disease
Neurologic disturbances are quite frequent in patients with
congenital heart disease. According to Tyler and Clark (187),
neurologic deficits occur in 25%, blackout spells in 12%, and
convulsions in 6% of the cases. It is therefore no surprise that
abnormal EEG tracings are fairly common in congenital heart
disease (188–194). Brain abscess is a well-known complication
of congenital heart disease, especially in infancy and early child-
hood. Focal polymorphic delta activity is usually very promi-
nent in these abscesses. This type of brain abscess is
unassociated with endocarditis and dissemination of infected
emboli. Diffuse cerebral hypoxia may give rise to impairment of
consciousness and marked diffuse slowing in the EEG. Focal
slowing may also occur. Such hypoxic lesion may gradually
become epileptogenic (194,195).
Cerebral hypoxia may develop in cyanotic as well as in non-
cyanotic forms of congenital heart disease, but it is more fre-
quently found in the former.
Acquired Heart Disease
CNS complications of acute or subacute bacterial endocarditis
are essentially infectious diseases and will be discussed under
that heading. In cardiac decompensation, a variety of causes and
mechanisms eventually leads to cerebral hypoxia with slowing
and terminal flattening of the EEG (172,196,197). EEG abnor-
malities are common in chronic bronchopneumopathy with cor
pulmonale. These changes may (198), or may not parallel the
degree of impaired consciousness (199).
Paroxysmal tachycardia may be associated with transient
EEG abnormalities (200,201), especially when the attack leads
to impairment of consciousness. Walsh et al. (202) stressed
paroxysmal cerebral discharges in conjunction with paroxysmal
atrial tachycardia, which was considered secondary to the cere-
bral events.
The Adams–Stokes syndrome is usually associated with a per-
manent heart block and a very slow pulse rate. The electroen-
cephalographer is advised to use EEG monitoring throughout
the recording. During the attack, slow activity appears after a
few seconds of asystolia with progressive amplitude of delta
waves (203) (see also Chapter 31).
Chapter 19 ■ Cerebrovascular Diseases and EEG 369
EEG in Syncope
Clinically, it can be difficult to differentiate syncope from an
epileptic seizure. Syncope is defined as transient loss of con-
sciousness and of postural control. Pathophysiologically, it is
based on global cerebral hypoxia as a result of transient hypoper-
fusion of the brain. Patients with syncope usually fall to the
ground with the onset of unconsciousness. In experimental stud-
ies, the frequency of tonic or myoclonic muscle activity during
syncope ranges between 12% and 100% with most studies find-
ing the frequency between 70% and 90%. Thus, convulsive syn-
cope is the rule rather than the exception, which can be explained
by the fact that convulsions are an intrinsic component of the
brain’s response to hypoxia. The most important criterion to dif-
ferentiate syncopal and epileptic convulsions is their specific phe-
nomenology. In contrast to epileptic muscle activity, syncopal
myoclonic movements are arrhythmic. Mostly, they are multi-
form rather than generalized; that is, they occur asynchronously
in different regions of the body. Only in rare cases do they last
longer than 30 seconds. Tonic phenomena are significantly less
frequent in syncope and are usually less pronounced. A tonic
extension of the body with flexion of the arms is more common
in profound cerebral hypoxia, for example, due to prolonged
asystole, or in syncope in children, such as breath-holding spells.
For the differentiation of syncope and epileptic seizures, post-
ictal phenomena are of particular importance. After syncope,
patients are fully reorientated within very few seconds, and even
after prolonged attacks lasting 1 to 2 minutes, the postictal con-
fusion or drowsiness lasts rarely longer than 30 seconds. Any
longer-lasting disorientation suggests an epileptic seizure. But it
must be kept in mind that partial seizures of frontal lobe origin
are frequently associated with a fast reorientation.
In rare cases, there is an interaction between syncopal and
epileptic mechanisms within a single attack. On the one hand,
this means that an epileptic seizure can emerge from syncope.
On the other hand, in certain forms of epileptic seizures, espe-
cially in complex partial seizures of temporal lobe origin, car-
diac arrhythmias can arise, which can then lead to syncope. If
the medical history suggests epileptic and syncopal phenomena
within the same attack, the diagnosis can only be made by
means of ictal EEG and ECG recordings.
The EEG during syncope shows generalized high-amplitude
slow-wave activity at the onset of unconsciousness, followed by
flattening of the EEG, and subsequently slow waves again before
the normal background activity returns (Brenner, 1997). This
sequence is independent of the mechanism of syncope and of
the clinical presentation as convulsive or nonconvulsive syn-
cope as it represents the common final path in terms of global
cerebral hypoxia.
The significance of EEG in the diagnosis of syncope is often
overestimated. It could be shown that in unselected patients with
syncope, the recording of a postictal EEG was not helpful.
Epileptiform patterns in an interictal EEG recording can indeed
corroborate a diagnosis of epilepsy. However, additional syncopal
attacks are not excluded. On the other hand, even in chronic
epilepsy epileptiform patterns can be absent during interictal EEG
recordings, thus also preventing a definite classification of the
370 Part III ■ Clinical EEG: General
attacks. Also, episodic seizures induced by alcohol or benzodi-
azepine withdrawal are usually not associated with epileptiform
patterns in the interictal EEG. Therefore, EEG recordings are not
recommended as standard tests alter syncope and should be
reserved for patients with positive history for epileptic seizures.
REFERENCES
1. Keilson MJ, Miller AE, Drexler ED. Early EEG and CTT scanning
in stroke—a comparative study [abstract]. Electroencephalogr Clin
Neurophysiol. 1985;61:21P.
2. Gloor P, Ball G, Schaul N. Brain lesions that produce delta waves
in the EEG. Neurology (Minneapolis). 1977;27:326–333.
3. Gastaut JL, Michel B, Sabet Hassan S, et al. Electroencephalography
in brain edema (127 cases of brain tumor investigated by cranial
computerized tomography). Electroencephalogr Clin Neurophysiol.
1979;46:239–255.
4. Chen J, Simon R. Ischemic tolerance in the brain. Neurology.
1997;48:306–311.
5. Hachinski V, Norris JW. The deteriorating stroke.. In: The Acute
Stroke. Philadelphia, PA: FA Davis (Quoted by Hachinski and
Norris, 1985).
6. Velho-Groneberg P. The EEG: an important (and often under-
rated) tool in the diagnosis of stroke. Am J EEG Technol. 1986;
26:213–224.
7. Kok NKD, Van der Drift JHA. The significance of the EEG for local-
ization and prognosis of supratentorial ischemic disease. A clinical
pathological study. Preliminary report. Rev EEG Neurophysiol. 1971;
1:295–307.
8. Van der Drift JHA. The EEG in cerebrovascular disease. In: Vinken
PJ, Bruyn GW, eds. Handbook of Clinical Neurology. Vol 11.
Amsterdam: North-Holland; 1972:267–291.
9. Van der Drift JHA, Kok NKD. The EEG in cerebrovascular disorders
in relations to pathology. In: A. Remond, Ed-in-chief. Handbook
of Electroencephalography and Clinical Neurophysiology. Vol 14A.
Amsterdam: Elsevier; 1972:12–30, 47–64.
10. Hilker BL, Dohmen C, Bosche B, et al. Early electroencephalogra-
phy in acute ischemic stroke: prediction of a malignant course?
Clin Neurol Neurosurg. 2007;109:45–49.
11. Aoki Y, Hiraga H, Ichijo S. EEG of moyamoya disease [abstract].
Electroencephalogr Clin Neurophysiol. 1977;43:490.
12. Kraaier V, Van Huffelen AC, Wieneke GH. Changes in quantitative
EEG and blood flow velocity due to standardized hyperventilation. A
model for transient ischemia in normal subjects. Electroencephalogr
Clin Neurophysiol. 1988;70:377–387.
13. Carney AL, Anderson EM, Burns E. Cerebral hemodynamic eval-
uation. In: Carney AL, Anderson EM, eds. Diagnosis and Treatment
of Brain Ischemia. New York, NY: Raven Press; 1981:335–359.
14. Anderson EM, Carney AL, Page L. EEG in the vascular laboratory.
In: Carney AL, Anderson EM, eds. Diagnosis and Treatment of
Brain Ischemia. New York, NY: Raven Press; 1981:325–334.
15. Hossmann KA. Viability thresholds and the penumbra of focal
ischemia. Ann Neurol. 1994;36:557–565.
16. Genvins A, Le J, Brickett P, et al. Seeing through the skull: advanced
EEG use MRIs to accurately measure cortical activity from the
scalp. Brain Topogr. 1991;4:125–131.
17. Jackel RA, Harner RN. Computed EEG topography in acute stroke.
Neurophysiol Clin. 1989;19:185–197.
18. Luu P, Tucker DM, Englander R, et al. Localizing acute stroke-
related EEG changes: assessing the effects of spatial undersampling.
J Clin Neurophysiol. 2001;18:302–317.
19. Fisher M, Garcia JH. Evolving stroke and the ischemic penumbra.
Neurology. 1996;47:884–888.
20. Kayser-Gatchalian MC, Neundörfer B. The prognostic value of
EEG in ischaemic cerebral insults. Electroencephalogr Clin
Neurophysiol. 1980;49:608–617.
21. Paddison RM, Ferris GS. The electroencephalograms in cerebral
vascular disease. Electroencephalogr Clin Neurophysiol. 1961;13:
99–110.
22. Van der Drift JHA, Magnus O. The EEG in cerebral ischemic lesions.
Correlations with clinical and pathological findings. In: Gastaut H,
Meyer JS, eds. Cerebral Annoxia and the Electroencephalogram.
Springfield. IL: Thomas; 1961:180–196.
23. Sainio K, Stenberg D, Kesimäki I, et al. Visual and spectral EEG
analysis in the evaluation of the outcome in patients with ischemic
brain infarction. Electroencephalogr Clin Neurophysiol. 1983;56:
117–124.
24. Burghaus L, Hilker R, Dohmen C, et al. Early electroencephalogra-
phy in acute ischemic stroke: prediction of a malignant course?
Clin Neurol Neurosurg. 2007;109:45–49.
25. Chatrian GE, Shaw GM, Leffman H. The significance of periodic
lateralized epileptiform discharges: an electrographic clinical and
pathological study. Electroencephalogr Clin Neurophysiol. 1964;17:
177–193.
26. Echlin FA, Arnett V, Zoll J. Paroxysmal high voltage discharges
from isolated and partially isolated human and animal cerebral
cortex. Electroencephalogr Clin Neurophysiol. 1952;4:147–164.
27. de la Paz D, Brenner RP. Bilateral independent periodic lateralized
epileptiform discharges. Arch Neurol. 1981;38:713–715.
28. Reiher J, Rivest J, Grand’Mason F. Periodic lateralized epileptiform
discharges with transitional rhythmic discharges. Electroencephalogr
Clin Neurophysiol. 1991;78:12–17.
29. Pohlmann-Eden B, Hoch DB, Cochius JL, et al. Periodic lateralized
epileptiform discharges: a critical review. J Clin Neurophysiol. 1996;
13:519–530.
30. Pohlmann-Eden B, Hoch OB, Cochius JI, et al. Stroke and
epilepsy: critical review of the literature. Part I: Epidemiology and
risk factors. Cerebrovasc Dis. 1996;6:332–338.
31. Fitzpatrick W, Lowry N. PLEDS: clinical correlates. Can J Neurol
Sci. 2007;34:443–450.
32. Friedlander WJ. Electroencephalographic changes in acute brain
stem vascular lesions. Neurology (Minneapolis). 1959;9:24–34.
33. Plum F, Posner JB. Diagnosis of Stupor and Coma. Philadelphia, PA:
Davis; 1966.
34. Hawkes CH, Bryan-Smyth L. The electroencephalogram in the
“locked-in” syndrome. Neurology. 1974;24:1015–1018.
35. Markand ON. Electroencephalogram in the “locked-in” syndrome.
Electroencephalography. 1976;40:529–534.
36. Chase TN, Moretti L, Prensky EL. Clinical and electroencephalo-
graphic manifestations of vascular lesions of the pons. Neurology
(Minneapolis). 1968;18:357–365.
37. Westmoreland BF, Klass DW, Sharbrough FW, et al. “Alpha coma”:
electroencephalographic, clinical, pathological and etiological cor-
relation. Arch Neurol. 1975;32:713–718.
38. Vignaendra V, Wilkus RJ, Copass MK, et al. Electroencephalographic
rhythms of alpha frequency in comatose patients after cardiopul-
monary arrest. Neurology. 1974;24:582–588.
39. Carroll WM, Mastaglia FL. Alpha and beta coma in drug intoxica-
tions uncomplicated by cerebral hypoxia. Electroencephalogr Clin
Neurophysiol. 1979;46:95–105.
40. Jouvet M. Coma and other disorders of consciousness. In: Vinken
PJ, Bruyn GW, eds. Handbook of Neurology. Vol 12. Amsterdam:
North-Holland; 1969:64–79.

41. Kenangil G, Orken DN, Yalcin D, et al. Triphasic EEG pattern in
bilateral paramedian thalamic infarction. Clin EEG Neurosci. 2008;
39:185–190.
42. Schreiner A, Pohlmann-Eden B, Schwartz A, et al. Epileptic
seizures in subcortical vascular encephalopathy. J Neurol Sci. 1995;
130:171–177.
43. De Reuck J, Decoo D, Boon P, et al. Late-onset epileptic seizures in
patients with leukoaraiosis: a positron emission tomographic
study. Eur Neurol. 1996;36:20–24.
44. De Reuck J, Santens P, Decoo D, et al. Positron emission tomo-
graphic study of late-onset cryptogenic symptomatic seizures. Clin
Neurol Neurosurg. 1995;97:208–212.
45. Lhermitte F, Gautier JC, Derouesne C. Anatomopathologie et
physiopathologie des stenoses carotidiennes. Rev Neurol (Paris).
1966;115:641.
46. Lhermitte F, Guiraud B. Ischemic accidents in the middle cerebral
artery. Arch Neurol. 1968;19:248.
47. Lhermitte F, Gautier JC. Sites of cerebral arterial occlusions. In:
Williams D, ed. Modern Trends in Neurology. London: Butterworths;
1975:123–140.
48. Rasheva M. Epileptic seizures in the acute stage of embolic stroke
[abstract]. Electroencephalogr Clin Neurophysiol. 1981;52:78P.
49. Rasheva M, Stamenov E, Todorova P, et al. Dynamics of electrical
activity of the brain in cerebellar haemorrhage [abstract].
Electroencephalogr Clin Neurophysiol. 1981;52:78P.
50. Naquet R, Vigouroux RA. EEG in cerebral gas emboli from cardiac
surgery. In: Remond A, Ed-in-chief. Handbook of Electroen -
cephalography and Clinical Neurophysiology. Vol 14A. Amsterdam:
Elsevier; 1972:38–44.
51. Müller HR, Klingler M. The electroencephalogram in cerebral fat
embolism. Electroencephalogr Clin Neurophysiol. 1965;18:278–386.
52. Scherzer E. EEG—Veränderungen bei zerebraler Fettembolie.
Psychiatr Neurol. 1967;153:337.
53. Hegedüs K, Fekete I, Tury F, et al. Experimental focal cerebral
ischaemia in rabbits. J Neurol (Berl). 1985;232:223–230.
54. Zülch KJ, Behrend RCH. The pathogenesis and topography of
anoxia and ischemia of the brain in man. In: Gastaut H, Meyer JS,
eds. Cerebral Anoxia and the Electroencephalogram. Springfield, IL:
Thomas; 1961:144–163.
55. Alajouanine T, Lecasble R, Remond A. Elements graphiques parox-
ystiques lents de survenue periodiques, Correlations electroclin-
iques. Rev Neurol (Paris). 1955;93:477–478.
56. Barolin C, Scherzer E, Naquet R, et al. Etude electroclinique des etats
de mal epileptiques survenant chez les apolectiques. Rev Neurol
(Paris). 1962;107:242–243.
57. Markand ON, Daly DD. Pseudoperiodic lateralized paroxysmal
discharges in the electroencephalogram. Neurology (Minneapolis).
1971;21:975–981.
58. Matsuo F. Multifocal epileptic seizures associated with acute cere-
bral ischemia [abstract]. Electroencephalogr Clin Neurophysiol. 1985;
61:25P.
59. Millikan CH. Transient cerebral ischemia: definition and natural
history. In: McDowell FH, Sonnenblick H, Lesch M, eds. Current
Concepts in Cerebrovascular Disease. New York, NY: Grune and
Stratton; 1980:1–6.
60. Taghavy A, Hamer H. Parenchymal “damage” in transient ischemic
attacks (TIAs) and prolonged reversible ischemic neurological
deficits (PRINDs): the role of cranial CT and EEG. Int J Neurosci.
1992;66:251–261.
61. Castle J, Mlynash M, Lee K, et al. 2010. Agreement regarding diag-
nosis of transient ischemic attack fairly low among stroke-trained
neurologists. Stroke. 2010;41(7):1367–1370.
Chapter 19 ■ Cerebrovascular Diseases and EEG 371
62. Albers GW, Caplan LR, Easton JD, et al. TIA Working Group: tran-
sient ischemic attack—proposal for a new definition. N Engl J
Med. 2002;347:1713–1716.
63. De Reuck J, Van Maele G. Transient ischemic attacks and inhibitory
seizures in elderly patients. Eur Neurol. 2009;62:344–348.
64. Bearden S, Uthman B. Cerebral hemodynamic compromise asso-
ciated with limb shaking TIA and focal EEG slowing. Am J
Electroneurodiagnostic Technol. 2009;49:225–243.
65. Enge S, Lechner H, Logar C, et al. Clinical value of EEG in transient
ischemic attacks. In: Lechner H, Aranibar A, eds. EEG and Clinical
Neurophysiology. Amsterdam: Excerpta Medica; 1980:173–180.
66. Silva A, Gross PT. Use of EEG in the evaluation of TIAs [abstract].
Electroencephalogr Clin Neurophysiol. 1983;56:13P.
67. Garcin R, Pestel M. Thrombophlebites Cerebrales. Paris: Masson;
1949.
68. Dubois J. Les thrombophlebites cerebrales post-partum. Gynaecol
Obstet. 1956;55:478.
69. Huhn A. Die Hirnvenen- und Sinusthrombose. Fortschr Neurol.
1957;25:440.
70. Huhn A. Die venöse Abflußstörung. In: Quandt J, ed. Die zere-
bralen/Durchblutungsstörungen des Erwachsenenalters. Berlin: VEB
Verlag Volk und Gesundheit; 1969:773–800.
71. Carter AB. Clinical aspects of infarction. In: Vinken PJ, Bruyn GW,
eds. Handbook of Neurology. Amsterdam: North-Holland; 1972:
292–326.
72. Lemmi H, Little SC. Occlusion of intracranial venous structures. A
consideration of the clinical and electroencephalographic find-
ings. Arch Neurol. 1960;3:252–266.
73. Franco PJ. Electroencephalography in venous disturbances of the
brain. Rev Neuro-Psiquiatr. 1961;24:447–452.
74. Hubach H, Struck G. Zur Korrelation von EEG und patho-mor-
phologischen Befunden cerebraler Gefäßprozesse. Arch Psychiatr
Nervenkr. 1965;206:641.
75. Griewing B, Mertins L, Lütcke A, et al. Das Elektroenzephalogramm
in der Verlaufsbeobachtung und Diagnostik von Thrombosen des
Hirnnervensystems. EEG-Labor. 1991;13:103–117.
76. Hauser W, Annegers J, Kurland L. Incidence of epilepsy and unpro-
voked seizures in Rochester, Minnesota: 1935–1984. Epilepsia. 1993;
34:453–468.
77. Berges S, Moulin T, Berger E, et al. Seizures and epilepsy following
strokes: recurrence factors. Eur Neurol. 2000;43:3–8.
78. Ferro JM, Pinto F. Poststroke epilepsy: epidemiology, pathophysi-
ology and management. Drugs Aging. 2004;21:639–653.
79. Kilpatrick CJ, Davis SM, Tress BM, et al. Epileptic seizures in acute
stroke. Arch Neurol. 1990;47:157–160.
80. Lamy C, Domigo V, Semah F, et al. Early and late seizures after
cryptogenic ischemic stroke in young adults. Neurology. 2003;60:
400–404.
81. Lossius MI, Rønning OM, Slapø GD, et al. Poststroke epilepsy:
occurrence and predictors—a long-term prospective controlled
study (Akershus Stroke Study). Epilepsia. 2005;46:1246–1251.
82. Niedermeyer E. Über Epilepsie im höheren Lebensalter. Arch
Psychiatr Nervenkr. 1958;197:248–262.
83. Shinton RA, Gill JS, Melnick SC, et al. The frequency, characteristics
and prognosis of epileptic seizures at the onset of stroke. J Neurol
Neurosurg Psychiatr. 1988;51:273–276.
84. Bladin C, Alexandrov A, Bellavance A, et al. Seizures after stroke: a
prospective multicenter study. Arch Neurol. 2000;57:1617–1622.
85. Burn J, Dennis M, Bamford J, et al. Epileptic seizures after first
stroke. Br Med J. 1997;315:1582–1587.
86. Gupta S, Naheedy M, Elias D, et al. Post-infarction seizures: a clin-
ical study. Stroke. 1988;19:1477–1481.
372 Part III ■ Clinical EEG: General
87. Holmes GL. The electroencephalogram as a predictor of seizures
following cerebral infarction. Clin Electroencephalogr. 1980;11:
834.
88. Kammersgaard LP, Olsen T. Poststroke epilepsy in the
Copenhagen stroke study: incidence and predictors. J Stroke
Cerebrovasc Dis. 2005;14:210–214.
89. Kelly KM. Poststroke seizures and epilepsy: clinical studies and
animal models. Epilepsy Curr. 2002;2:173–177.
90. Lancman ME, Golimstok A, Norscini J, et al. Risk factors for
developing seizures after a stroke. Epilepsia. 1993;34:141–143.
91. Tettenborn B. Häufigkeit und Behandlung von zerebralen
Anfällen und Epilepsie nach Schlaganfall. Arch Neurol Psychiatr.
2004;155:169–172.
92. Tettenborn B, Vitek L, Grabow T, et al. Carotid stenosis is a risk
factor for recurrent seizures after cerebral ischemia. Epilepsia.
1999;40(suppl 7):98.
93. De Reuck J, De Groote L, Van Maele G, et al. The cortical involve-
ment of territorial infarcts as a risk factor for stroke-related
seizures. Cerebrovasc Dis. 2008;25:100–106.
94. De Reuck J, Goethals M, Claeys I, et al. EEG findings after a cere-
bral territorial infarct in patients who develop early- and late-
onset seizures. Eur Neurol. 2006;55:209–213.
95. De Reuck J, Sieben A, Van Maele G. Characteristics and outcomes
of patients with seizures according to the time of onset in rela-
tion to stroke. Eur Neurol. 2008;59:225–228.
96. Kotila M, Waltimo O. Epilepsy after stroke. Epilepsia. 1992;33:
495–498.
97. De Reuck J, Goethals M, Vonck K, et al. Clinical predictors of
late-onset seizures and epilepsy in patients with cerebrovascular
disease. Eur Neurol. 2005;54:68–72.
98. Jacoby A, Snape D, Baker GA. Determinants of quality of life in
patients with epilepsy. Neurol Clin. 2009;27:843–863.
99. Carrera E, Michel P, Despland PA, et al. Continuous assessment
of electrical epileptic activity in acute stroke. Neurology. 2006;67:
99–104.
100. Kilpatrick CJ, Davis SM, Hopper JL, et al. Early seizures after
stroke. Risk of late seizures. Arch Neurol. 1992;49:509–511.
101. Eliassen JC, Boespflug EL, Lamy M, et al. Brain-mapping tech-
niques for evaluating poststroke recovery and rehabilitation: a
review. Top Stroke Rehabil. 2008;15:427–450.
102. Strens LH, Asselman P, Pogosyan A, et al. Corticocortical cou-
pling in chronic stroke: its relevance to recovery. Neurology. 2004;
63:475–484.
103. Meinzer M, Elbert T, Wienbruch C, et al. Intensive language
training enhances brain plasticity in chronic aphasia. BMC Biol.
2004;2:20.
104. Cuspineda E, Machado C, Aubert E, et al. Predicting outcome in
acute stroke: a comparison between QEEG and Canadian neuro-
logical scale. Clin Electroencephalogr. 2003;34:1–4.
105. Finnigan SP, Rose SE, Chalk JB. Rapid EEG changes indicate
reperfusion after tissue plasminogen activator injection in acute
ischaemic stroke. Clin Neurophysiol. 2006;117:2338–2339.
106. Finnigan SP, Rose SE, Walsh M, et al. Quantitative EEG indices of
sub-acute ischaemic stroke correlate with clinical outcomes. Clin
Neurophysiol. 2007;118:2525–2532.
107. Finnigan SP, Rose SE, Chalk JB. Contralateral hemisphere delta
EEG in acute stroke patients precedes worsening of symptoms
and death. Clin Neurophysiol. 2008;119:1689–1695.
108. Deecke L, Scheid P, Kornhuber HH. Distribution of readiness
potential, pre-motion positivity, and motor potential of the
human cerebral cortex preceding voluntary finger movements.
Exp Brain Res. 1969;7:158–168.
109. Gilden L, Vaughan HG, Costa LD. Summated human EEG poten-
tials with voluntary movement. Electroencephalogr Clin
Neurophysiol. 1966;20:433–438.
110. Green JB, Bialy Y, Sora E, et al. High-resolution EEG in poststroke
hemiparesis can identify ipsilateral generators during motor
tasks. Stroke. 1999;30:2659–2665.
111. Platz T, Kim IH, Pintschovius H, et al. Multimodal EEG analysis
in man suggests impairment-specific changes in movement-
related electric brain activity after stroke. Brain. 2000;123:
2475–2490.
112. Verleger R, Adam S, Rose M, et al. Control of hand movements
after striatocapsular stroke: high-resolution temporal analysis of
the function of ipsilateral activation. Clin Neurophysiol.
2003;114:1468–1476.
113. Eder CF, Skoic D, Covickovic-Sternic N, et al. Symmetry of post-
movement beta-ERS and motor recovery from stroke: a low-
resolution EEG pilot study. Eur J Neurol. 2006;13:1312–1323.
114. Pfurtscheller G, Aranibar A. Event-related desynchronization
detected by power measurements of scalp EEG. Electroencephalogr
Clin Neurophysiol. 1977;42:817–826.
115. Pfurtscheller G, Aranibar A. Changes in central EEG activity in
relation to voluntary movement. I. Normal subjects. Prog Brain
Res. 1980;54:225–231.
116. Kopruner V, Pfurtscheller G. Multiparametric asymmetry score
(MAS)—distinction between normal and ischaemic brains.
Electroencephalogr Clin Neurophysiol. 1984;57:343–346.
117. Andrew C, Pfurtscheller G. Lack of bilateral coherence of post-
movement central beta oscillations in the human electroen-
cephalogram. Neurosci Lett. 1999;273:89–92.
118. Stancak A Jr, Pfurtscheller G. Effects of handedness on move-
ment-related changes of central beta rhythms. J Clin Neurophysiol.
1997;14:419–428.
119. Pfurtscheller G, Pichler-Zalaudek K, Ortmayr B, et al.
Postmovement beta synchronization in patients with Parkinson’s
disease. J Clin Neurophysiol. 1998;15:243–250.
120. Bassetti CL. Sleep and stroke. In: Kryger MH, Roth T, Dement
WC, eds. Principles and Practice of Sleep Medicine. 4th ed.
Philadelphia, PA: Elsevier Saunders; 2005:811–830.
121. Bassetti CL. Sleep and stroke. Semin Neurol. 2005;25:19–32.
122. Bassetti CL, Valko P. Poststroke hypersomnia. Sleep Med Clin.
2006;1:139–155.
123. Carlsson GE, Moller A, Blomstrand C. Consequences of mild
stroke in persons 75 years—a 1 year follow-up. Cerebrovasc Dis.
2003;16:383–388.
124. Braun AR, Balkin TJ, Wesensten NJ, et al. Regional cerebral blood
flow throughout the sleep–wake cycle. Brain. 1997;120:1173–1197.
125. Arpa J, Rodriguez-Albarino R, Izal E, et al. Hypersomnia after
segmenta pontine hematoma: case report. Neurologia. 1995;10:
140–144.
126. Bastiju H, Nighoghossian N, Salord F, et al. Mesodiencephalic
infarct with hypersomnia: sleep recording in two cases. J Sleep
Res. 1994;3(suppl 1):16.
127. Hachinski VC, Mamelak M, Norris JW. Clinical recovery and sleep
architecture degradation. Can J Neurol Sci. 1990;17:332–335.
128. Müller C, Ackermann P, Bischof M, et al. Visual and spectral
analysis of sleep EEG in acute hemispheric stroke. Eur Neurol.
2002;48:164–171.
129. Gottselig JM, Bassetti CL, Ackermann P. Power and coherence of
sleep spindle frequency activity following hemispheric stroke.
Brain. 2002;125:373–383.
130. Giubilei F, Iannilli M, Vitale A. Sleep patterns in acute ischemic
stroke. Acta Neurol Scand. 1992;86:567–571.

131. Vock J, Ackermann P, Bischof M. Evolution of sleep and sleep
EEG after hemispheric stroke. Brain. 2002;125:373–383.
132. Toole JF, Patel AN. Cerebrovascular Disorders. 2nd ed. New York,
NY: McGraw-Hill; 1974.
133. Heine G. Vergleichende Untersuchung über Hirndurchblutungs-
größe und Hirnstrombild beim Menschen. Verh Dtsch Ges
Kreislaufforsch. 1953;19:196–210.
134. Aguglia U, Tinuper P, Farnarier G, et al. 1984. Electroencephalo-
graphic and anatomo-clinical evidences of posterior cerebral
damage to hypertensive encephalopathy. Clin Electroencephalogr.
15:53–60.
135. Benna P, Bergamini L, Tarenzi L, et al. Hypertensive encephalopa-
thy: association with unusual EEG changes [abstract]. Electroen-
cephalogr Clin Neurophysiol. 1984;58:74P.
136. Fountain NB. Is it time for routine EEG monitoring after intrac-
erebral hemorrhage? Neurology. 2007;69:1312–1313.
137. Vespa PM, Nuwer MR, Juhasz C. Early detection of vasospasm
after acute subarachnoid hemorrhage using continuous EEG ICU
monitoring. Electroencephalogr Clin Neurophysiol. 1997;103:
607–615.
138. Claassen J, Jetté N, Chum F, et al. Electrographic seizures and
periodic discharges after intracerebral hemorrhage. Neurology.
2007;69:1356–1365.
139. Vespa PM, O’Phelan K, Shah M, et al. Acute seizures after intrac-
erebral hemorrhage. Neurology. 2003;60:1441–1446.
140. Jasper HH, Van Buren J. Interrelationship between cortex and
subcortical structures: clinical electroencephalographic studies.
Electroencephalogr Clin Neurophysiol. 1953;4:168–188.
141. Hirose G, Saeki M, Kosoegawa H, et al. Delta waves in the EEGs
of patients with intracerebral hemorrhage. Arch Neurol. 1981;
38:170–175.
142. Onofrj M, Curatola L, Malatesta P, et al. Delayed P3 event-related
potentials (ERPs) in thalamic hemorrhage. Electroencephalogr
Clin Neurophysiol. 1992;83:52–61.
143. Steegman AT. Primary pontile hemorrhage. J Nerv Ment Dis.
1951;114:35–65.
144. Loeb C, Poggio GF. Electroencephalogram in a case with pon-
tomesencephalic haemorrhage. Electroencephalogr Clin Neuro-
physiol. 1953;5:295–296.
145. Loeb C, Rosadini G, Poggio GF. Electroencephalograms during
coma. Neurology. 1959;9:610–618.
146. Neundörfer BL, Meyer-Wahl L, Meyer JG. Alpha-EEG und
Bewußtlosigkeit. Ein kasuistischer Beitrag zur lokaldiagnostis-
chen Bedeutung des Alpha-EEG beim bewußtlosen Patienten. Z
EEG-EMG. 1974;5:106–114.
147. Radermecker J. Severe acute necrosis of the pons with long survival:
electroclinical symptoms and absence of cerebral lesions [abstract].
Electroencephalogr Clin Neurophysiol. 1967;23:281–282.
148. Gerloff C, Altenmüller E, Dichgans J. Disintegration and reorgan-
ization of cortical motor processing in two patients with cerebel-
lar stroke. Electroencephalogr Clin Neurophysiol. 1996;98:59–68.
149. Roseman E, Bloom BM, Schmidt RP. The electroencephalo-
gram in intracranial aneurysms. Neurology (Minneapolis). 1951;1:
25–38.
150. Christian W. Klinische Elektroenzephalographie. 2nd ed. Stuttgart:
Thieme; 1975.
151. Daly DD, Markand ON. Focal brain lesions. In: Duly DD, Pedley
TA, eds. Current Practice of Clinical Electroencephalography.
2nd ed. New York, NY: Raven Press; 1990:335–370.
152. Logar C, Enge S, Körner EM, et al. The value of EEG in subarach-
noidal haemorrhagia. EEG EMG Z Elektroenzephalogr Elektromyogr
Verwandte Geb. 1982;13:68–72.
Chapter 19 ■ Cerebrovascular Diseases and EEG 373
153. Labar DR, Fisch BJ, Fink ME, et al. Quantitative EEG following
subarachnoid hemorrhage: a comparison of methods [abstract].
Electroencephalogr Clin Neurophysiol. 1989;72:2P–3P.
154. Echlin F. EEG changes following spasms of the basilar and verte-
bral arteries produced by experimental subarachnoid hemorrhage
[abstract]. Electroencephalogr Clin Neurophysiol. 1965;19:319.
155. Levitt P. The effects of subarachnoid blood on the electrocor-
ticogram of the cat. In: Proceedings of the American EEG Society,
Hans Berger Award Lecture; Washington, DC; September 1970.
156. Heros RC, Zervas NT, Varsos V. Cerebral vasospasm after sub-
arachnoid hemorrhage: an update. Ann Neurol. 1983;14:599–608.
157. Kiloh LG, McComas AJ, Osselton JW. Clinical Electroencephalo-
graphy. 3rd ed. London: Butterworths; 1972.
158. Rumpl E, Bauer G, Stampfei G. Der Angiospasmus bei der
Subarachnoidalblutung als wichtige Ursache für fokale Verän -
derungen im EEG Z. EEG-EMG. 1977;8:200–204.
159. Margerison JH, Binnie CD, McCaul LR. Electroencephalographic
signs employed in the location of ruptured intracranial arterial
aneurysms. Electroencephalogr Clin Neurophysiol. 1970;28:
296–306.
160. Walton JN. The electroencephalographic sequelae of sponta-
neous subarachnoid haemorrhage. Electroencephalogr Clin
Neurophysiol. 1953;5:41.
161. Scott DF, Cabral R. Development and prevention of seizures after
neurosurgical procedures including ruptured cerebral
aneurysms. In: Proceedings of the American Epilepsy Society. New
York; December 3, 1975.
162. Husby J, Norlen G, Petersen Z. Electroencephalographic find-
ings in intracranial arterial and arteriovenous aneurysms and
subarachnoid hemorrhages. Acta Psychiatr Scand. 1953;28:
387–400.
163. Groethuysen UC, Bickford RG, Svien HJ. The EEG in arteriove-
nous anomalies of the brain. Arch Neurol Psychiatry. 1955;74:
506–513.
164. Mosmans PCM, Jonkman EV. The significance of the collateral
vascular system of the brain in hunt and steal syndromes. Clin
Neurol Neurosurg. 1980;82:145–156.
165. Yeh HS, Kashiwagi S, Tew J, et al. Surgical management of epilepsy
with cerebral arteriovenous malformations. J Neurosurg. 1990;
72:216–223.
166. Walton A. Pre- and postoperative findings in an arteriovenous
aneurysm. A case study. Am J EEG Technol. 1994;34:55–65.
167. Paiva T, Campos J, Baeta E, et al. EEG monitoring during
endovascular embolization of cerebral arteriovenous malforma-
tions. Electroencephalogr Clin Neurophysiol. 1995;95:3–13.
168. Aita JA. Neurologic Manifestations of General Diseases.
Springfield, IL: Thomas; 1975.
169. Taylor RM, Pacella BC. Electroencephalograms in scleroderma.
J Nerv Ment Dis. 1949;107:42–47.
170. Krump JE. Modifications electroencephalographiques des mal-
adies du collagene. Rev Neurol (Paris). 1955;92:593–607.
171. Stein J, Stava Z. Electroencefalogram u sklerodermie. Czechoslov
Derm. 1961;36:501–512 (Quoted after Gibbs and Gibbs, 1964).
172. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 3. 2nd
ed. Reading, MA: Addison-Wesley; 1964.
173. Finn R, Kamaldeen S, Moona G, et al. The EEG in systemic lupus
erythematosus. Electroencephalogr Clin Neurophysiol. 1980;49:91P.
174. Kogeorgos J, Scott DF. Neuropsychiatric and EEG features in 74
cases of systemic lupus erythematosus with cerebral involvement
[abstract]. Electroencephalogr Clin Neurophysiol. 1982;53:1P.
175. Gottwald W, Sturm U. EEG findings in lupus erythematosus.
Electroencephalogr Clin Neurophysiol. 1984;57:34P–35P.
374 Part III ■ Clinical EEG: General
176. Molaie M, Fariello RG. EEG detection of cerebral involvement in
Wegener granulomatosis [abstract]. Electroencephalogr Clin
Neurophysiol. 1981;51:37P.
177. Hwang P, Andermann F, Metrakos K, et al. The seizures of
morphea [abstract]. Electroencephalogr Clin Neurophysiol. 1981;
52:47P.
178. Alexander EL, Alexander GE. Aseptic meningoencephalitis in
primary Sjögren syndrome. Neurology. 1983;33:593–598.
179. Caselli RJ, Scheithauer BW, Bowles CA, et al. The treatable
dementia of Sjögren syndrome. Ann Neurol. 1991;30:98–101.
180. Bernsmeier A, Held K. Thromboangiitis obliterans cerebri. In:
Vinken PJ, Bruyn GW, eds. Handbook of Neurology. Vol 11.
Amsterdam: North-Holland; 1972:292–326; Vol 12: 384–397.
181. Nishimoto A, Takeuchi S. Moyamoya disease. Abnormal arterial
networks in the cerebral basal region. In: Vinken PJ, Bruyn GW,
eds. Handbook of Neurology. Vol 12. Amsterdam: North-Holland;
1972:352–383.
182. Ohyama H, Niizuma H, Kinjo T, et al. Changes of EEG and tissue
pCO2 with hyperventilation in children with Moyamoya disease
[abstract]. Electroencephalogr Clin Neurophysiol. 1983;56:55P.
183. Cravioto H, Feigin I. Non-infectious granulomatous angiitis with
a predilection for the nervous system. Neurology (Minneapolis).
1959;9:599–609.
184. Kolodny EH, Reibeiz JJ, Caviness VS, et al. Granulomatous
angiitis of the central nervous system. Arch Neurol. 1968;19:
510–524.
185. Hill FS, Hughes JG, Davis BC. Electroencephalographic findings
in sickle cell anemia. Pediatrics. 1950;6:277–285.
186. Neidengard L, Niedermeyer E. The electroencephalogram in neuro-
logical complications of sickle cell anemia (SS-hemoglobinopathy).
Clin Electroencephalogr. 1975;6:68–74.
187. Tyler HR, Clark DB. Incidence of neurological complications in
congenital heart disease. Arch Neurol Psychiatry. 1957;77:17–22.
188. Losev JE. Kocenke funkcionajnogo softojanija kory golovnogo
mozga ditej a vrozdenymi serdca osnovanij elektroencefalo-
graficeskich issledovanij. Vop Okhrany Materin Dets. 1958;34:
66–75 (Quoted after Lesny et al., 1961).
189. Shev EE, Robinson SJ. Electroencephalographic findings associ-
ated with congenital heart disease. Preliminary report. Electroen-
cephalogr Clin Neurophysiol. 1958;10:253–258.
190. Shev EE, Robinson SJ. Electroencephalographic findings associ-
ated with congenital heart disease. In: Gastaut H, Meyer JS, eds.
Cerebral Anoxia and the Electroencephalogram. Springfield, IL:
Thomas; 1961:578–589.
191. Beaussart M, Dupuis D, Dupuis R. Electroencephalographic stud-
ies of cyanotic congenital heart disease. In: Gastaut H, Meyer JS,
eds. Cerebral Anoxia and the Electroencephalogram. Springfield, IL:
Thomas; 1961.
192. Lesny I, Bor I, Vlach V. EEG changes in children suffering from con-
genital heart disease: influence of O2 inhalation. Electroencephalogr
Clin Neurophysiol. 1961;13:173–179.
193. Bekeny G, Horänyi B, Peter A, et al. Elektroenzephalographische
Untersuchungen bei kongenitalen Herzfehlern. In: Jenenser EEG-
Symposion. 30 Jahre Elektroenzephalographie. Berlin: VER Verlag
Volk und Gesundheit; 1963:194–200.
194. Kalyanaraman K, Niedermeyer E, Rowe R, et al. The electroen-
cephalogram in congenital heart disease. Arch Neurol (Chicago).
1968;18:98–106.
195. Niedermeyer E. The EEG in cardiac diseases. In: Remond A, Ed.-
in-chief. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 14A. Amsterdam: Elsevier; 1972:65–67.
196. Obrist WD, Bissell LF. The electroencephalogram of aged
patients with cardiac and cerebral vascular disease. J Gerontol.
1955;10:315–330.
197. Krump JE. Die Lebenswandlung des Elektroenzephalogramms
bei Herz-und Kreislaufkranken. Verh Dtsch Ges Kreislaufforsch.
1958;24:258–265.
198. Davidson LAG, Jefferson JM. Electroencephalographic studies in
respiratory failure. Br Med J. 1959;2:296.
199. Goulon M, Picidalo JJ, Christophe M, et al. Clinical, electroen-
cephalographic and biological correlations in 34 cases of chronic
bronchopneumopathy with asphyxia. In: Gastaut H, Meyer JS,
eds. Cerebral Anoxia and the Electroencephalogram. Springfield,
IL: Thomas; 1961:565–577.
200. Subbotnik SI, Feinberg YS, Spielberg PJ. Electroencephalography
in paroxysmal tachycardia. Electroencephalogr Clin Neurophysiol.
1955;7:577–584.
201. Tucker JS, Yoe RH. Simultaneous EEG-EKG recording: study of a
case with complete heart block and ventricular tachycardia.
Electroencephalogr Clin Neurophysiol. 1956;8:129–132.
202. Walsh G, Masland W, Goldensohn ES. Paroxysmal cerebral dis-
charge associated with atrial tachycardia [abstract]. Electroen-
cephalogr Clin Neurophysiol. 1968;24:187.
203. Broser F. Die cerebralen vegetativen Anfalle. Berlin: Springer; 1958.

Dementia and EEG
INTRODUCTION
Only a few years after his first report on the human electroen-
cephalogram (EEG), Hans Berger described EEG abnormalities
in dementia, including an autopsy-proven case of Alzheimer
disease (AD) (1,2). Since then, over 2000 research papers and
more than 200 reviews have been published on the EEG in
dementia. According to DSM IV criteria, dementia is a syn-
drome characterized by multiple cognitive deficits that cause
impairment in social or occupational functioning and represent
a significant decline from a previous level of functioning (3).
Dementia has now become a major health problem, especially
in the aged Western population. The prevalence of dementia in
the Western population in those aged over 65 has been esti-
mated between 3.6% and 10.3% (4). In 50% to 60% of patients
dementia is due to AD (4). In Europe alone, 5 million people
are suffering from dementia (5).
In view of the importance of dementia as a health problem
and the large number of studies that have been conducted, one
might expect that the role of the EEG in dementia is now well
defined. Surprisingly, this turns out not to be the case. Despite
recent efforts at evidence-based guidelines for the diagnosis of
dementia, opinions on the usefulness of the EEG still vary con-
siderably. According to some authors, the usefulness of the EEG
is very limited and should be restricted to excluding nonconvul-
sive status epilepticus or supporting a diagnosis of sporadic
Creutzfeldt–Jakob disease (6). In an American guideline the
EEG is not even considered as an option (7). In contrast, other
authors have advocated a more widespread use of the EEG in
dementia (8,9). Summarizing the EEG literature up to 1996,
Jonkman reached the following three conclusions: (i) there
remains a need for a laboratory technique to confirm the diag-
nosis since a 100% accuracy cannot be obtained by clinical
methods only; (ii) the sensitivity of the EEG is higher or equal
to the best other laboratory techniques; (iii) all AD patients
should be studied at least once by EEG as well as by MRI.
In a recent European guideline, the possibility that the EEG
could be considered not simply as a tool to exclude rare condi-
tions but also as an adjunct in the diagnosis of specific causes of
dementia is considered (5). Compared to the 2001 USA guide-
line, the 2007 European guideline reflects the tendency to move
away from a purely exclusionary approach to dementia diagno-
sis that dominated the older literature (10). Currently, there is a
tendency to use laboratory investigations such as brain imaging,
cerebrovascular fluid markers, and the EEG not just to exclude
rare and possibly treatable causes of dementia but also to obtain
CHAPTER
CORNELIS JAN STAM 20
positive arguments in favor of specific diseases causing demen-
tia such as AD, frontotemporal lobar dementia (FTLD), vascular
dementia (VaD), and so on (11,12). In a recent review on the
usefulness of EEG and ERPs as biomarkers Jackson and Snyder
indicate that they anticipate expanded application of quantita-
tive EEG (qEEG) as a reliable and sensitive biomarker of mild
cognitive impairment (MCI) and AD (13).
In this chapter we will give an overview of the literature on
EEG in dementia. This overview cannot be complete, in view of
the huge number of published research papers; instead we will
present a selection of the most important studies, with an
emphasis on more recent investigations that make use of cur-
rent diagnostic criteria and proper statistical methods. First, the
most important EEG changes in normal aging will be discussed.
This topic is important to gain a proper understanding of the
borderline between normal changes due to aging and the earli-
est signs of brain pathology in MCI and early AD. EEG changes
in MCI, which is in many respects a transitional phase between
normal aging and early dementia, will be discussed separately.
Next, EEG changes in the major categories of dementia will be
described. Emphasis will be given to the most common causes
of dementia such as AD, VaD, dementia with Lewy bodies
(DLB), and FTLD. Rare causes of dementia that are particularly
relevant for EEG such as Creutzfeldt–Jakob disease will also be
described. The practical use of the EEG in the work up of
patients with cognitive complaints will be considered taking
into account the objective data from the literature as well as the
controversies between various guidelines.
EEG CHANGES IN NORMAL AGING
Compared to the extensive EEG changes that occur in the ear-
liest years of life, reflecting developmental processes in the
brain, the EEG remains relatively stable after the age of 18.
However, EEG changes do occur with normal aging, especially
after the age of 60, and it is important to delineate these from
the earliest signs of pathology. EEG changes with normal aging
have been described extensively in a classic review by Klass and
Brenner (14). A more recent review has been published by
Rossini et al. (15). One problem confronting EEG studies in the
elderly is the definition of “normal.” On the one hand, “normal”
can be defined in a statistical sense as referring to what is most
prevalent in a population. A consequence of this approach
could be that EEG findings will be biased by the presence of as
yet subclinical pathology in an elderly population. For example,
in a study on 52 normal subjects aged 20 to 91 years,
375
376 Part III ■ Clinical EEG: General

Hartikainen et al. found abnormal EEGs in 21% (16). Such
findings raise the question whether the EEG is very prone to
false-positive findings or whether a substantial percentage of
subjects were suffering from subclinical pathology.
Alternatively, “normal” can be defined as complete absence of
disease or “successful” aging. Taken to the extreme, this
approach might imply that the EEG of even the very old should
not be different from that of an 18 year old, what is, of course,
not realistic. In practice it is not so easy to determine which def-
inition of “normal” holds; it depends strongly on the inclusion
criteria used in EEG studies of healthy elderly. However, the
question what is normal remains important, and we will come
back to it when appropriate. Here we will discuss age-related
changes in the major EEG rhythms, as well as some borderline
phenomena such as focal temporal slow activity.
The physiologic alpha rhythm of awake, healthy adults is
characterized by a frequency range between 8 and 13 Hz, a pos-
terior temporal and occipital dominance and a clear reactivity to
eye opening. The characteristics of the alpha rhythm, and in par-
ticular the alpha peak frequency, are very stable properties
within a person and have a strong genetic basis. Several of these
characteristics have been reported to change with normal aging,
most notably the alpha peak frequency. A slowing of the alpha
frequency was described already in 1954 by Obrist (17). After the
age of 85 years the alpha rhythm may slow to 7 to 8 Hz (6).
However, even in the very old the alpha frequency is not
expected to drop below 7.5 to 8 Hz (18). In addition to slowing,
the alpha rhythm in the elderly may be characterized by the
presence of a benign rhythmic subharmonic variant of the alpha
rhythm (19). This variant is characterized by rhythmic activity
in the theta or delta band, at a frequency that is a subharmonic
of the alpha rhythm, with clear reactivity to eye opening. Often
traces of normal alpha can be found in the same record as well
(Fig. 20.1). Reports on alpha rhythm slowing are probably
biased at least to some extent by the presence of subclinical
pathology. In our view slowing of the alpha rhythm below 8 Hz,
even in the very old, is more likely to reflect pathology than nor-
mal aging. In a single subject, slowing of the alpha peak fre-
quency over time by more than 1 Hz is also a sign of pathology.
Several studies have addressed changes in the amount and
amplitude of the alpha rhythm with normal aging. A reduced
percentage time alpha and reduced alpha reactivity were
reported by Markand (20). A reduction of alpha band power was
also described by Hatikainen et al. (16). According to Babiloni et
al. the power decrease in healthy elderly subjects may be limited
to the lower alpha band of 8 to 10.5 Hz (21). The lower alpha
band has been associated with attentional processes, whereas the
upper alpha band may be related to different types of cognitive

Figure 20.1 Male, 55 years. Alternation of normal alpha rhythm on the left and subharmonic slow variant of alpha rhythm
on the right. This can be a normal finding in elderly subjects. Calibration mark: vertical, 50 V; Horizontal, 1 second. Filter
settings: time constant 1 second. Low-pass filter, 35 Hz. Source derivation.

processes, memory, and perceptual performance, but in different
ways and with different reactivities (22). In addition to a
decrease in power, the spatial distribution of EEG rhythms
becomes more uniform in the elderly (23). Even though the
power of the alpha rhythm may decrease with aging, it should be
stressed that this is a fairly nonspecific finding. The so-called
low-voltage EEG with amplitudes less than 20 V can be a nor-
mal variant (14). An exception is the low-voltage EEG in
Huntington disease, which will be discussed later in this chapter.
Compared to the alpha rhythm, much less is known about
changes in other rhythms with aging. The mu rhythm may be
less prevalent in elderly (24). Slowing of the mu rhythm has not
been the subject of separate studies but is probably in line with
alpha rhythm changes. Differences in peak frequency between
mu and alpha of more than 1 Hz should probably be considered
to be abnormal. Of interest, the alpha rhythm in AD can be
slowed with relative sparing of the mu rhythm. The amount of
beta decreases in the elderly (16). Surprisingly, several reports
indicate that elderly women have more beta than men (16,25).
This finding is suggestive of postmenopausal hormonal
changes, but there is no clear evidence for this conjecture.
Quantitative analysis has shown that absolute power in delta
and theta bands decreases with normal aging (16). In the eldest
group, delta power was related to subtle cognitive deficits and
cholinergic dysfunction.
A controversial issue is the interpretation of focal abnormal-
ities in the anterior temporal areas in the EEG of the elderly.
Such abnormalities have been reported by several authors
(26–28). These abnormalities are highly prevalent, occurring in
over one third of subjects over 60 years, often with a preference
for the left temporal areas (27). Intermittent temporal slowing
has been related to white matter hyperintensities on MRI (26).
White matter hyperintensities in healthy elderly have also been
associated with decreased EEG coherence and cognitive abnor-
malities in another study (29). In a study in 27 healthy elderly
subjects with a mean age of 78 years (range: 65 to 83 years)
Visser et al. describe temporal theta or delta in 16 subjects
(59%) (28). These temporal EEG abnormalities were associated
with impaired word fluency and ventricular widening on CT.
Klass and Brenner discuss the possible clinical significance of
these temporal slow waves extensively (14). They propose to
distinguish between more malignant and more benign forms of
temporal slow waves. They introduce the concept of “benign
temporal transients of the elderly” (BTTE) and give nine crite-
ria that have to be fulfilled for temporal slow waves in order to
be classified as BTTE. If the criteria are not fulfilled, the tempo-
ral slow waves should be considered pathologic and may indi-
cate temporal, possibly vascular pathology.
The introduction of the concept of BTTE is a laudable
attempt to reduce the number of false-positive findings in EEGs
of elderly which would be likely to occur if all focal temporal
slow waves would be classified as abnormal. However, the crite-
ria for BTTE are somewhat complex and elaborate, and no
information is available on their intra- and interrater reliability.
Also, it is not clear whether the association between focal tem-
poral theta or delta, cognitive dysfunction and structural
abnormalities does not hold for the supposedly benign category
Chapter 20 ■ Dementia and EEG 377
of BTTE. At present it cannot be excluded that focal temporal
theta and delta may be very sensitive early biomarkers of tem-
poral lobe dysfunction, either of vascular or degenerative ori-
gin. The significant association between temporal focal
abnormalities and MCI in a large cohort of memory clinic
patients would favor such a view (30). What would be helpful is
a simpler classification scheme of temporal abnormalities in the
elderly, allowing for different grades of abnormalities and not
just a classification as benign and malignant.
Another phenomenon that can be encountered infrequently
in the EEG of healthy elderly subjects is a type of rhythmic,
anteriorly dominant delta that occurs at the transition between
wakefulness and sleep. This activity resembles frontal intermit-
tent rhythmic delta activity (FIRDA) and has been referred to as
“sleep onset FIRDA” (31). However, it should be stressed that
FIRDA in the EEG of adults is always abnormal, whereas sleep
onset FIRDA is considered to be a benign variant. A proper
evaluation depends on the context, including the state of wake-
fulness as well as the presence of other EEG abnormalities.
Pathologic FIRDA is unlikely to occur in the context of an oth-
erwise entirely normal EEG. Also related to drowsiness is the
phenomenon of “subclinical rhythmic EEG discharge of adults”
(SREDA) (32). SREDA is characterized by paroxysms of rhyth-
mic, sometimes sharply formed activity, mainly in the theta
band. The paroxysms have a sudden onset and end, and the
activity does not show the evolution of amplitude or frequency
that is typical of epileptic seizures. The distribution is either
unilateral or bilateral, predominantly in the temporal areas.
SREDA is not associated with any kind of brain pathology and
is considered to be an entirely benign variant.
MILD COGNITIVE IMPAIRMENT
A major trend in current studies on diagnosis in dementia is the
emphasis on early diagnosis. Patients seek medical attention at
a stage when there are complaints, possibly objective cognitive
deficits, but not sufficient criteria for a diagnosis of dementia.
There is an urgent need for tools to be able to either reassure
these patients that their complaints are innocent or to antici-
pate the advent of full-blown dementia reliably and take appro-
priate actions. The introduction of the concept of MCI reflects
these current trends. MCI is a condition characterized by objec-
tive memory disturbances in the absence of other cognitive
deficits (33). Importantly, the objective cognitive deficits are
insufficient to interfere with social or occupational functioning.
The original concept of MCI was based on memory complaints
and could be called amnestic MCI. More recently, the concept
of MCI has been expanded to include nonamnestic as well as
multidomain types of MCI as well (12). In all cases, MCI is con-
sidered to be a transitional phase between normal aging and
dementia, characterized by complaints, objective deficits, but
no interference with daily functioning. MCI patients have a
clearly increased risk to develop dementia. The risk of develop-
ing dementia is 15% per year for a patient with MCI, compared
to 1% to 2% for the healthy elderly population (12). Although
MCI is clearly a risk factor for dementia, not all patients with
MCI become demented. One major challenge is to identify
378 Part III ■ Clinical EEG: General
those patients who will not go on to develop dementia at an
early stage. This is one of the important objectives of EEG stud-
ies in this field.
EEG changes in MCI subjects compared to healthy elderly
subjects are subtle but have been described in many studies.
Several authors have reported an increase in theta power in MCI
compared to healthy controls (34–37). Increased theta, but not
alpha or beta, power correlates with hippocampal volume loss
(34,38). Jiang et al. found increased alpha and beta band power
in addition to increased theta power in MCI (39). EEG power
was negatively correlated with MMSE scores. In the study of
Babiloni et al. the strength of delta and alpha 1 sources was
decreased in MCI subjects, and these changes also correlated
with MMSE changes (40). Surprisingly, lower alpha 1 source
power was negatively associated with vascular abnormalities on
MRI (41). Liedorp et al. reported a significant association
between focal abnormalities and MCI (30). In contrast, a com-
bination of focal and diffuse abnormalities showed a negative
association with MCI and rather favored a diagnosis of AD, VaD,
or DLB.
Some studies have attempted to increase the sensitivity by
recording the EEG during tasks or by considering measures of
functional connectivity. The notion of “functional connectiv-
ity” refers to statistical interdependencies between measures of
brain activity (such as EEG, MEG, and BOLD) recorded over
different brain regions; such correlation might reflect func-
tional interactions taking place between those brain regions
(42). Pijnenburg et al. showed that during a working memory
task the synchronization likelihood, a general measure of cor-
relations between signals, was significantly higher in MCI
compared to the control subjects in the lower alpha band (8 to
10 Hz) (43). The increased task-related connectivity might
suggest a compensatory mechanism. In another study involv-
ing working memory tasks, the EEG coherences in all bands
were significantly higher in the MCI group in comparison with
those in the control group (39). MCI patients showed less
alpha band reactivity during a picture memory task (44).
Findings with respect to resting state functional connectivity
seem to be less consistent. Some studies showed no difference
between two groups at rest (39,45). In contrast Babiloni did
find a reduced frontoparietal alpha 1 band synchronization
likelihood in AD and MCI subjects, correlating with lower
MMSE scores (46).
An important question is whether the EEG changes in MCI
predict conversion to dementia. Huang et al. found decreased
alpha global field power and more anterior located theta, alpha,
and beta sources in progressive as opposed to stable MCI (47).
In the study of Jelic et al. alpha and theta relative power and
mean frequency from left temporal occipital regions proved to
be reliable predictors for conversion to AD (48). Progressive
MCI patients displayed lower event-related theta synchroniza-
tion than stable MCI (49). In a study in 69 subjects fulfilling cri-
teria for MCI, of whom 24 converted to AD, Rossini et al.
showed increased midline coherence as well as stronger delta,
theta, and alpha 1 sources in converters compared to noncon-
verters (50). In contrast, Luckhaus et al. reported an association
between decreased alpha activity at baseline and risk of conver-
sion in MCI (51). In the study of van der Hiele et al. increased
theta activity during eyes closed and less alpha reactivity during
eyes open and memory activation at baseline indicated lower
global cognitive and executive performance at follow-up (52).
Prichep et al. showed that the EEG might even predict future
cognitive decline in subjects who have subjective memory com-
plaints but do not even fulfill criteria for MCI (53). Baseline
EEG in 44 normal elderly subjects with subjective complaints
could predict future cognitive decline during an impressive 7-
year follow-up. Decliners were characterized by increased theta
power, slowing of the mean frequency, and abnormal func-
tional connectivity especially in the right hemisphere in their
baseline EEGs.
To conclude, EEG changes can be demonstrated in MCI, and
some of these changes may correlate with structural changes
(notably hippocampal atrophy), cognitive deficits, and an
increased risk of conversion to dementia. The most consistent
finding seems to be an increase in theta power, in particular in
the anterior temporal areas. Studies during cognitive tasks or
employing measures of functional connectivity show less con-
sistent results, possibly due to unknown compensation effects.
Many authors claim that the EEG can predict conversion to
dementia in MCI, but there is no agreement on which EEG
measure is most important. In addition to increased theta
power, slowing of the alpha rhythm and decreased alpha power
may be important. So far, many of these studies are mainly con-
cerned with predictions at the group level. It remains to be
shown that reliable predictions can be made at the level of indi-
vidual patients in prospective studies.
ALZHEIMER DISEASE
Alzheimer disease was first described by Alois Alzheimer in 1907
(54). It is the most frequent cause of dementia in the aging
Western population, accounting for more than 50% to 60% per-
cent of all cases (4). Neuropathologically, AD is characterized by
neuronal loss, atrophy, deposition of amyloid beta 1–42 protein
in senile plaques, and neurofibrillary tangles composed of
abnormally phosphorylated tau protein. Different stages can be
distinguished in the neuropathology of AD, mainly depending
on the distribution of neurofibrillary tangles, which is limited to
the hippocampus in early stages and spreads out to entorhinal
and association cortex in later phases (55). Although the deposi-
tion of amyloid is often considered an important hallmark of the
disease (56), neurofibrillary tangles may be more closely related
to disease symptoms. With the advent of new PET tracers it has
now become possible to trace the neuropathology of AD in vivo
(57,58). 2-(1-(6-[(2-[18F]Fluoroethyl) (methyl)amino]-2-
naphthyl)ethylidene)malononitrile ([18F]FDDNP) is a tracer
that binds to neurofibrillary tangles and beta-amyloid senile
plaques. FDDNP binding has been shown to correlate with cog-
nitive disturbances in AD patients (58). In contrast, Pittsburgh
Compound-B (PIB) that binds selectively to amyloid did not
correlate with cognitive dysfunction in AD (57). While the diag-
nostic usefulness of these new PET tracers remains to be eluci-
dated, they present a major breakthrough in the direct
neuroimaging of the underlying pathology in AD.
 Chapter 20 ■ Dementia and EEG 379

Clinically, AD is characterized by early memory loss, in par-
ticular impaired encoding of new material. Later, cortical func-
tions become disrupted resulting in aphasia, apraxia, and
agnosia. Psychiatric and behavioral problems are also common.
Average duration of survival after diagnosis in AD is 5 to 8
years (59). A diagnosis of AD depends mainly on clinical find-
ings, with neuropathologic confirmation as a gold standard.
However, the interrater agreement of a diagnosis of AD is only
moderate with a kappa of 0.63 (60). Currently, the most
important research criteria for AD are the NINCDS-ADRDA
criteria (61). According to Waldemar et al. these criteria have a
sensitivity of 81% and a specificity of 70% across studies with
postmortem confirmation (5). Even the autopsy findings may
not constitute a straightforward gold standard: kappa scores
for neuropathologic interrater reliability vary between 0.65
and 0.59 depending on criteria used (62). The modest agree-
ment between clinical diagnostic criteria and neuropathologic
findings illustrates the need for sensitive, specific, and reliable
biomarkers for AD. However, they may also point to more fun-
damental limitations in our concepts of disease entities. We
will return to this in our discussion of the clinical usefulness of
the EEG.
EEG changes in AD were already described by Hans Berger (1).
There is general agreement that the EEG in AD is characterized
by a relatively nonspecific, stage-dependent pattern of diffuse
slowing (Fig. 20.2). In the earliest phase of the disease the EEG
can be normal, especially in late onset cases (14). However, in
general, a normal EEG argues against a diagnosis of AD (30).
Power decreases in the beta and alpha band and increases in the
theta and delta band (63–68). Alpha peak frequency decreases
and reactivity of the alpha rhythm to eye opening also dimin-
ishes (68). Increase in theta and decrease in beta power are
probably the first and most sensitive EEG changes in AD
(10,66,69,70). The increase in theta may be more outspoken in
early onset AD (71,72). A slowing of the alpha peak frequency
occurs only after the increase in theta (70). Later alpha power
decreases and finally in a relatively late stage delta increases. The
increase in slow activity is most prominent over left temporal
areas (65,73). Furthermore sources of alpha and beta display a
shift toward more frontal regions in AD (74). An increase in
gamma band power has been reported in AD during rest and
cognitive tasks, but it is questionable whether EEG activity
above 20 Hz can be reliably distinguished from muscle artefact
(75,76). In the course of the disease, there is a further decrease
of beta and alpha power, slowing of the alpha peak frequency,
and a further increase in theta and delta power (66,77). Focal
abnormalities are not typical and may be related to the frequent
occurrence of vascular pathology in AD. FIRDA is rare and can

Figure 20.2 Male, 87 years. Alzheimer disease. Irregular, mildly slowed alpha rhythm, more pronounced at posterior tem-
poral areas and diminished reactivity to eye closure. Calibration mark: vertical, 50 V; Horizontal, 1 second. Filter settings:
time constant 1 second. Low-pass filter, 35 Hz. Source derivation.
380 Part III ■ Clinical EEG: General
Figure 20.3 Female, 74 years. Mild cognitive impairment. Spike-wave complexes in the left temporal regions. Calibration
mark: vertical, 50 V; Horizontal, 1 second. Filter settings: time constant 1 second. Low-pass filter, 35 Hz. Source derivation.
be found in only 2.9% of EEGs in AD (78). Epileptiform abnor-
malities are also rare, but AD is associated with a significantly
increased risk of epilepsy, especially in early onset and severe
AD (79). Temporal epileptiform activity may mimic AD
(80–82). An example can be found in Figure 20.3.
EEG abnormalities pointing to abnormal functional interac-
tions between brain regions have been reported frequently in
AD. Clinical symptoms may be due to a considerable extent to
disrupted communication between distributed brain areas, jus-
tifying a description of AD as a disconnection syndrome (83).
Functional connectivity can be studied by computing linear
(such as coherence) or nonlinear (such as synchronization like-
lihood or mutual information) measures of correlation
between EEG signals recorded from different brain regions. A
decrease of EEG coherence in the alpha band in AD has been
reported in many studies (69,84–90). Nonetheless van der Hiele
et al. reported no changes of alpha band coherence in AD,
although these authors found also that theta relative power was
increased in AD patients and this increase was related to
decreased performance in all cognitive domains (68). A lower
coherence in the beta band has also been described (89,91).
Several studies agree with respect to a decrease of coherence in
the theta band interhemispherically (69,92) and to a decrease of
coherence in the alpha band (92), more specifically in the tem-
poro–parieto–occipital region (88) and more focused on the
left temporal region (69). Changes in the delta band are less evi-
dent, with a study reporting decreased coherence both inter-
and intrahemispherically (92) while another one (88) noted
increased coherence between frontal and posterior regions but
restricted to a few AD patients.
The normal parietal to frontal direction of information flow
in the alpha and beta band is disrupted in AD patients (93).
Nonlinear measures also showed a reduced functional connec-
tivity in AD (42,45,94,95).
In addition to spectral analysis and assessment of functional
coherence, several studies have explored advanced types of sig-
nal analysis in AD. Artificial neural networks are nonlinear clas-
sifiers that can improve the distinction between EEGs recorded
in healthy subjects and patients suffering from MCI or AD
(96–98). Nonlinear EEG analysis attempts to reconstruct the
dynamic processes in neural networks underlying the EEG
(99,100). Nonlinear analysis has been applied to the EEG in AD
in many studies (4,101). Measures such as the correlation
dimension, the largest Lyapunov exponent, and the Kolmogorov
entropy are used to characterize changes in the complexity of
brain networks. In AD, several authors have reported a loss of
complexity of brain dynamics (97,102–111). These changes may
correlate with the severity of clinical symptoms. However, it
should be stressed that results from nonlinear EEG analysis may
be biased by more basic linear properties of the EEG (112).

These properties can be measured more easily with spectral
analysis and coherence and may depend on the frequency band
considered (113,114). More recently, complex brain networks
have been characterized with graph theory (115,116). While
healthy subjects have brain networks with the so-called small-
world topology, characterized by a combination of high cluster-
ing and short path lengths, AD patients may have more random
brain networks due to loss of critical communication lines
(117–120).
The diagnosistic accuracy of EEG changes, based on measures
of EEG slowing or disturbed functional connectivity, fluctuates
around 80%. Summarizing 16 studies published between 1983
and 1995, Jonkman reports a total accuracy between 54% and
100% with a median of 81% (8). An accuracy of 100% was
obtained in several studies of the EEG during random eye
movement (REM) sleep (121–123). In one of the few studies
relating EEG abnormalities to the gold standard of neu-
ropathology, Robinson et al. showed that 87% of patients with
an autopsy-proven diagnosis of AD had an abnormal EEG
(124). However, in the matching control group 35% had abnor-
mal EEGs. Strijers et al. reported an accuracy of 84% in distin-
guishing between controls and AD patients using either EEG or
hippocampus atrophy (125). An earlier study reported a sensi-
tivity of 85% and a specificity of 88% of medial temporal atro-
phy for the detection of mild-to-moderate AD: 85% and 88%
(126). Using visual analysis and a semiquantitative rating, the
grand total EEG (GTE) score, Claus et al. related the diagnostic
gain of the EEG to the a priori likelihood of a diagnosis of AD
(127). They showed that in the case of an abnormal EEG a diag-
nostic gain could be obtained when the a priori likelihood of
AD was between 30% and 40%. Few studies have directly com-
pared the accuracy of visual and quantitative analysis
(128–130). These older studies suggest a comparable accuracy
between 75% and 80% for visual and computer analysis. Adler
et al. reported that left temporal alpha coherence and global
theta power allowed an identification of AD patients with a sen-
sitivity of 87% and a specificity of 77% (69).
We can conclude that the test characteristics of the EEG do
not compare unfavorably with those of some other biomarkers
(12). In addition it should be noted that these results can be
obtained with relatively simple semiquantitative visual assess-
ment of the EEG (127). Furthermore, compared to other labo-
ratory studies, the EEG is relatively cheap, generally available,
and noninvasive. However, EEG studies are confronted with the
same problem as studies of other biomarkers, namely the prob-
lems related to the lack of a good gold standard. We will come
back to this issue in a discussion of the practical use of the EEG
in dementia diagnosis.
Both EEG slowing and disturbed functional connectivity
can correlate with severity of dementia and cognitive deficits, as
well as structural changes. EEG slowing correlates with ventric-
ular widening, cortical atrophy, and worse psychometric per-
formance in AD (67). An anterior shift of alpha and beta
sources, and stronger sources in the low-frequency bands corre-
late with severity of dementia (74). Delta power correlates neg-
atively and alpha 1 power positively with the AD patient’s
MMSE scores (65). Decreased cortical alpha power correlates
Chapter 20 ■ Dementia and EEG 381
with hippocampal atrophy (131). Lower alpha band reactivity
correlates with decreased performance on tests of global cogni-
tion, memory, and executive functioning; brain atrophy corre-
lates with psychomotor speed (68,132). Loss of coherence in the
fast frequencies relates to impaired performance on neuropsy-
chologic tests (36,133). More specifically, Dunkin et al. found
correlations between beta band coherence between right tem-
porooccipital electrodes (P4, T4, T6) and scores on the Boston
Naming task, and a correlation between scores on block design
and fascicle coherence (defined, for left and right hemisphere,
as long distance coherence between P3/4 and F3/4, F1/2 and
F7/8) (133). Lower beta band (14 to 22 Hz) mean synchroniza-
tion likelihood (averaged over all pairs of electrodes) correlates
with lower MMSE scores in MCI and AD patients (45). Loss of
interhemispheric coherence in AD correlates with atrophy of
the corpus callosum (89). Loss of global field synchronization
(an average measure of coupling between all brain regions) in
AD correlates with MMSE and CDR scores (91).
EEG changes in AD can predict to a certain extent the fur-
ther course of the disease. In general, EEG slowing already pres-
ent at the time of initial diagnosis predicts a poor outcome
(8,134). Helkala et al. showed that AD patients with more severe
EEG abnormalities as compared to those with relatively normal
EEGs had a worse prognosis with more severe neurologic
abnormalities and a greater risk of institutionalization later in
the disease (135). Claus et al. showed that slowing on qEEG pre-
dicts cognitive and functional decline in AD (136). In a later
study, the authors showed that mortality in patients with early
AD was predicted by a decrease in alpha and beta activity deter-
mined by spectral analysis of the EEG (137). Delta and alpha
power could predict loss of activities of daily life and inconti-
nence in the study of Nobili et al. (138).
EEG changes in AD may provide some understanding of the
genetic and other factors involved in the pathophysiology of this
disorder apart from their diagnostic and prognostic meaning.
Purely genetic causes of AD are rare. Genetic changes in genes
for amyloid beta precursor protein, presenilin 1, and presenilin
2 account for less than 5% of AD cases (12). Most cases of AD
are sporadic, but genetic risk factors are important. This is espe-
cially relevant in the case of the epsilon 4 (e4) allele of the
apolipoprotein E (ApoE) gene. The ApoE e4 allele is a genetic
risk modifier for AD (139). Considering the fact that ApoE e4 is
a risk factor for AD and that EEG changes such as increased
theta and loss of functional connectivity could be early bio-
markers of AD pathology, it is interesting to study the possible
relationship between genetics and EEG markers. Lehtovirta et al.
showed that AD patients with ApoE e4 genotype had more pro-
nounced slow wave activity (140–142). In contrast, Jelic et al.
could not confirm the relation between ApoE e4 and EEG slow-
ing but did report a decreased functional connectivity in AD
patients with the e4 allele (143). Kramer et al. used the synchro-
nization likelihood to determine the relation between ApoE
genotype and functional connectivity in subjects with subjective
memory complaints and AD patients (144). In this study sub-
jects with ApoE e4 had higher synchronization in the lower and
upper alpha band than subjects without ApoE e4, irrespective
of diagnosis. Compared to controls, AD patients had lower
382 Part III ■ Clinical EEG: General
synchronization in the upper alpha and beta band. This study
suggests that both the e4 allele as well as AD pathology influence
functional connectivity, but they do so in opposite directions
and partly in different frequency bands. In a large study involv-
ing 145 individuals with AD, their unaffected relatives and unre-
lated individuals, Ponomareva et al. showed that AD patients
with the ApoE e4 allele had a stronger loss of alpha power than
e4 noncarriers (145). The resting EEG of nondemented relatives
of AD patients showed no differences related to ApoE genotype.
During hyperventilation, however, only e4 allele carriers showed
synchronous high-voltage slow and sharp waves. This excessive
synchronization is somewhat reminiscent of the increased syn-
chronization reported by Kramer et al. but does not agree with
the loss of coherence found by Jelic et al. (143,144). Apparently,
the relation between ApoE genotype and EEG is complex and
depends on the study conditions, the EEG measure investigated,
and the presence or absence of AD pathology.
In contrast, the relation between cholinergic loss and EEG
changes in AD seems to be more straightforward. There is con-
siderable evidence for a loss of central cholinergic function in
AD (146,147). This cholinergic loss is presumably due to a loss
of neurons in the nucleus basalis of Meynert, which projects to
medial temporal and neocortical brain areas (146). Impaired
cholinergic function in AD is held responsible for at least some
of the cognitive deficits in this disorder. This concept is known
as the “cholinergic hypothesis” (148). The cholinergic hypothe-
sis is of considerable interest, since it suggests a possible way to
treat AD symptoms by activating the central cholinergic system.
This could be achieved by administration of drugs that inhibit
the breakdown of acetylcholine by acetylcholine esterase.
Several acetylcholine esterase inhibitors such as tacrine,
donepezil, rivastigmine, and galantamine have been used clini-
cally in AD, with varying results.
Cholinergic loss could explain EEG changes in AD. In an ani-
mal study, Buzsaki et al. showed that damage to the nucleus
basalis results in a slowing of the EEG (149). In another animal
study, Villa et al. showed that loss of acetylcholine results in an
increase in low-frequency coupling and a decrease in high-fre-
quency coupling (150). These experimental findings might
explain why alpha and beta band connectivity are lost in AD,
whereas low-frequency coupling may be spared or even
increased. Administration of scopolamine, which blocks central
cholinergic activity, to healthy subjects decreases EEG inter-
hemispheric coherence in the delta and beta-1 bands in the rest-
ing state. (151). Increase in delta and theta and decrease of alpha
and beta after scopolamine administration have also been
reported in healthy subjects (152). Scopolamine induces a pat-
tern of memory and cognitive deficits in healthy subjects that
resembles the changes in early AD (153). The general slowing of
the EEG in AD has been interpreted as a cholinergic effect (154).
If we assume that EEG slowing in AD is due to cholinergic
loss, we might expect that treatment with drugs that stimulate
the central cholinergic system will reverse the EEG slowing.
Furthermore, EEG changes might predict which patients will
show the best response to treatment. Some evidence for both
conjectures is available. EEG slowing in AD is reversed by treat-
ment with cholinesterase inhibitors (155–158). EEG slowing at
onset, or a relatively strong EEG change after a single test dose
of the drug, may predict long-term response to cholinesterase
inhibitors (156,159–161).
Obviously, the cholinergic hypothesis of AD has received a
lot of attention in AD in view of its therapeutical implications.
However, treatment with cholinesterase inhibitors is only mod-
erately effective in a subset of patients. This suggests that other
factors than a loss of central cholinergic activity may be impor-
tant. For instance, receptors may be downregulated or even dis-
appear from the membranes, and if there is neuronal death this
will certainly lead to a decrease in targets of cholinergic inputs.
According to Dringenberg other neurotransmitters in addition
to acetylcholine have to be considered (162). Changes in several
different neurotransmitters could perhaps explain the global
cortical hyperexcitability of the cortex in AD while the brain is
at rest (163). In the earliest phase of AD there may be no evi-
dence of cholinergic dysfunction (164). In MCI the cholinergic
system may even by upregulated, possibly as a compensatory
mechanism (143). A relation between EEG changes and amy-
loid beta1–42 is suggested by findings in patients with Down
syndrome. In Down syndrome, an extra chromosome 21 is
present. The gene for amyloid precursor protein lies on chro-
mosome 21, and this has been related to the high risk of early
onset of AD in Down patients. In line with this concept, AD in
Down is predicted by EEG slowing (165). Furthermore, EEG
slowing correlates with AD pathology in Down. Another
important factor in AD pathology is the abnormally phospho-
rylated tau protein. Jelic et al. have reported a correlation
between tau protein in the cerebrospinal fluid and the
alpha/delta ratio of the EEG in AD patients (166). Babiloni et al.
have shown that homocysteine levels, related to vascular
pathology, are related to EEG slowing in mild AD (167). In gen-
eral, the simultaneous occurrence of AD and vascular pathol-
ogy may be more important than has been realized in the past.
According to Schreiter-Gasser et al. the presence of vascular
pathology could magnify the effect of AD pathology on demen-
tia severity (168).
FRONTOTEMPORAL LOBAR DEMENTIAS
Originally, frontal lobe dementia was synonymous with Pick
disease, a rare degenerative disorder characterized by typical
neuronal inclusions bodies (Pick bodies). Currently, the con-
cept has been broadened and includes degenerative frontal
dementia with or without Pick bodies, progressive aphasia, and
semantic dementia. This spectrum of disorders is referred to as
FTLD. FTLD is characterized by early behavioral and language
problems, frontal-type cognitive dysfunction, and relatively late
occurrence of memory complaints. In the early phase, FTLD is
often confused with psychiatric disorders, including personality
disorders and depression.
In view of the frontotemporal dominance of the pathology
in FTLD, it can be expected that the EEG and in particular the
posterior alpha rhythm remains normal until late in the disease.
Several studies have reported normal or only mildly disturbed
EEGs in FTLD (169–172). The EEG in FTLD may show signif-
icantly less slowing than in AD (171,173,174). Functional

connectivity, as measured with the synchronization likelihood,
is lower in FTLD compared to AD; FTLD values are compara-
ble to nondemented controls (174). EEG coherence in FTLD
was also comparable to controls (173). Using qEEG analysis,
Yener et al. reported EEG abnormalities in FTLD patients (175).
On the basis of these findings FTLD patients could be distin-
guished from both AD patients as well as healthy controls. More
recently in a study with neuropathologic confirmation of diag-
nosis, Chan et al. showed that EEGs of FTLD patients could not
be distinguished from EEGs of AD patients; however, a very
simple visual classification scheme was used in this study and
no comparison was made with EEGs of age-matched nonde-
mented controls (176). The discrepancies between the various
studies could be due to variability of the neuropathologic pat-
tern of degeneration. In a study with neuropathologic confir-
mation of the diagnosis, Passant et al. showed that onset before
65 years is associated with behavioral–mood changes and mild
EEG abnormalities, while onset after 65 years is related to lan-
guage disturbances and a normal EEG (177).
PARKINSON DISEASE WITH
AND WITHOUT DEMENTIA:
DEMENTIA WITH LEWY BODIES
Next to AD, Parkinson disease is one of the most prevalent neu-
rodegenerative disorders. Parkinson disease is a multisystem
disorder characterized by motor symptoms such as akinesia,
rigidity, tremor and postural instability, cognitive and neu-
ropsychiatric problems. Mild executive cognitive dysfunction
can occur early in the disease, while full-blown dementia is a
frequent complication in the later stages of the disease. A dis-
tinction has been made between Parkinson disease with
dementia (PDD) and DLB, mainly based on the time interval
between onset of extrapyramidal and cognitive symptoms; this
distinction may be a bit artificial, and in practice considerable
overlap can be expected between PDD and DLB.
In Parkinson disease without dementia (PD), the EEG is
probably relatively preserved, at least in the early stages.
However, even in nondemented Parkinson patients a correla-
tion between motor disability and slowing of the background
activity has been found (178). EEG slowing emerges with pro-
gressive cognitive disturbances (179). In contrast to PD, PDD is
often associated with a slowing and impaired reactivity of the
EEG (178,180,181). As in AD, EEG slowing in PDD is respon-
sive to treatment with rivastigmine (182).
Extensive slowing of the EEG has also been reported in DLB
(183). In the study of Liedorp et al. a normal EEG or an EEG
with only focal abnormalities argued strongly against a diagno-
sis of DLB, whereas an EEG with both focal and diffuse abnor-
malities strongly favored a diagnosis of DLB (30). Of interest,
FIRDA may occur frequently in DLB, but not in AD
(78,183,184). In the study of Roks et al., FIRDA was found in
2.9% of the patients with AD and in 33.3% of the patients with
DLB (78). Some authors report comparable slowing of the EEG
in DLB and AD (185,186). However, slowing of the EEG back-
ground is more severe in DLB than in AD according to several
Chapter 20 ■ Dementia and EEG 383
authors (78,187–189). Using a semiquantitative visual analysis
(GTE) Roks et al. could distinguish between DLB and AD with
a sensitivity of 72% and a specificity of 85% at a GTE cutoff of
9.5. Kai et al. showed increased delta and theta power as well as
intrahemispheric coherence in DLB as compared to AD (189).
AD patients had less beta power mainly in the posterior
regions. In DLB patients treatment with the acetylcholine
esterase inhibitor donepezil resulted in a power decrease in
delta and theta bands but a similar response was not present in
AD patients. Increased delta band power variability, increased
delta coherence, and decreased alpha coherence were demon-
strated in DLB compared to AD and elderly controls by
Andersson et al. (190). Bonanni et al. studied 50 DLB, 50 AD,
and 40 PDD patients with relatively MCI; diagnosis was con-
firmed in most patients after a 2-year follow-up (187). DLB
patients had a slower and more variable dominant frequency
compared to AD patients. Only half of the PDD patients had
EEG abnormalities comparable to DLB patients. The authors
conclude that the EEG may be useful to distinguish between AD
and DLB. This conclusion is in line with some of the other stud-
ies that report more extensive slowing and a higher prevalence
of FIRDA in DLB. Furthermore, such a distinction can be made
with relatively high sensitivity and specificity using simple
visual assessment of the EEG.
HUNTINGTON DISEASE
Huntington disease is a rare neurodegenerative disorder with
an autosomal-dominant inheritance. It is caused by multiple
CAG repeats of the Huntington gene on chromosome 4. It is
characterized by severe dementia, psychiatric disturbances, and
involuntary movements, notably chorea. Neuropathologically
there is conspicuous atrophy of the caudate nucleus. A charac-
teristic finding in up to 40% of Huntington patients is the
occurrence of a low-voltage EEG (191). Streletz et al. reported
increased theta and decreased alpha in Huntington patients
(192). The EEG may show changes in nondemented mutation
carriers for Huntington disease. van der Hiele et al. found less
alpha band reactivity during memory activation in mutation
carriers compared to nonmutation carriers (193). However,
compared to genetic testing the clinical significance of the EEG
in Huntington disease is probably quite limited.
PROGRESSIVE SUPRANUCLEAR PALSY
AND CORTICOBASAL GANGLIONIC
DEGENERATION
Progressive supranuclear palsy (PSP; also called Steele–
Richardson–Olzewski syndrome) is another rare disease char-
acterized by frontal-type cognitive dysfunction, dementia,
extrapyramidal motor symptoms (in particular axial rigidity),
and supranuclear vertical gaze palsy. Fowler et al. compared
EEG findings in 22 patients with PSP with EEGs of 22 matched
AD patients (194). PSP patients had EEGs with low voltage,
preserved alpha peak frequency, and modest increase of theta;
one patient had FIRDA. Findings in the AD control group were
384 Part III ■ Clinical EEG: General
largely similar. Compared to PSP patients, patients with corti-
cobasal ganglionic degeneration (CBD) had more focal abnor-
malities, often contralateral to the dominantly affected side and
ipsilateral to the atrophy on MRI (195). In this study the preva-
lence of FIRDA was relatively high in both disorders: 20% in
CBD and 35.7% in PSP.
VASCULAR DEMENTIA
The diagnosis of VaD depends on three conditions: (i) a
dementia syndrome; (ii) demonstration of vascular lesions in
the brain; and (iii) evidence of a causal relation between the
vascular lesions and the dementia. Although both dementia and
vascular lesions are common, it is not at all easy to prove they
are causally related in a single patient. To complicate the issue,
vascular lesions are increasingly considered to be important in
AD as well. The NINDS-AIREN criteria emphasize the impor-
tance of radiologic evidence for vascular lesions (196). According
to Holmes et al., these criteria have a sensitivity of 43% and a
specificity of 95% (197).
The contribution of the EEG to diagnosis and prognosis in
VaD is more limited than in AD. According to Jonkman, the lit-
erature up to 1996 shows no consensus on the ability of the
EEG to distinguish between AD and VaD (8). Several of the
studies discussed in this review did not use the NINDS-AIREN
criteria and depended on older concepts such as “multi-infarct
dementia.” In the studies of Erkinjuntti et al. degree of diffuse
abnormalities/mean frequency background did not distinguish
between AD and MID (198,199). Leuchter et al. investigated
coherence between EEG signals recorded from brain areas
linked by two different types of connections: areas linked by
bands of long corticocortical fibers showed greatly diminished
coherence in subjects with AD but not in MID patients. On the
contrary areas linked by broad connective networks showed the
largest decreases in coherence among MID subjects. These find-
ings can be interpreted as implying that AD is a neocortical
“disconnection syndrome” in which there is a loss of long cor-
ticocortical tracts, whereas vascular disease (MID) most promi-
nently affects broad fiber networks that may be affected by
diffuse subcortical vascular damage (86,87). According to
Rosen et al. the EEG in multi-infarct dementia displays
increased slow activity in combination with relatively preserved
posterior activity, whereas in AD slowing is combined with
abnormalities of posterior activity (171). Robinson et al. failed
to find EEG differences between AD and VaD (124).
More recent studies often identify EEG differences between
VaD and AD. Jeong et al. used nonlinear EEG analysis to com-
pare AD and VaD (200). AD patients showed a loss of complex-
ity reflected in lower values of the correlation dimension and
largest Lyapunov exponent, whereas complexity was pathologi-
cally increased in VaD (200). Individual analysis of alpha fre-
quency and power could distinguish between AD and VaD in
the study of Moretti et al. (201). Frontoparietal synchronization
likelihood in the alpha 1 band was more reduced in VaD
patients than in AD in a study of Babiloni et al. (63). This more
pronounced loss of long distance connectivity in VaD patients
seems to be in disagreement with the coherence findings of
Leuchter et al., which suggests that connections between areas
linked by long corticocortical fibers (see above) would be more
affected in AD (86,87). Quantitative but not visual analysis can
distinguish between patients with subcortical VaD with mild or
moderate dementia (202). qEEG measures correlated signifi-
cantly with MMSE scores in this study. In a large study based
upon visual analysis, VaD was associated with the combination
of diffuse and focal abnormalities; normal EEGs made a diag-
nosis of VaD less likely (30). Recent reviews agree that the EEG
can sometimes distinguish VaD from AD, but question the clin-
ical usefulness of this, since other laboratory investigation such
as MRI may be more effective in this respect (6,11). Again, the
significant overlap between AD and VaD especially in the
elderly population should be stressed.
DELIRIUM AND METABOLIC/ TOXIC
ENCEPHALOPATHIES
Delirium is a clinical syndrome that can be caused by meta-
bolic, toxic, or infectious encephalopathies, frequently in the
context of pre-existing neurodegenerative brain disorders such
as AD. A relation between EEG slowing and disturbed arousal
in delirious patients was already observed in the classic study by
Romano and Engel (203). The sensitivity of the EEG to brain
dysfunction in delirium has been confirmed with quantitative
analysis (204). Delirious patients showed increased delta and
theta, and decreased alpha as well as a slowing of alpha peak fre-
quency; these changes were correlated with the severity of cog-
nitive disturbances. EEG slowing in delirium is often
nonspecific, making a distinction between delirium and AD
difficult; moreover, both conditions frequently occur together.
Even so, recent studies have attempted to increase the specificity
of the EEG in this respect (205,206). Prolonged activity, consist-
ing of EEG recording during 3 minutes with eyes open, signifi-
cantly improved the ability of the EEG to distinguish between
delirious patients with or without underlying dementia and
cognitively unimpaired elderly (205). In another study, delir-
ium in dementia was characterized by slower alpha peak fre-
quency, loss of alpha power, increased theta and delta; these
changes correlated with severity of delirium and cognitive dis-
turbances (206).
Other EEG patterns may suggest specific causes of delirium
(6,14,207,208). Generalized triphasic waves are often associated
with metabolic encephalopathy due to hepatic or renal disease
but can also occur in other conditions such as postanoxic states.
In metabolic encephalopathies it is especially the rate of change
that determines the EEG abnormalities (6). These triphasic
waves are often responsive to benzodiazepine administration,
but this is not a proof of their epileptic origin. Low-voltage
records with excess fast activity can be found in withdrawal
from alcohol or sedative drugs. Focal abnormalities should
always arouse suspicion of underlying structural lesions. In high
doses antidepressives and antipsychotics, in particular clozap-
ine, often produce epileptiform abnormalities in addition to
slowing. Nonconvulsive status epilepticus is a rare but impor-
tant and treatable cause of delirium; it often arises in elderly
 Chapter 20 ■ Dementia and EEG 385

patients after sudden withdrawal of antiepileptic drugs. The
EEG may be normal in alcohol dementia in subjects younger
than 60 years but can show abnormalities including a low-volt-
age record in older subjects (209). In these subjects slowing of
the EEG correlates with cortical atrophy. EEG changes in
Wernicke–Korsakov syndrome may vary widely.
CREUTZFELDT–JAKOB DISEASE
Creutzfeldt–Jakob disease (CJD) is a transmissible spongiform
encephalopathy caused by accumulation of abnormal isoform
of host-encoded cellular prion protein. Different forms of CJD
have been described: sporadic, iatrogenic, and variant. The new
variant of CJD is possibly related to bovine spongiform
encephalopathy. Iatrogenic CJD is caused by transmission of
infectious prion protein partials through tissue or medical
instruments. Most forms are sporadic. Other prion disorders
are Kuru, familial fatal insomnia, and Gertmann–Straussler–
Scheinker syndrome. Clinically, CJD is characterized by a rap-
idly progressive dementia with myoclonus. Cerebellar, pyrami-
dal, and extrapyramidal motor symptoms as well as psychiatric
symptoms also occur frequently. Early visual symptoms are
characteristic for the Heidenhain variant of CJD.
Although CJD is a rare cause of dementia, it is considered an
indication for EEG recording in view of the characteristic peri-
odic sharp wave complexes (PSWC). The EEG is an essential
part of the diagnostic process in the WHO diagnostic classifica-
tion criteria of sporadic CJD (210). According to these criteria,
a definite diagnosis of CJD requires neuropathologic confirma-
tion or demonstration of the abnormal prion protein. Probable
CJD requires (i) a rapidly (within 2 years) progressive demen-
tia; (ii) typical PSWC in the EEG and/or 14-3-3 protein in the
cerebrospinal fluid; and (iii) two of the following: (a)
myoclonus, (b) ataxia and visual signs and symptoms, (c)
extrapyramidal and/or pyramidal signs and symptoms; and (d)
akinetic mutism. Possible CJD can be diagnosed if the above
criteria are met with the exception of the EEG or 14-3-3 protein
abnormalities.
EEG changes in CJD occur in different stages (211). In the first
stage, the EEG shows nonspecific slowing and sometimes FIRDA.
In the middle stage, the EEG is characterized by PSWC in about
two third of patients (Fig. 20.4). In the final stage, background
activity between the complexes becomes increasingly flat, and
finally the complexes may disappear. The PSWC may be localized
in the occipital regions in the Heidenhain variant of CJD. The
PSWC disappear during sleep, sedation, and stimulation. Often a

Figure 20.4 Female, 61 years. Progressive cognitive disturbances, myoclonus, and impaired consciousness. Diagnosis:
Creutzfeldt–Jakob disease. The EEG shows generalized periodic sharp wave complexes (PSWC) with a repetition frequency
of slightly above 1/ sec. Calibration mark: vertical, 50 V; Horizontal, 1 second. Filter settings: time constant 1 second.
Low-pass filter, 35 Hz. Average reference.
386 Part III ■ Clinical EEG: General
clear time locking to clinical myoclonus is absent. If no PSWC are
seen after 12 weeks of disease onset, this argues against a diagno-
sis of CJD. PSWC are not assumed to occur in variant CJD (212),
but exceptions to this rule have been described (213). In the ini-
tial stages PSWC can occur unilaterally (214). This raises the
question whether PSWC are distinct from PLEDs (lateralized
periodic discharges or LPD according to the new terminology
(215)). According to some authors PSWC and PLEDs may be dif-
ferent phenomena, but the distinction seems to be a bit artificial
(211,216). The detection of EEG abnormalities in suspected CJD
can be improved with advanced EEG analysis techniques such as
independent component analysis (ICA) (216,217).
Strict criteria for PSWC in CJD were presented by Steinhof
et al. (218). To classify as PSWC the following criteria should be
met: (i) Strictly periodic cerebral potentials, the majority with a
duration between 100 and 600 msec and an intercomplex inter-
val between 500 and 2000 msec; (ii) generalized and lateralized
complexes accepted; and (iii) at least five repetitive intervals
with a duration difference of less than 500 msec to rule out
semiperiodic activity. If these criteria are met, PSWC have a
sensitivity of 64%, specificity of 91%, a positive predictive value
of 95%, and a negative predictive value of 49% (219). Zerr et al.
reported a specificity of 74%, a positive predictive value of 93%,
and a sensitivity of 66% for PSWC (220). These numbers
should be compared to the 14-3-3 protein in the cerebrospinal
fluid: sensitivity 84% and specificity 94% (220,221). In an
autopsy series of 201 patients suspected of CJD, van Everbroeck
et al. found PSWC in 52% of autopsy-confirmed CJD cases, but
also in three patients with AD, one patient with DLB, and one
patient with VaD (222). In general, autopsy-confirmed studies
indicate that PSWC have a fairly high specificity, but a rather
low sensitivity for CJD. This could be due to the fact that PSWC
develop later in the disease or may not appear at all in a sub-
stantial percentage of patients. Genetic forms only display
PSWC in 10% of cases (223). Furthermore, PSWC are associ-
ated with methionine homozygosity or methionine/valine het-
erozygosity at codon 129 of the prion protein on chromosome
20, while PSWC hardly occur in CJD patients who are homozy-
gotes for valine at codon 129 (6,211).
No completely satisfactory explanation of PSWC in CJD
exists. According to Tschampa et al. a reduction of paralbumin-
positive cells in the thalamic reticular nucleus could be responsi-
ble for both PSWC and myoclonus in CJD (224). These authors
stress that “functional integrity (at least in parts) of the subcor-
ticocortical or thalamocortical network is a prerequisite for the
generation of PSWC.” In agreement with this, modeling studies
have shown that typical periodic sharp waves can be reproduced
in a model of the alpha rhythm assuming (i) an increased excita-
tion/inhibition balance and (ii) intact but abnormally active cor-
ticothalamic loops (225,226). PSWC may reflect pathologic,
highly nonlinear oscillations in corticothalamic networks.
HERPES ENCEPHALITIS
The EEG is always abnormal in acute encephalitis. Herpes sim-
plex encephalitis typically shows periodic discharges, mainly
over the temporal lobes, but they may appear only after several
days. Since treatment has to start directly, the EEG does not play
a large role in initial diagnosis. The EEG may provide prognos-
tic information (6).
SUBACUTE SCLEROSING PANENCEPHALITIS
Subacute sclerosing panencephalitis (SSPE) is a neurodegener-
ative complication of measles infection in childhood (6). The
EEG may show characteristic high-voltage generalized com-
plexes occurring in a periodic fashion with very long intervals
(periodic long interval diffuse discharges [PLIDDs]); these
complexes have first been described by Radermecker and Poser
(227). The complexes are usually associated with myoclonic
jerks. It should be pointed out that PLIDDs are not fully specific
for SSPE, since they have also been described in phencyclidine
intoxication (228,229).
HUMAN IMMUNODEFICIENCY VIRUS
EEGs of 47 patients with the acquired immunodeficiency syn-
drome (AIDS) or AIDS-related complex (ARC) were described
by Gabuzda et al. (230). EEG abnormalities, related to demen-
tia or opportunistic CNS infections, were found in 67% of
patients with AIDS and 36% of patients with ARC. No EEG
abnormalities were described in a group of neuropsychologi-
cally unimpaired patients infected with HIV (231). In a study by
Harrison et al. patients with AIDS had more signs of neurologic
dysfunction, cognitive problems, and EEG and MRI abnormali-
ties than asymptomatic seropositive subjects (232).
CLINICAL USE OF EEG IN DEMENTIA
Following the review of EEG changes in the major categories of
dementia, we return to our initial questions: why are there such
large differences of opinion concerning the role of the EEG
between different dementia guidelines? What is the clinical use-
fulness of the EEG in dementia, and how should it be used? We
will address these questions and attempt to give some tentative
answers.
Methodologic problems are one source of controversies
between various studies. Many studies, mainly but not exclu-
sively the older ones, suffer from various problems such as
small group size; bad matching of groups in terms of age, gen-
der, and disease duration; and influence of centrally acting
drugs. Results may not be representative since they concern
patients referred to secondary or tertiary referral centers for
dementia diagnosis. There is a lack of consensus on which EEG
measures to use and insufficient knowledge on the repro-
ducibility of these measures. Studies that fail to reproduce ear-
lier findngs may be underrepresented in the literature due to
publication bias. Inadequate statistical analysis hampers some
studies and could be responsible for the reporting of false-pos-
itive findings. Despite these methodologic problems, which are
not unique for EEG research in dementia, many excellent stud-
ies have been published, and our knowledge has increased con-
siderably over the last decades. EEG findings in normal aging,
MCI, and several major categories of dementia, as well as their

possible pathophysiologic substrate, are now fairly well known.
Other problems may therefore be more important in translat-
ing this knowledge to clinical practice.
One major problem is the lack of a good gold standard (12).
It is almost impossible to carry out a laboratory investigation if
there is no gold standard to serve as reference. Currently the
closest one can get to a gold standard in dementia research is a
combination of clinical diagnosis and neuropathologic confir-
mation. Standards of clinical diagnosis have improved consid-
erably over the past two decades with the introduction of
guidelines for several of the major diseases causing dementia.
Similar standards are being introduced for neuropathologic
diagnosis. However, clinical diagnosis according to well-defined
guidelines correlates only moderately with neuropathologic
diagnosis; furthermore, neuropathologic confirmation is often
not available and neuropathologic diagnosis itself also has a
limited interrater reproducibility. Under these conditions it is
almost impossible to determine reliably the sensitivity and
specificity of a laboratory test such as the EEG in distinguishing
between the normal healthy state and various conditions that
can give rise to dementia. One solution is to use information on
the further course of the disease to validate the initial diagnosis.
An excellent example of this can be found in the study of
Bonanni et al. (187). What could also be helpful is to give up the
idea of eternal platonic truths underlying all our observations.
Concepts such as “Alzheimer disease” or “vascular dementia” or
“dementia with Lewy bodies” are man-made attempts to detect
regularities in our observations; our concepts can be enriched
and may even have to evolve by incorporating all clinical, neu-
ropathologic, and laboratory information available into consis-
tent and useful entities.
The final and perhaps most important problem of EEG
studies in dementia may be that many of them seem to be
addressing the wrong problem. If the purpose of the investiga-
tion is to describe the pathophysiology of a clinical dementia
syndrome, it makes sense to compare age-matched healthy con-
trols to subjects with a specific disorder causing dementia.
However, if the goal is to determine the clinical usefulness of
the EEG in dementia, such a comparison is useless: distinguish-
ing between a healthy person and a clearly demented patient is
a trivial problem from a clinical point of view. Neither the EEG
nor any other laboratory investigation is a test “to see if a
patient is demented.” To be clinically useful, a laboratory test
has to give an answer to a clinical problem, and the answer has
to have consequences for diagnosis, prognosis, and treatment.
Ultimately it has to be demonstrated that doing a test, as com-
pared to not doing the test, benefits the health and quality of life
of the patient. Very few EEG studies attempt to address these
questions, and this may be a major source of discrepancies
between various guidelines. Guidelines that focus upon sensi-
tivity and specificity of comparisons between healthy subjects
and dementia may be far more optimistic in their assessment of
the usefulness of the EEG than guidelines that stress demon-
strated clinical usefulness.
In view of these problems, we now have to address our final
question: in what conditions can the EEG be expected to be really
Chapter 20 ■ Dementia and EEG 387
helpful in dementia diagnosis? The answer to this question can
only be tentative since the proper kind of clinical studies simply
have not yet been conducted as indicated in the previous section.
First of all it should be stressed that the results of an EEG exam-
ination should also be considered within their clinical context,
that is, taking the other relevant information (clinical, neuropsy-
chological, radiologic) into account. Ideally this should be done
in the context of a multidisciplinary meeting of all disciplines
involved, but it may not be easy to do this outside specialized
memory clinics. Active participation of the clinical neurophysiol-
ogist in such multidisciplinary teams may greatly enhance the
contribution of the EEG to the diagnostic process and may also
serve as an educational experience for all involved. Second, it is
possible to identify a number of clinical situations where the EEG
could contribute to solving a real clinical problem: (i) differenti-
ating between FTLD and early onset AD. In relative young
patients the clinical spectrum of both disorders can overlap con-
siderably, AD patients presenting with behavioral problems, and
FTLD patients also having cognitive complaints. In our view, a
clearly abnormal EEG favors a diagnosis of early onset AD and
argues strongly against FTLD (174); (ii) differentiating between
AD and DLB. Again, the differential diagnosis can be difficult in
some patients and may have consequences for treatment (for
instance, avoidance of neuroleptics in DLB). The EEG is likely to
be more abnormal in DLB than in AD and may show specific
abnormalities such as FIRDA much more often in DLB than in
AD (78); (iii) and differential diagnosis of delirium. Here the
EEG can differentiate between organic and psychiatric condi-
tions, suggest metabolic encephalopathy or withdrawal syn-
dromes, and rule out nonconvulsive status epilepticus; (iv)
epilepsy-related cognitive disturbances. Temporal lobe epilepsy
can be associated with cognitive disturbances, in particular mem-
ory complaints, and can mimic AD (80–82); (v) diagnosis of
Creutzfeldt–Jakob disease according to WHO standards (211).
REFERENCES
1. Berger H. Uber das Elektrenkephalogramm des Menschen. Dritte
Miteilung. Arch Psychiatr Nervenkr. 1931;94:16–60.
2. Berger H. Uber das Elektrenkephalogramm des Menschen. Funfte
Miteilung. Arch Psychiatr Nervenkr. 1932;98:231–254.
3. American Psychiatric Association. Diagnostic and Statistical
Manual of Mental Sisorders. 4th ed. (DSM IV). Washington, DC:
APA; 1994.
4. Jeong J. EEG dynamics in patients with Alzheimer’s disease. Clin
Neurophysiol. 2004;115:1490–1505.
5. Waldemar G, Dubois B, Emre M, et al. Recommendations for the
diagnosis and management of Alzheimer’s disease and other dis-
orders associated with dementia: EFNS guideline. Eur J Neurol.
2007;14:e1–e26.
6. Smith SJM. EEG in neurological conditions other than epilepsy:
when does it help, what does it add? J Neurol Neurosurg Psychiatry.
2005;76(suppl II):ii8–ii12.
7. Knopman DS, DeKosky ST, Cummings JL, et al. Practice parame-
ter: diagnosis of dementia (an evidence-based review). Report of
the Quality Standards Subcommittee of the American Academy of
Neurology. Neurology. 2001;56:1143–1153.
388 Part III ■ Clinical EEG: General
8. Jonkman EJ. The role of the electroencephalogram in the diagno-
sis of dementia of the Alzheimer type: an attempt at technology
assessment. Neurophysiol Clin. 1997;27:211–219.
9. Rosen I. Electroencephalography as a diagnostic tool in dementia.
Dement Geriatr Cogn Disord. 1997;8:110–116.
10. Boerman RH, Scheltens P, Weinstein HC. Clinical neurophysiol-
ogy in the diagnosis of Alzheimers disease. Clin Neurol Neurosurg.
1994;96:111–118.
11. Adamis D, Sahu S, Treloar A. The utility of EEG in dementia: a
clinical perspective. Int J Geriatr Psychiatry. 2005;20:1038–1045.
12. Van der Flier WM, Scheltens P. Use of laboratory and imaging
investigations in dementia. J Neurol Neurosurg Psychiatry.
2005;76(suppl V):v45–v52.
13. Jackson CE, Snyder PJ. Electroencephalography and event-related
potentials as biomarkers of mild cognitive impairment and mild
Alzheimer’s disease. Alzheimers Dement. 2008;4(1 suppl 1):
S137–S143.
14. Klass DW, Brenner RP. Electroencephalography of the elderly.
J Clin Neurophysiol. 1995;12:116–131.
15. Rossini PM, Rossi S, Babiloni C, et al. Clinical neurophysiology of
aging brain: from normal aging to neurodegeneration. Prog
Neurobiol. 2007;83:375–400.
16. Hartikainen P, Soininen H, Partanen J, et al. Aging and spectral
analysis of EEG in normal subjects: a link to memory and CSF
AChE. Acta Neurol Scand. 1992;86:148–155.
17. Obrist WD. The electroencephalogram in normal aged adults.
Electroenceph Clin Neurophysiol. 1954;6:235–244.
18. Hubbard O, Sunde D, Goldensohn ES. The EEG in centenarians.
Electroenceph Clin Neurophysiol. 1976;40:407–417.
19. Aird RB, Gastaut Y. Occipital and posterior electroencephalographic
rhythms. Electroenceph Clin Neurophysiol. 1959;11:637–656.
20. Markand ON. Alpha rhythms. J Clin Neurophysiol. 1990;7:
163–189.
21. Babiloni C, Frisoni G, Steriade M, et al. Frontal white matter vol-
ume and delta EEG sources negatively correlate in awake subjects
with mild cognitive impairment and Alzheimer’s disease. Clin
Neurophysiol. 2006;117:1113–1129.
22. Klimesch W. EEG alpha and theta oscillations reflect cognitive
and memory performance: a review and analysis. Brain Res Rev.
1999;29:169–195.
23. Dustman RE, LaMarche JA, Cohn NB, et al. Power spectral analy-
sis and cortical coupling of EEG for young and old normal adults.
Neurobiol Aging. 1985;6:193–198.
24. Schnell RG, Klass DW. Further observation son the Rolandic
arceau rhythm. Electroenceph Clin Neurophysiol. 1966;20:95.
25. Giaquinto S. Nolfe G. The EEG in the normal elderly: a contribu-
tion to the interpretation of aging and dementia. Electroenceph
Clin Neurophysiol. 1986;63:540–546.
26. Oken BS, Kaye JA. Electrophysiologic function in the healthy,
extremely old. Neurology. 1992;42:519–526.
27. Silverman AJ, Busse EW, Barnes RH. Studies in the processes of
aging: electroencephalographic findings in 400 elderly subjects.
Electroenceph Clin Neurophysiol. 1955;7:67–74.
28. Visser SL, Hooijer C, Jonker C, et al. Anterior temporal focal
abnormalities in EEG in normal aged subjects; correlations with
psychopathological and CT brain scan findings. Electroenceph
Clin Neurophysiol. 1987;66:1–7.
29. Cook IA, Leuchter AF, Morgan ML, et al. Cognitive and physio-
logic correlates of subclinical structural brain disease in elderly
healthy control subjects. Arch Neurol. 2002;59:1612–1620.
30. Liedorp M, van der Flier WM, Hoogervorst EL, et al. Associations
between patterns of EEG abnormalities and diagnosis in a large
memory clinic cohort. Dement Geriatr Cogn Disord. 2008;27:
18–23.
31. Katz RI, Horowitz GR. Electroencephalogram in the septuagenar-
ian: studies in a normal geriatric population. J Am Geriatr Soc.
1982;30:273–275.
32. Westmoreland BF, Klass DW. A distinctive rhythmic EEG dis-
charge of adults. Electroenceph Clin Neurophysiol. 1981;51:
186–191.
33. Petersen RC, Smith GE, Waring SC, et al. Mild cognitive impair-
ment: clinical characterization and outcome. Arch Neurol.
1999;56:303–308.
34. Grunwald M, Busse F, Hensel A, et al. Correlation between corti-
cal theta activity and hippocampal volumes in health, mild cogni-
tive impairment and mild dementia. J Clin Neurophysiol. 2001;18:
178–184.
35. Grunwald M, Bussse F, Hensel A, et al. Theta-power differences in
patients with mild cognitive impairment under rest condition
and during haptic tasks. Alzheimer Dis Assoc Disord. 2002;16:
40–48.
36. Jelic V, Shigeta M, Julin P, et al. Quantitative electroencephalogra-
phy power and coherence in Alzheimer’s disease and mild cogni-
tive impairment. Dementia. 1996;7:314–323.
37. Zappoli R, Versari A, Paganini M, et al. Brain electrical activity
(quantitative EEG and bit-mapping neurocognitive CNF compo-
nents), psychometrics and clinical findings in presenile subjects
with initial mild cognitive decline or probable Alzheimer-type
dementia. Ital J Neurol Sci. 1995;16:341–376.
38. Grunwald M, Hensel A, Wolf H, et al. Does the hippocampal atro-
phy correlate with the cortical theta power in elderly subjects with
a range of cognitive impairment? J Clin Neurophysiol. 2007;
24:22–26.
39. Jiang ZY. Study on EEG power and coherence in patients with
mild cognitive impairment during working memory task.
J Zhejiang Univ Sci B. 2005;6:1213–1219.
40. Babiloni C, Binetti G, Cassetta E, et al. Sources of cortical rhythms
change as a function of cognitive impairment in pathological
aging: a multicenter study. Clin Neurophysiol. 2006;117:252–268.
41. Babiloni C, Frisoni GB, Pievani M, et al. White-matter vascular
lesions correlate with alpha EEG sources in mild cognitive
impairment. Neuropsychologia. 2008;46:1707–1720.
42. Lee L, Harrison LM, Mechelli A. A report of the functional con-
nectivity workshop, Dusseldorf 2002. NeuroImage. 2003;19:
457–465.
43. Pijnenburg YA, v d Made Y, van Cappellen van Walsum AM, et al.
EEG synchronization likelihood in mild cognitive impairment
and Alzheimer’s disease during a working memory task. Clin
Neurophysiol. 2004;115:1332–1339.
44. van der Hiele K, Vein AA, Kramer CG, et al. Memory activation
enhances EEG abnormality in mild cognitive impairment.
Neurobiol Aging. 2007;28:85–90.
45. Stam CJ, van der Made Y, Pijnenburg YAL, et al. EEG synchroniza-
tion in mild cognitive impairment and Alzheimer’s disease. Acta
Neurol Scand. 2003;108:90–96.
46. Babiloni C, Ferri R, Binetti G, et al. Fronto-parietal coupling of
brain rhythms in mild cognitive impairment: a multicentric EEG
study. Brain Res Bull. 2006;69:63–73.
47. Huang C, Wahlund L, Dierks T, et al. Discrimination of
Alzheimer’s disease and mild cognitive impairment by equivalent
EEG sources: a cross-sectional and longitudinal study. Clin
Neurophysiol. 2000;111:1961–1967.
48. Jelic V, Johansson S-E, Almkvist O, et al. Quantitative electroen-
cephalography in mild cognitive impairment: longitudinal

changes and possible prediction of Alzheimer’s disease. Neurobiol
Aging. 2000;21:533–540.
49. Missonnier P, Gold G, Herrmann FR, et al. Decreased theta event-
related synchronization during working memory activation is
associated with progressive mild cognitive impairment. Dement
Geriatr Cogn Disord. 2006;22:250–259.
50. Rossini PM, Del Percio C, Pasqualetti P, et al. Conversion from
mild cognitive impairment to Alzheimer’s disease is predicted by
sources and coherence of brain electroencephalography rhythms.
Neuroscience. 2006;143:793–803.
51. Luckhaus C, Grass-Kapanke B, Blaeser I, et al. Quantitative EEG
in progressing vs stable mild cognitive impairment (MCI): results
of a 1-year follow-up study. Int J Geriatr Psychiatry. 2008;
23:1148–1155.
52. van der Hiele K, Bollen EL, Vein AA, et al. EEG markers of future
cognitive performance in the elderly. J Clin Neurophysiol.
2008;25:83–89.
53. Prichep LS, John ER, Ferris SH, et al. Prediction of longitudinal
cognitive decline in normal elderly with subjective complaints using
electrophysiological imaging. Neurobiol Aging. 2006;27:471–481.
54. Alzheimer A. Über eine eigenartige Erkrankung der Hirnrinde.
Allgemeine Zeitschrift für Psychiatrie. 1907;64:146–148.
55. Braak H, Braak E. Neuropathological staging of Alzheimer-
related changes. Acta Neuropathol. 1991;82:239–259.
56. Selkoe DJ. Cell biology of the amyloid beta-protein precursor and
the mechanism of Alzheimer’s disease. Annu Rev cell Biol.
1994;10:373–403.
57. Klunk WE, Engler H, Nordberg A, et al. Imaging brain amyloid in
Alzheimer’s disease with Pittsburgh Compound-B. Ann Neurol.
2004;55:306–319.
58. Shoghi-Jadid K, Small GW, Agdeppa ED, et al. Localization of
neurofibrillary tangles and beta-amyloid plaques in the brains of
living patients with Alzheimer disease. Am J Geriatr Psychiatry.
2002;10:24–35.
59. Bracco L, Gallato R, Grigoletto F, et al. Factors affecting course
and survival in Alzheimer’s disease. A 9-year longitudinal study.
Arch Neurol. 1994;51:1213–1219.
60. Farrer LZ, Cupples LA, Blackburn S, et al. Interrater agreement
for diagnosis of Alzheimer’s disease. The mirage study. Neurology.
1994;44:652–256.
61. McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of
Alzheimer’s disease: report of the NINCDS-ADRDA Work Group
under the auspices of Department of Health and Human Services
Task Force on Alzheimer’s disease. Neurology. 1984;34:939–944.
62. Paulus W, Bancher C, Jellinger K. Interrater reliability in the
neuropathologic diagnosis of Alzheimer’s disease. Neurology.
1992;42:329–332.
63. Babiloni C, Binetti G, Cassetta E, et al. Mapping distributed
sources of cortical rhythms in mild Alzheimer’s disease. A multi-
centric EEG study. Neuroimage. 2004;22:57–67.
64. Czigler B, Csikós D, Hidasi Z, et al. Quantitative EEG in early
Alzheimer’s disease patients—power spectrum and complexity
features. Int J Psychophysiol. 2008;68:75–80.
65. Gianotti LR, Künig G, Lehmann D, et al. Correlation between dis-
ease severity and brain electric LORETA tomography in
Alzheimer’s disease. Clin Neurophysiol. 2007;118:186–196.
66. Penttilä M, Partanen JV, Soininen H, et al. Quantitative analysis of
occipital EEG in different stages of Alzheimer’s disease.
Electroenceph Clin Neurophysiol. 1985;60:1–6.
67. Saletu B, Anderer P, Paulus E, et al. EEG brain mapping in diag-
nostic and therapeutic assessment of dementia. Alzheimer Dis
Assoc Disord. 1991;5(suppl 1):S57–S75.
Chapter 20 ■ Dementia and EEG 389
68. van der Hiele K, Vein AA, Reijntjes RH, et al. EEG correlates in the
spectrum of cognitive decline. Clin Neurophysiol. 2007;118:
1931–1939.
69. Adler G, Brassen S, Jajcevic A. EEG coherence in Alzheimer’s
dementia. J Neural Transm. 2003;110:1051–1058.
70. Coben LA, Danziger WL, Berg L. Frequency analysis of the rest-
ing awake EEG in mild senile dementia of Alzheimer type.
Electroencephalogr Clin Neurophysiol. 1983;55:372–380.
71. Schreiter-Gasser U, Gasser T, Ziegler P. Quantitative EEG analysis
in early onset Alzheimer’s disease: a controlled study.
Electroenceph Clin Neurophysiol. 1993;86:15–22.
72. Schreiter-Gasser U, Gasser T, Ziegler P. Quantitative EEG analysis
in early onset Alzheimer’s disease: correlation with severity, clin-
ical characteristics, visual EEG and CCT. Electroenceph Clin
Neurophysiol. 1994;90:267–272.
73. Rice DM, Buchsbaum MS, Starr A, et al. Abnormal slow activity
in left temporal areas in senile dementia of the Alzheimer type. J
Gerontol Med Sci. 1990;45:M145–M151.
74. Dierks T, Ihl R, Frölich L, et al. Dementia of the Alzheimer type:
effects on the spontaneous EEG described by dipole sources.
Psychiatry Res. 1993;50:151–162.
75. Whitham EM, Pope KJ, Fitzgibbon SP, et al. Scalp electrical
recording during paralysis: quantitative evidence that EEG fre-
quencies above 20 Hz are contaminated by EMG. Clin
Neurophysiol. 2007;118:1877–1888.
76. van Deursen JA, Vuurman EF, Verhey FR, et al. Increased EEG
gamma band activity in Alzheimer’s disease and mild cognitive
impairment. J Neural Transm. 2008;115:1301–1311.
77. Coben LA, Danziger W, Storandt M. A longitudinal EEG study of
mild senile dementia of Alzheimer type: changes at 1 year and at
2.5 years. Electroenceph Clin Neurophysiol. 1985;61:101–112.
78. Roks G, Korf ESC, van der Flier WM, et al. The use of EEG in the
diagnosis of dementia with Lewy bodies. J Neurol Neurosurg
Psychiatry. 2008;79:377–380.
79. Amatniek JC, Hauser WA, DelCastillo-Castaneda C, et al.
Incidence and predictors of seizures in patients with Alzheimer's
disease. Epilepsia. 2006;47:867–872.
80. Hogh P, Smith SJ, Scahill RI, et al. Epilepsy presenting as AD: neu-
roimaging, electrical features, and response to treatment.
Neurology. 2002;58:298–301.
81. Tombini M, Koch G, Placidi F, et al. Temporal lobe epileptic activ-
ity mimicking dementia: a case report. Eur J Neurol. 2005;12:
805–806.
82. Zeman AZ, Boniface SJ, Hodges JR. Transient epileptic amnesia: a
description of the clinical and neuropsychological features in 10
cases and a review of the literature. J Neurol Neurosurg Psychiat.
1998;64:435–443.
83. Cook IA, Leuchter AF. Synaptic dysfunction in Alzheimer’s dis-
ease: clinical assessment using quantitative EEG. Behav Brain Res.
1996;78:15–23.
84. Besthorn C, Forstl H, Geiger-Kabisch C, et al. EEG coherence in
Alzheimer disease. Electroenceph Clin Neurophysiol. 1994;90:242–245.
85. Dunkin JJ, Leuchter AF, Newton TF, et al. Reduced EEG coherence
in dementia: state or trait marker? Biol Psychiatry. 1994;35:
870–879.
86. Leuchter AF, Spar JE, Walter DO, et al. Electroencephalographic
spectra and coherence in the diagnosis of Alzheimer’s-type and
multi-infarct dementia. A pilot study. Arch Gen Psychiatry.
1987;44:993–998.
87. Leuchter AF, Newton TF, Cook AA, et al. Changes in brain func-
tional connectivity in Alzheimer-type and multi-infarct demen-
tia. Brain. 1992;115:1543–1561.
390 Part III ■ Clinical EEG: General
88. Locatelli T, Cursi M, Liberati D, et al. EEG coherence in
Alzheimer’s disease. Electroenceph Clin Neurophysiol. 1998;
106:229–237.
89. Pogarell O, Teipel SJ, Juckel G, et al. EEG coherence reflects
regional corpus callosum area in Alzheimer’s disease. J Neurol
Neurosurg Psychiatry. 2005;76:109–111.
90. Sloan EP, Fenton GW, Kennedy NS, et al. Neurophysiology and
SPECT cerebral blood flow patterns in dementia. Electroenceph
Clin Neurophysiol. 1994;91:163–170.
91. Park YM, Che HJ, Im CH, et al. Decreased EEG synchronization
and its correlation with symptom severity in Alzheimer’s disease.
Neurosci Res. 2008;62:112–117.
92. Knott V, Mohr E, Mahoney C, et al. Electroencephalographic coher-
ence in Alzheimer’s disease: comparisons with a control group and
population norms. J Geriatr Psychiatry Neurol. 2000;13:1–8.
93. Babiloni C, Ferri R, Binetti G, et al. Directionality of EEG syn-
chronization in Alzheimer’s disease subjects. Neurobiol Aging.
2009;30:93–102.
94. Babiloni C, Ferri R, Moretti DV, et al. Abnormal fronto-parietal
coupling of brain rhythms in mild Alzheimer’s disease: a multi-
centric EEG study. Eur J Neurosci. 2004;19:2583–2590.
95. Jeong J, Gore JC, Peterson BS. Mutual information analysis of the
EEG in patients with Alzheimer’s disease. Clin Neurophysiol.
2001;112:827–835.
96. Anderer P, Saletu B, Kloppel B, et al. Discrimination between
demented patients and normals based upon topographic EEG
slow wave activity: comparison between z statistics, discriminant
analysis and artificial neural network classifiers. Electroenceph
Clin Neurophysiol. 1994;91;108–117.
97. Pritchard WS, Duke DW, Coburn KL, et al. EEG-based, neural-
net predictive classification of Alzheimer’s disease versus control
subjects is augmented by non-linear EEG measures. Electroenceph
Clin Neurophysiol. 1994;91:118–130.
98. Rossini PM, Buscema M, Capriotti M, et al. Is it possible to auto-
matically distinguish resting EEG data of normal elderly vs. mild
cognitive impairment subjects with high degree of accuracy? Clin
Neurophysiol. 2008;119:1534–1545.
99. Stam CJ. Nonlinear dynamical analysis of EEG and MEG: review
of an emerging field. Clin Neurophysiol. 2005;116:2266–2301.
100. Stam CJ. Nonlinear Brain Dynamics. New York: Nova Science
Publishers; 2006.
101. Jeong J. Nonlinear dynamics of EEG in Alzheimer’s disease. Drug
Dev Res. 2002;56:57–66.
102. Besthorn C, Sattel H, Geiger-Kabisch C, et al. Parameters of EEG
dimensional complexity in Alzheimer’s disease. Electroenceph
Clin Neurophysiol. 1995;95:84–89.
103. Hornero R, Abásolo D, Escudero J, et al. Nonlinear analysis of
electroencephalogram and magnetoencephalogram recordings in
patients with Alzheimer’s disease. Philos Transact A Math Phys
Eng Sci. 2009;367:317–336.
104. Jelles B, Strijers RLM, Hooijer Ch, et al. Nonlinear EEG analysis
in early Alzheimer’s disease. Acta Neurologica Scandinavica.
1999;100:360–368.
105. Jeong J, Kim SY, Han SH. Non-linear dynamical analysis of the
EEG in Alzheimer’s disease with optimal embedding dimension.
Electroenceph Clin Neurophysiol. 1998;106:220–228.
106. Pritchard WS, Duke DW, Coburn KL. Altered EEG dynamical
responsivity associated with normal aging and probable
Alzheimer’s disease. Dementia. 1991;2:102–105.
107. Stam CJ, Tavy DLJ, Jelles B, et al. Non-linear dynamical analysis
of multi channel EEG data: clinical applications in dementia and
Parkinson's disease. Brain Topogr. 1994;7:141–150.
108. Stam CJ, Jelles B, Achtereekte HAM, et al. Investigation of EEG
non-linearity in dementia and Parkinson’s disease. Electroenceph
Clin Neurophysiol. 1995;95:309–317.
109. Stam CJ, Jelles B, Achtereekte HAM, et al. Diagnostic usefulness
of linear and nonlinear quantitative EEG analysis in Alzheimer’s
disease. Clin Electroenceph. 1996;27:69–77.
110. Woyshville MJ, Calabrese JR. Quantification of occipital EEG
changes in Alzheimer’s disease utilizing a new metric: the fractal
dimension. Biol Psychiatry. 1994;35:381–387.
111. Yagyu T, Wackermann J, Shigeta M, et al. Global dimensional
complexity of multichannel EEG in mild Alzheimer’s disease
and age-matched controls. Dement Geriatr Cogn Disord.
1997;8:343–347.
112. Jelles B, Jonkman EJ, Birgelen JH van, et al. Decrease of non-lin-
ear structure in the EEG of Alzheimer patients compared to
controls. Clin Neurophysiol. 1999;110:1159–1167.
113. Jelles B, Scheltens P, van der Flier WM, et al. Global dynamical
analysis of the EEG in Alzheimer’s disease: frequency-specific
changes of functional interactions. Clin Neurophysiol.
2008;119:837–841.
114. Van Cappellen van Walsum A-M, Pijnenburg YAL, Berendse
HW, et al. A neural complexity measure applied to MEG data in
Alzheimer’s disease. Clin Neurophysiol. 2003;114:1034–1040.
115. Reijneveld JC, Ponten SC, Berendse HW, et al. The application
of graph theoretical analysis to complex networks in the brain.
Clin Neurophysiol. 2007;118:2317–2331.
116. Stam CJ, Reijneveld JC. Graph theoretical analysis of complex
networks in the brain. Nonlinear Biomed Phys. 2007;1:3.
117. He Y, Chen ZJ, Evans AC. Small-world anatomical networks in
the human brain revealed by cortical thickness from MRI. Cereb
Cortex. 2007;17:2407–2419.
118. Stam CJ, Jones BF, Nolte G, et al. Small-world networks and
functional connectivity in Alzheimer’s disease. Cereb Cortex.
2007;17:92–99.
119. Stam CJ, de Haan W, Daffertshofer A, et al. Graph theoretical analy-
sis of magnetoencephalographic functional connectivity in
Alzheimer’s disease. Brain. 2008, October 24. [Epub ahead of print]
120. Supekar K, Menon V, Rubin D, et al. Network analysis of intrin-
sic functional brain connectivity in Alzheimer’s disease. PLoS
Comput Biol. 2008;4:e1000100.
121. Petit D, Montplaisir J, Lorrain D, et al. Spectral analysis of the
rapid eye movement sleep EEG in right and left temporal reigns;
a biological marker for diagnosis of probable Alzheimer’s dis-
ease. Neurology. 1992;42:765–770.
122. Prinz PN, Larsen LH, Moe KE, et al. EEG markers of early
Alzheimer’s disease in computer selected tonic REM sleep.
Electroenceph Clin Neurophysiol. 1992;83:36–43.
123. Moe KE, Larsen LH, Prinz PN, et al. Major unipolar depression
and mild Alzheimer’s disease: differentiation by quantitative
tonic REM EEG. Electroenceph Clin Neurophysiol. 1993;86:
238–246.
124. Robinson DJ, Merskey H, Blume WT, et al.
Electroencephalography as an aid in the exclusion of
Alzheimer’s disease. Arch Neurol. 1994;51:280–284.
125. Strijers RLM, Scheltens Ph, Jonkman EJ, et al. Diagnosing
Alzheimer’s disease in community-dwelling elderly: a compari-
son of EEG and MRI. Dement Geratr Cogn Disord. 1997;8:
198–202.
126. Scheltens P, Leys D, Barkhof F, et al. Atrophy of medial temporal
lobes on MRI in “probable” Alzheimer’s disease and normal age-
ing: diagnostic value and neuropsychological correlates. J Neurol
Neurosurg Psychiatry. 1992;55:967–972.

127. Claus JJ, Strijers RLM, Jonkman EJ, et al. The diagnostic value of
electroencephalography in mild senile Alzheimer’s disease. Clin
Neurophysiol. 1999;110:825–832.
128. Brenner R, Ulrich RF, Spiker DG, et al. Computerized EEG spec-
tral analysis in elderly normal, demented and depressed sub-
jects. Electroenceph Clin Neurophysiol. 1986;64:483–492.
129. Brenner RP, Reynolds III ChF, Ulrich RF. Diagnostic efficacy of
computerized spectral versus visual EEG analysis in elderly nor-
mal, demented and depressed subjects. Electroenceph Clin
Neurophysiol. 1988;69:110–117.
130. Hooijer Ch, Jonker C, Posthuma J, et al. Reliability, validity and
follow-up of the EEG in senile dementia: sequelae of sequential
measurement. Electroenceph Clin Neurophysiol. 1990;76:
400–412.
131. Babiloni C, Frisoni GB, Pievani M, et al. Hippocampal volume
and cortical sources of EEG alpha rhythms in mild cognitive
impairment and Alzheimer disease. Neuroimage. 2009;44:
123–135.
132. van der Hiele K, Vein AA, van der Welle A, et al. EEG and MRI
correlates of mild cognitive impairment and Alzheimer’s dis-
ease. Neurobiol Aging. 2007;28:1322–1329.
133. Dunkin JJ, Osato S, Leuchter A. Relationships between EEG
coherence and neuropsychological tests in dementia. Clin
Electroenceph. 1995;26:47–59.
134. Lopez OL, Brenner RP, Becker JT, et al. EEG spectral abnormal-
ities and psychosis as predictors of cognitive and functional
decline in probable Alzheimer’s disease. Neurology. 1997;48:
1521–1525.
135. Helkala EL, Laulumaa V, Soininen H, et al. Different patterns of
cognitive decline related to normal or deteriorating EEG in a 3-
year follow-up study of patients with Alzheimer’s disease.
Neurology. 1991;41:528–532.
136. Claus JJ, Gool WA van, Teunisse S, et al. Predicting survival in
patients with early Alzheimer’s disease. Dement Geriatr Cogn
Disord. 1998;9:283–293.
137. Claus JJ, Ongerboer de Visser BW, Walstra GJM, et al.
Quantitative spectral electroencephalography in predicting sur-
vival in patients with early Alzheimer disease. Arch Neurol.
1998;55:1105–1111.
138. Nobili F, Copello F, Vitali P, et al. Timing of disease progression
by quantitative EEG in Alzheimer’s patients. J Clin Neurophysiol.
1999;16:566–573.
139. Strittmatter WJ, Roses AD. Apolipoprotein E and Alzheimer’s
disease. Proc Natl Acad Sci USA. 1995;92:4725–4727.
140. Lehtovirta M, Soininen H, Helisalmi S, et al. Clinical and neu-
ropsychological characteristics in familial sporadic Alzheimer’s
disease: relation to apolipoprotein E polymorphism. Neurology.
1996;46:413–419.
141. Lehtovirta M, Partanen J, Kononen M, et al. Spectral analysis of
EEG in Alzheimer’s disease: relation to Apolipoprotein E poly-
morphism. Neurobiol Ageing. 1996;4:523–526.
142. Lehtovirta M, Partanen J, Kononen M, et al. A longitudinal
quantitative EEG study of Alzheimer’s disease: relation to
apolipoprotein E polymorphism. Dem Geriatr Cogn Disord.
2000;11:29–35.
143. Jelic V, Julin P, Shigeta M, et al. Apolipoprotein E H4 allele
decreases functional connectivity in Alzheimer’s disease as
measured by EEG coherence. J Neurol Neurosurg Psychiatry.
1997;63:59–65.
144. Kramer G, van der Flier WM, de Langen C, et al. EEG functional
connectivity and ApoE genotype in Alzheimer’s disease and
controls. Clin Neurophysiol. 2008;119:2727–2732.
Chapter 20 ■ Dementia and EEG 391
145. Ponomareva NV, Korovaitseva GI, Rogaev EI. EEG alterations in
non-demented individuals related to apolipoprotein E genotype
and to risk of Alzheimer disease. Neurobiol Aging. 2008;29:
819–827.
146. Coyle JT, Price DL, DeLong MR. Alzheimer’s disease: a disorder
of cortical cholinergic innervation. Science. 1983;219:
1181–1190.
147. Collerton D. Cholinergic function and intellectual decline in
Alzheimer’s disease. Neuroscience. 1986;19:1–28.
148. Perry EK, Perry RH, Blessed G, et al. Necropsy evidence of cen-
tral cholinergic deficits in senile dementia. Lancet. 1977;1:189.
149. Buzsaki G, Bickford RG, Ponomareff G, et al. Nucleus basalis
and thalamic control of neocortical activity in the freely moving
rat. J Neurosci. 1988;8:4007–4026.
150. Villa AEP, Tetko IV, Dutoit P, et al. Non-linear cortico-cortical
interactions modulated by cholinergic afferences from the rat
basal forebrain. BioSystems. 2000;58:219–228.
151. Kikuchi M, Wada Y, Koshino Y, et al. Effects of scopolamine on
interhemispheric EEG coherence in healthy subjects: analysis
during rest and photic stimulation. Clin Electroencephalogr.
2000;31:109–115.
152. Ebert U, Grossmann M, Oertel R, et al. Pharmacokinetic-phar-
macodynamic modeling of the electroencephalogram effects of
scopolamine in healthy volunteers. J Clin Pharmacol. 2001;41:
51–60.
153. Weingartner H. Models of memory dysfunctions. Ann N Y Acad
Sci. 1985;444:359–369.
154. Riekkinen P, Buzsaki G, Riekkinen P Jr, et al. The cholinergic
system and EEG slow waves. Electroenceph Clin Neurophysiol.
1991;78:89–96.
155. Agnoli A, Martucci N, Manna V, et al. Effect of cholinergic and
anticholinergic drugs on short-term memory in Alzheimer’s
dementia: a neuropsychological and computerized electroen-
cephalographic study. Clin Neuropharmacol. 1983;6:311–323.
156. Babiloni C, Cassetta E, Dal Forno G, et al. Donepezil effects on
sources of cortical rhythms in mild Alzheimer’s disease:
Responders vs. Non-Responders. Neuroimage. 2006;31:
1650–1665.
157. Gianotti LR, Künig G, Faber PL, et al. Rivastigmine effects on
EEG spectra and three-dimensional LORETA functional imag-
ing in Alzheimer’s disease. Psychopharmacology (Berl). 2008;
198:323–332.
158. Rodriguez G, Vitali P, De Leo C, et al. Quantitative EEG changes
in Alzheimer patients during long-term Donepezil therapy.
Neuropsychobiology. 2002;46:49–56.
159. Adler G, Brassen S, Chwalek K, et al. Prediction of treatment
response to rivastigmine in Alzheimer’s dementia. J Neurol
Neurosurg Psychiat. 2004;75:292–294.
160. Almkvist O, Jelic V, Hellstrom-Lindahl E, et al. Responder char-
acteristics to a single oral dose of cholinesterase inhibitor: a dou-
ble-blind placebo-controlled study with Tacrine in Alzheimer
patients. Dement Geriatr Cogn Disord. 2001;12:22–32.
161. Knott V, Mohr E, Mahoney C, et al. Pharmaco-EEG test dose
response predicts cholinesterase inhibitor outcome in Alzheimer’s
disease. Methods Find Exp Clin Pharmacol. 2000;22:115–122.
162. Dringenberg HC. Alzheimer’s disease: more than a ‘cholinergic
disorder’—evidence that cholinergic—monoaminergic interac-
tions contribute to EEG slowing and dementia. Behav Brain Res.
2000;115:235–249.
163. Di Lazzaro V, Oliviero A, Pilato F, et al. Motor cortex hyperex-
citability to transcranial magnetic stimulation in Alzheimer’s
disease. J Neurol Neurosurg Psychiatry. 2004;75:555–559.
392 Part III ■ Clinical EEG: General
164. Davis KL, Mohs RC, Marin D, et al. Cholinergic markers in eld-
erly patients with early signs of Alzheimer’s disease. J Am Med
Assoc. 1999;45:178–179.
165. Visser FE, Kuilman M, Oosting J, et al. Use of electroencephalog-
raphy to detect Alzheimer’s disease in Down’s syndrome. Acta
Neurol Scand. 1996;94:97–103.
166. Jelic V, Blomberg M, Dierks T, et al. EEG slowing and cere-
brospinal fluid tau levels in patients with cognitive decline.
Neuroreport. 1988;9:157–160.
167. Babiloni C, Bosco P, Ghidoni R, et al. Homocysteine and elec-
troencephalographic rhythms in Alzheimer disease: a multicen-
tric study. Neuroscience. 2007;145:942–954.
168. Schreiter-Gasser U, Rousson V, Hentschel F, et al. Alzheimer dis-
ease versus mixed dementias: an EEG perspective. Clin
Neurophysiol. 2008;119:2255–2259.
169. Johannesson G, Brun A, Gustafson I, et al. EEG in presenile
dementia related to cerebral blood flow and autopsy findings.
Acta Neurol Scand. 1977;56:89–103.
170. Johannesson G, Hagberg B, Gustafson I, et al. EEG and cognitive
impairment in presenile dementia. Acta Neurol Scand.
1979;59:225–240.
171. Rosén I, Gustafson L, Risberg J. Multichannel EEG frequency
analysis and somatosensory-evoked potentials in patients with
different types of organic dementia. Dementia. 1993;4:43–49.
172. Stigsby B, Johannesson G, Ingvar DH. Regional EEG analysis
and regional cerebral blood flow in Alzheimer’s disease and
Pick’s disease. Electroenceph Clin Neurophysiol. 1981;51:
537–547.
173. Forstl H, Besthorn C, Hentschel F, et al. Frontal lobe degenera-
tion and Alzheimer’s disease: a controlled study on clinical
findings, volumetric brain changes and quantitative electroen-
cephalography data. Dementia. 1996;7:27–34.
174. Pijnenburg YA, Strijers RL, Made YV, et al. Investigation of rest-
ing-state EEG functional connectivity in frontotemporal lobar
degeneration. Clin Neurophysiol. 2008;119:1732–1738.
175. Yener GG, Leuchter AF, Jenden D, et al. Quantitative EEG in
frontotemporal dementia. Clin Electroencephalogr. 1996;27:
61–68.
176. Chan D, Walters RJ, Sampson EL, et al. EEG abnormalities in
frontotemporal lobar degeneration. Neurology. 2004;62:
1628–1630.
177. Passant U, Rosén I, Gustafson L, et al. The heterogeneity of fron-
totemporal dementia with regard to initial symptoms, qEEG
and neuropathology. Int J Geriatr Psychiatry. 2005;20:983–988.
178. Neufeld MY, Inzelberg R, Korczyn AD. EEG in demented and
non-demented parkinsonian patients. Acta Neurol Scand.
1988;78:1–5.
179. Caviness JN, Hentz JG, Evidente VG, et al. Both early and
late cognitive dysfunction affects the electroencephalogram
in Parkinson’s disease. Parkinsonism Relat Disord. 2007;13:
348–354.
180. Domitrz I, Friedman A. Electroencephalography of demented
and non-demented Parkinson’s disease patients. Parkinsonism
Relat Disord. 1999;5:37–41.
181. Neufeld MY, Blumen S, Aitkin I, et al. EEG frequency analysis in
demented and nondemented parkinsonian patients. Dementia.
1994;5:23–28.
182. Fogelson N, Kogan E, Korczyn AD, et al. Effects of rivastigmine
on the quantitative EEG in demented Parkinsonian patients.
Acta Neurol Scand. 2003;107:252–255.
183. Crystal HA, Dickson DW, Lizardi JE, et al. Antemortem diagno-
sis of diffuse Lewy body disease. Neurology. 1990;40:1523–1528.
184. Calzetti S, Bortone E, Negrotti A, et al. Frontal intermittent
rhythmic delta activity (FIRDA) in patients with dementia with
Lewy bodies: a diagnostic tool? Neurol Sci. 2002;23(suppl.
2):S65–S66.
185. Barber PA, Varma AR, Lloyd JJ, et al. The electroencephalogram
in dementia with Lewy bodies. Acta Neurol Scand. 2000;
101:53–56.
186. Londos E, Passant U, Brun A, et al. Regional cerebral blood flow
and EEG in clinically diagnosed dementia with Lewy bodies and
Alzheimer’s disease. Arch Gerontol Geriatr. 2003;36:231–245.
187. Bonanni L, Thomas A, Tiraboschi P, et al. EEG comparisons in
early Alzheimer’s disease, dementia with Lewy bodies and
Parkinson’s disease with dementia patients with a 2-year follow-
up. Brain. 2008;131:690–705.
188. Briel RCG, McKeith IG, Barker WA, et al. EEG findings in
dementia with Lewy bodies and Alzheimer’s disease. J Neurol
Neurosurg Psychiatry. 1999;66:401–403.
189. Kai T, Asai Y, Sakuma K, et al. Quantitative electroencephalo-
gram analysis in dementia with Lewy bodies and Alzheimer’s
disease. J Neurol Sci. 2005;237:89–95.
190. Andersson M, Hansson O, Minthon L, et al. Electroencephalogram
variability in dementia with Lewy bodies, Alzheimer’s disease and
controls. Dement Geriatr Cogn Disord. 2008;26:284–290.
191. Scott DF, Heathfield KW, Toone B, et al. The EEG in
Huntington’s chorea: a clinical and neuropathological study.
J Neurol Neurosurg Psychiatry. 1972;35:97–102.
192. Streletz LJ, Reyes PF, Zalewska M, et al. Computer analysis of
EEG activity in dementia of the Alzheimer’s type and
Huntington’s disease. Neurobiol Aging. 1990;11:15–20.
193. van der Hiele K, Jurgens CK, Vein AA, et al. Memory activation
reveals abnormal EEG in preclinical Huntington’s disease. Mov
Disord. 2007;22:690–695.
194. Fowler CJ, Harrison MJ. EEG changes in subcortical dementia: a
study of 22 patients with steele-richardson-olszewski (SRO)
syndrome. Electroencephalogr Clin Neurophysiol. 1986;64:
301–303.
195. Tashiro K, Ogata K, Goto Y, et al. EEG findings in early-stage
corticobasal degeneration and progressive supranuclear palsy: a
retrospective study and literature review. Clin Neurophysiol.
2006;117:2236–2242.
196. Román GC, Tatemichi TK, Erkinjuntti T, et al. Vascular demen-
tia: diagnostic criteria for research studies. Report of the
NINDS-AIREN International Workshop. Neurology. 1993;43:
250–260.
197. Holmes C, Cairns N, Lantos P, et al. Validity of current clinical
criteria for Alzheimer’s disease, vascular dementia and dementia
with Lewy bodies. Br J Psychiatry. 1999;174:45–50.
198. Erkinjuntti T. Differential diagnosis between Alzheimer’s disease
and vascular dementia: evaluation of common clinical methods.
Acta Neurol Scand. 1987;76:433–442.
199. Erkinjuntti T, Larsen T, Sulkava R, et al. EEG in the differential
diagnosis between Alzheimer’s disease and vascular dementia.
Acta Neurol Scand. 1988;77:36–43.
200. Jeong J, Chae J-H, Kim SY, et al. Nonlinear dynamic analysis of
the EEG in patients with Alzheimer’s disease and vascular
dementia. J Clin Neurophysiol. 2001;18:58–67.
201. Moretti DV, Babiloni C, Binetti G, et al. Individual analysis of
EEG frequency and band power in mild Alzheimer’s disease.
Clin Neurophysiol. 2004;115:299–308.
202. Gawel M, Zalewska E, Szmidt-Sałkowska E, et al. Does EEG
(visual and quantitative) reflect mental impairment in subcorti-
cal vascular dementia? J Neurol Sci. 2007;257:11–16.

203. Romano J, Engel GL. Delirium: I. Electroencephalographic data.
Arch Neurol Psychiatry. 1944;51:356–377.
204. Koponen H, Partanen J, Pääkkönen A, et al. EEG spectral analy-
sis in delirium. J Neurol Neurosurg Psychiatry. 1989;52:980–985.
205. Thomas C, Hestermann U, Walther S, et al. Prolonged activation
EEG differentiates dementia with and without delirium in frail
elderly patients. J Neurol Neurosurg Psychiatry. 2008;79:119–125.
206. Thomas C, Hestermann U, Kopitz J, et al. Serum anticholinergic
activity and cerebral cholinergic dysfunction: an EEG study in
frail elderly with and without delirium. BMC Neurosci.
2008;9:86.
207. Boutros NN. A review of indications for routine EEG in clinical
psychiatry. Hosp Community Psych. 1992;43:716–719.
208. Jacobson S, Jerrier H. EEG in delirium. Semin Clin Neuropsychiatry.
2000;5:86–92.
209. Newman SE. The EEG manifestations of chronic ethanol abuse:
relation to cerebral cortical atrophy. Ann Neurol. 1978;3:299–304.
210. World Health Organization. Consensus on criteria for diagnosis
of sporadic CJD. Wkly Epidemiol Rec. 1998;73:361–365.
211. Wieser HG, Schindler K, Zumsteg D. EEG in Creutzfeldt-Jakob
disease. Clin Neurophysiol. 2006;117:935–951.
212. Will RG, Zeidler M, Stewart GE, et al. Diagnosis of new variant
Creutzfeldt-Jakob disease. Ann Neurol. 2000;47:575–582.
213. Binelli S, Agazzi P, Giaccone G, et al. Periodic electroencephalo-
gram complexes in a patient with variant Creutzfeldt-Jakob dis-
ease. Ann Neurol. 2006;59:423–427.
214. Cambier DM, Kantarci K, Worrell GA, et al. Lateralized and
focal clinical, EEG, and FLAIR MRI abnormalities in Ceutzfelt-
Jakob disease. Clin Neurophysiol. 2003;114:1724–1728.
215. Hirsch LJ, Brenner RP, Drislane FW, et al. The ACNS subcom-
mittee on research terminology for continuous EEG monitor-
ing: proposed standardized terminology for rhythmic and
periodic EEG patterns encountered in critically ill patients. J
Clin Neurophysiol. 2005;22:128–135.
216. Hung CI, Wang PS, Soong BW, et al. Blind source separation of
concurrent disease-related patterns from EEG in Creutzfeldt-
Jakob disease for assisting early diagnosis. Ann Biomed Eng.
2007;35:2168–2179.
217. Wang PS, Wu YT, Hung CI, et al. Early detection of periodic
sharp wave complexes on EEG by independent component
analysis in patients with Creutzfeldt-Jakob disease. J Clin
Neurophysiol. 2008;25:25–31.
218. Steinhoff BJ, Racker S, Herrendorf G, et al. Accuracy and relia-
bility of periodic sharp wave complexes in Creutzfeldt-Jakob
disease. Arch Neurol. 1996;53:162–166.
Chapter 20 ■ Dementia and EEG 393
219. Steinhoff BJ, Zerr I, Glatting M, et al. Diagnostic value of peri-
odic complexes in Creutzfeldt-Jakob disease. Ann Neurol.
2004;56:702–708.
220. Zerr I, Pocchiari M, Collins S, et al. Analysis of EEG and CSF
14-3-3 proteins as aids to the diagnosis of Creutzfeldt-Jakob
disease. Neurology. 2000;55:811–815.
221. Zerr I, Bodemer M, Otto M, et al. Diagnosis of Creutzfeldt-
Jakob disease by two-dimensional gel electrophoresis of cere-
brospinal fluid. Lancet. 1996;348:846–849.
222. Van Everbroeck B, Dobbeleir I, De Waele M, et al. Differential
diagnosis of 201 possible Creutzfeldt-Jakob disease patients.
J Neurol. 2004;251;298–304.
223. Kovacs GG, Trabbattoni G, Hainfellner JA, et al. Mutations of
the prion protein gene phenotypic spectrum. J Neurol.
2002;249:1567–1582.
224. Tschampa HJ, Herms JW, Schulz-Schaeffer WJ, et al. Clinical
findings in sporadic Creutzfeldt-Jakob disease correlate with
thalamic pathology. Brain. 2002;125:2558–2566.
225. Stam CJ, van Woerkom TCAM, Keunen, RWM. Non-linear
analysis of the electroencephalogram in Creutzfeldt-Jakob dis-
ease. Biol Cybern. 1997,;77:247–256.
226. Stam CJ, Vliegen JHR, Nicolai J. Investigation of the dynamics
underlying periodic complexes in the EEG. Biol Cybern. 1999;
80:57–69.
227. Radermecker J. Aspects electroencephalographiques dans trois
cas d’encephalite subaigue. Acta Neurol Psychiatr Belg. 1949;49:
222–232.
228. Fariello RJ, Black JA. Pseudoperiodic bilateral EEG paroxysms in
a case of phencyclidine intoxication. J Clin Psychiatry.
1978;39:579–581.
229. Stockard JJ, Werner SS, Aalbers JA, et al. Electroencephalographic
findings in phencyclidine intoxication. Arch Neurol. 1976;33:
200–203.
230. Gabuzda DH, Levy SR, Chiappa KH. Electroencephalography in
AIDS and AIDS-related complex. Clin Electroencephalogr.
1988;19:1–6.
231. Nuwer MR, Miller EN, Visscher BR, et al. Asymptomatic HIV
infection does not cause EEG abnormalities: results from the
Multicenter AIDS Cohort Study (MACS). Neurology. 1992;42:
1214–1219.
232. Harrison MJ, Newman SP, Hall-Craggs MA, et al. Evidence of
CNS impairment in HIV infection: clinical, neuropsychological,
EEG, and MRI/MRS study. J Neurol Neurosurg Psychiatry. 1998;
65:301–307.

Metabolic Disorders and EEG
TRUDY PANG, MEGAN SELVITELLI, DONALD L. SCHOMER AND ERNST NIEDERMEYER
INTRODUCTION AND HISTORICAL
REMARKS
Neuronal electrogenesis is dependent on metabolic homeosta-
sis. Nutritive and energy-providing metabolic systems represent
the fuel for neuronal and glial structures and abnormalities in
their composition may lead to pronounced clinical and elec-
troencephalographic changes. These changes are usually
reversible, but sometimes irreversible progression ensues.
Cerebral metabolic disturbances and/or electroencephalo-
gram (EEG) changes due to blood alkalosis from hyperventila-
tion and anoxia are presented elsewhere. This chapter deals
mainly with alterations of blood sugar levels, failure of the
hepatic and renal organ systems, states of electrolyte imbal-
ances, changes due to endocrine disorders, and certain toxic-
metabolic conditions. Some chronic metabolic disturbances,
such as inborn errors of metabolism, are discussed under the
heading of degenerative disorders.
Berger’s observation of slow EEG activity induced by hypo-
glycemia in schizophrenic patients treated with insulin marked
the beginning of EEG studies in human metabolic disorders
(1). Over the ensuing years, EEG studies in metabolic disorders
remained somewhat in the background when compared with
the activity in the fields of epileptic seizure disorder and struc-
tural cerebral lesions. Most of the metabolic disorders belong to
the domain of internal medicine, a field of understandably lim-
ited interest in EEG investigations.
Major progress was made when the work of Foley, and
Bickford and Butt on hepatic coma showed very pronounced
slowing and formation of triphasic waves (2,3). The contribu-
tions of Condon, Krump, Thiebaut, Cadilhac, and Glaser have
provided further important information on the EEG correlates
of metabolic encephalopathies (4–8).
There has been persistent interest in the EEG of metabolic
disorders. This is probably due to the discovery of renal
encephalopathies after dialysis in chronic renal insufficiency (9).
Furthermore, EEG monitoring techniques are now being used to
study the evolution of hepatic encephalopathy and the require-
ment for and efficacy of hepatic and renal transplantation.
HYPOGLYCEMIA
Cerebral functioning depends on an adequate supply of glu-
cose. Hypoglycemia-induced EEG changes have been shown
experimentally by Moruzzi, Himwich, Gellhorn and Kessler,
and Davis (10–13). It was found that the electrical activity of
CHAPTER
21
the cortex disappears earlier than that of deep structures. Level
of awareness, blood sugar level, and EEG changes do not neces-
sarily parallel each other (12).
Early clinical EEG observations were reported by Berger,
Davis, Hoagland, Gibbs, Engel, Weinland and Weinland, Lafon,
Gibbs and Murray, Hetzel and Niedermeyer, Shagass and
Roswell, and Regan and Browne-Mayers (1,13–22). It was
found that very low blood sugar levels were compatible with a
normal EEG and also with wakefulness (23). Clinical observa-
tions had already shown that impairment of consciousness and
disturbance of central nervous system (CNS) functions were
dependent mainly on the rapid fall of the blood sugar level,
rather than on absolute values (24). Patients with insulin-
dependent diabetes mellitus may be unaware of hypoglycemia.
Their nearly normal glycosylated hemoglobin levels result in
normal cerebral glucose uptake and preserved cerebral metab-
olism (25). Most of the aforementioned work was based on
administration of insulin for therapeutic purposes in schizo-
phrenic patients (Sakel cure).
All of this work shows an extremely impressive degree of
slowing in the EEG. In particular, the EEG demonstrates a
decrease in alpha and an increase in theta and delta frequencies
with increasing hypoglycemia. Most often this is most marked
in the frontal region (26,27). In some patients, hypoglycemia
may also promote epileptiform discharges (28). Sperling found
that induced hypoglycemia in medically refractory complex
partial epilepsy patients may activate or accentuate temporal
lobe slowing and epileptiform discharges (29). In a more con-
trolled manner, induced hypoglycemia in adult cats results in
an initial diffuse slowing of the EEG, followed by occasional
spike waves and then near-isoelectric activity (30). Along this
continuum and commonly seen in diabetic patients, Kirchner
and colleagues demonstrated that induced hypoglycemia
results in a faster onset of seizures in rats exposed to flurothyl,
which corresponded with implanted hippocampal depth elec-
trode recorded discharges (31).
Spontaneous states of hypoglycemia may mimic complex
partial seizures, due to the concomitant impaired neuropsychi-
atric state. However, the prolonged nature of these attacks,
atypical semiology, unresponsiveness to anticonvulsant thera-
pies, and provocation by fasting helps to differentiate between
these possibilities. Wang and fellow researchers recently
demonstrated an initial theta and delta slowing on the EEG, fol-
lowed by sporadic spike and sharp waves during a patient’s typ-
ical attacks. Intravenous glucose administration terminated the
event (32). Prolonged fasting may be required to confirm the
395
396 Part III ■ Clinical EEG: General
diagnosis of an insulinoma, which requires surgical removal to
promote cessation of events (33). Thus, the hypoglycemic
attacks induced by an insulinoma may result initially in
impaired cognition, while increasing hypoglycemia may pro-
mote spike-wave discharges and seizures.
HYPERGLYCEMIA
Hyperglycemic states are much less eloquent and predictable in
their EEG manifestations than are states of hypoglycemia.
Gibbs and colleagues found mixed slow and fast frequencies
and some intermingled spiking in patients with blood sugar
levels exceeding 400 mg/100 mL (15). In advanced diabetic
coma, however, very pronounced slowing may occur, and the
record may become indistinguishable from a hypoglycemic
state (6). According to Cadilhac and colleagues, the EEG abnor-
malities may persist for several days despite effective treatment
(4). Occasionally, epileptogenic lesions may result from coma-
tose states, but such sequelae are more likely to be caused by the
hypoglycemic state rather than hyperglycemic state (34).
A distinction has to be made between ketotic and nonketotic
hyperglycemia. The latter entity was described by DiBenedetto
and colleagues and appears to be rich in neurologic complica-
tions (35). Epileptic seizures of focal character are common in
this condition (36–38), as are, above all, focal motor seizures
with rolandic EEG spike foci and contralateral clonic motions
(39,40). The focal motor seizures may be movement induced
(41,42). Other focal seizures including visual hallucinations
with forced head and eye deviation and aphasia may also rarely
occur (43–45). The hyperglycemic states do not have to be non-
ketotic in order to cause focal seizures, but in such cases, a
ketotic nonacidotic situation is found (36). Correction of the
hyperglycemic state with insulin, rather than treatment with
anticonvulsant medications, controls the epileptic seizures.
LIVER DISEASE (HEPATIC
ENCEPHALOPATHY)
Liver disease may produce variable degrees of EEG abnormali-
ties, from mild slowing to isoelectric activity. The presence or
absence of notable EEG changes is dependent on the patho-
physiologic mechanisms involved. For this reason, some discus-
sion of the pathophysiologic mechanisms and their clinical
expression is needed.
A distinction must be made between the concept of hepatic
encephalopathy and acute massive hepatic cellular failure. Portal-
systemic hepatic encephalopathies are based on a portocaval
anastomosis, whereby protein from gastrointestinal venous
absorption bypasses hepatic metabolism and is carried to the
systemic venous system. This condition is found in chronic cir-
rhosis of the liver with portal-systemic venous shunts. In con-
trast, acute fulminant hepatic failure results in sudden
hepatocellular breakdown and dysfunction. In both situations,
toxic (mainly nitrogenous) substances reach the brain and con-
tribute to the cerebral pathology and clinical symptoms
described below. In advanced stages, the distinction between
both the pathophysiologic and EEG changes of a full-fledged
hepatic coma becomes indistinguishable, regardless of the
mechanism (9).
Patients presenting with hepatic encephalopathy demon-
strate a notable decline in mental functions, including con-
structional apraxias, increased reaction times, and impaired
memory and concentration (46,47). In addition, altered per-
sonality traits may occur, such as apathy, euphoria, childishness,
and irritability. Motorically, patients demonstrate bradykinesia,
asterixis, parkinsonian traits, and pyramidal signs, which may
be present before hypo- and hyperreflexia emerge in profound
coma (46). The level of consciousness decreases from confusion
to lethargy and may progress to deep coma with unresponsive-
ness. The concept of minimal hepatic encephalopathy has also
been raised, in which patients appear clinically normal, with
subtle deficits on neuropsychiatric testing, including changes in
attention and psychomotor speed (47). The demonstration of
minimal hepatic encephalopathy may correlate with decreased
ability to drive and decreased quality of life.
The biochemical and neuropathologic mechanisms
of hepatic encephalopathy are still being elucidated.
Hyperammonemia and glutamate alterations are strongly sus-
pected in both acute and chronic hepatic encephalopathy, as is
the role of increased GABA tone. Shawcross and Jalan summa-
rized the ammonia hypothesis, detailing the role of hyperam-
monemia in contributing to increased astrocytic pH that results
in calcium-dependent release of glutamate, which contributes
to NMDA activation and excitotoxicity (47). In contrast, with
chronic liver disease, hyperammonemia contributes to
decreased chronic astrocytic glutamate stores, resulting in inac-
tivation of the glutamate transporter and decreased postsynap-
tic glutamate receptors on neurons and astrocytes, thereby
contributing to increased cerebral inhibition. However, these
authors also note that there is increased GABAergic cerebral
tone that also contributes to chronic cerebral inhibition in liver
cirrhosis. Ahboucha and Butterworth reviewed PET studies
that suggest a role for increased GABAA receptor activation, also
resulting in increased postsynaptic neuron inhibition (48).
Neuropathologic correlates of acute and chronic liver disease
include astrocytic and cytotoxic edema with the former and
Alzheimer II astrocytic change in the latter (47).
The altered pathophysiology of hepatic encephalopathy may
be demonstrated by both evoked potentials and brain MRI, as
well as EEG. The varying dynamics of cognitive brain dysfunc-
tions are mirrored by altered event-related P300 potentials,
even with a minor degree of dysfunction and with less sensitiv-
ity to medication use than the EEG (49). Even in the absence of
overt hepatic encephalopathy, brain MRI may be abnormal and
demonstrate low-grade cerebral edema (50).
The electroencephalographic changes of hepatic encephalopa-
thy have been well known since the studies of Foley et al. and
Bickford and Butt (2,3). The degree of slowing often parallels the
level of blood ammonia (51). Similarly, the degree of EEG slow-
ing correlates with severity of liver disease (52). Amodio and
Gatta summarized the common changes associated with hepatic
encephalopathy, including low-frequency alpha background
rhythms with sudden intermittent random theta waves superim-
posed in either a temporal, frontal, or generalized distribution
 Chapter 21 ■ Metabolic Disorders and EEG 397

Figure 21.2 Hepatic encephalopathy in a 42-year-old patient.

Predominant activity around 2.5 Hz, most notable over the frontocen-
tral regions. Several triphasic waves are seen in channels 4 and 15 of
the right portion of the tracing.
Figure 21.1 Hepatic encephalopathy in a 59-year-old patient. The EEG
demonstrates moderate slowing of the background rhythm (5 to 6 Hz),
with some intermixed 8 to 10 Hz alpha waves of lower voltage.
More recently, several authors have questioned the diagnos-
tic specificity of triphasic waves for hepatic encephalopathy.
Fisch and Klass demonstrated triphasic waves in numerous
(53). Subsequently, slow theta background activity may predom- conditions, including stroke, hypertensive encephalopathy,
inate (3,9,54–57). As the clinical presentation worsens, the EEG hyper/hyponatremia, hypercalcemia, hypoglycemia, encephali-
may progressively show more disorganization with diffuse and tis, cerebral abscess, septic shock, dementia, lithium toxicity,
asynchronous theta and delta slowing over both hemispheres. and the postictal state. However, no specific characteristic of the
With severe coma, the background activity may decrease in triphasic wave morphology could reliably differentiate the eti-
amplitude and frequency to the point of becoming isoelectric ology of the triphasic wave from the possibilities noted above
(53,58). Girier and colleagues investigated the EEG findings in (63). Aguglia and colleagues and Blatt and Brenner similarly
fulminant hepatic failure in children and described the stages of noted the occurrence of triphasic waves with etiologies ranging
progressive EEG deterioration (59). Inferences with regard to the from stroke, brain tumors, Binswanger’s subcortical
need for liver transplant can be made on the basis of the EEG encephalopathy, and lithium-treated aged psychiatric patients
findings. Figure 21.1 demonstrates the mild-to-moderate degree (64,65). More recently, Kenangil and colleagues and Eriksson
of slowing that may be seen with hepatic encephalopathy, while
Figure 21.2 demonstrates more marked slowing.
Bickford and Butt introduced the term triphasic waves, clas-
sically described as moderate to high-amplitude slow waves
occurring with a frequency of 1.5 to 3 Hz with an initial sharp
positive transient, followed by a smaller negative component
(Figs. 21.3 and 21.4) (3). The discharges are often noted to
have a frontal to occipital delay in expression, occur singly or
in runs lasting 0.5 to 2 seconds, and arise from a slowed and
disorganized background (60). Triphasic waves may be seen in
fully awake patients or may commonly occur with worsening
cognitive status. However, these waveforms resolve with sleep. Figure 21.3 Examples of triphasic waves taken from two patients with
In general, the incidence of triphasic waves in hepatic hepatic encephalopathy. Note wave morphology. The numbers 1, 2, and
encephalopathy lies around 25% (57,61,62). 3 indicate the phase of the triphasic wave.
398 Part III ■ Clinical EEG: General
Figure 21.4 Triphasic waves of unusually paroxysmal configuration in
a 59-year-old patient with hepatic encephalopathy and no history of
seizures. Also note the distribution of the potentials and a posterior
maximum of the triphasic waves; this is not altogether uncommon.
and Wictor also described the occurrence of triphasic waves
with bilateral paramedian thalamic stroke and CNS Lyme dis-
ease, respectively (66,67). Karnaze and Bickford also reviewed
the medical records of patients whose EEGs demonstrated
triphasic waves (68). Of the 50 patients identified, more than
half had hepatic encephalopathy as the likely etiology of the
triphasic waves. However, other associated conditions included
renal failure (10), anoxia (9), hyperosmolarity (2), and hypo-
glycemia (1). Of note, only 43% of hepatic encephalopathy
patients demonstrated all the diagnostic features of triphasic
waves noted above.
The presence of triphasic waves appears to be a poor prog-
nostic indicator. Karnaze and Bickford found that patients with
triphasic waves experienced mortalities ranging from 30% with
renal encephalopathy, 60% with hepatic encephalopathy to
100% with anoxia (68). Bahamon-Dussan and colleagues simi-
larly found that patients with triphasic waves, regardless of eti-
ology, had a 50% 1-month mortality rate and 77% mortality
rate at 22 months of follow-up (69). However, these studies
were performed over 20 years ago and advances in medicine
may have lowered the mortality rates associated with these con-
ditions. No recent evaluation of the prognostic significance of
triphasic waves has been published.
Quantification of EEG data has been done by some authors
(9,55,70–75). Amodio and colleagues recently compared differ-
ent EEG hepatic encephalopathy classification systems (modi-
fied Parsons-Smith, Conn’s, spectral analysis according to Van
der Rijt and ANNES [artificial neural network expert system])
to assess for correlations between the individual classification
schemes with neuropsychiatric test results, severity of liver dis-
ease, biochemical indices of liver function, and risk of death
(76). While all classification schemes, including visual assess-
ment and spectral analysis, were appropriately correlated with
these variables, the greatest reliability was noted with automatic
quantification systems. In addition, spectral EEG analysis has
also been used to identify minimal hepatic encephalopathy fea-
tures, including an increase in theta power in the posterior divi-
sions and less commonly a decrease in mean dominant
frequency and increase in relative delta power (53).
EEG spectral analysis may help identify the initial changes
associated with minimal hepatic encephalopathy, and the EEG
may assist with the prognosis and treatment response to liver
transplantation. Interestingly, while spectral EEG analysis may
improve after orthotopic liver transplantation, neuropsychi-
atric impairments may remain. As cerebral disturbances after
liver transplant are not uncommon, caused by factors such as
general anesthetics, impaired liver function, sepsis, or hypoxia
(77), the EEG may be useful in the assessment of cerebral dys-
function and possible electrographic seizures. In addition, the
EEGs of post-transplant patients who do not survive demon-
strate a greater incidence of epileptiform abnormalities than
survivors (78). Thus, the EEG may assist with the differential
diagnosis of post-transplant cerebral disturbances and assist
with post-transplant prognosis.
Despite the paroxysmal appearance of triphasic waves, clini-
cally overt epileptic manifestations are not very common with
hepatic encephalopathy and occur much less frequently than in
renal encephalopathies (79–81). Surprisingly, however, Adams
and Foley found an approximately equal incidence of seizures
in renal and hepatic failure (82). Similarities between triphasic
waves and spike-wave complexes cannot be denied. The bound-
aries between these abnormalities were discussed by Hughes,
Klass, and Ghigo (83–85). Ficker et al. demonstrated various
forms of spikes, focal sharp waves, and generalized spike waves
in patients with hepatic encephalopathy (86). These patients
also had epileptic seizures and poor outcomes. In addition,
Eleftheriadis and colleagues also discussed the case of a gentle-
man presenting with continuous generalized tonic–clonic
seizures in the setting of decompensated hepatitis C liver cir-
rhosis (87). Therefore, a continuum of epileptiform discharges
and electrographic seizures may be seen with hepatic
encephalopathy.
Chronic hepatocerebral degeneration may lead to a variety of
EEG changes. The acquired form is neurologically characterized
by dysarthria, ataxia, intention tremor, and choreoathetosis
affecting chiefly the cranial musculature (88,89). This clinical
entity is considered to occur much more frequently than the
familial form (Wilson disease) (88). The acquired form may
develop in various chronic liver diseases and often leads to the
above-described picture of hepatic encephalopathy. Such a
development is less common in Wilson disease. Nevsimalova
and colleagues found only mild alterations in EEG recordings
of patients diagnosed with Wilson disease, with predominantly
generalized intermittent theta or delta slowing, occasionally
with associated sharp waves (90). Only 21% of patients
demonstrated focal findings, of which 15% demonstrated focal

epileptiform discharges. Giagheddu and colleagues noted
abnormal EEGs in only 3 of 23 patients studied. These record-
ings demonstrated temporal theta slowing, left paramedian
epileptiform activity (in a patient with a prior thalamotomy),
and diffuse slow activity with triphasic waves (in a patient expe-
riencing symptoms of hepatic encephalopathy) (91).
RENAL DISEASE (RENAL
ENCEPHALOPATHY, UREMIA)
Various gradations of EEG abnormalities have been reported in
both acute as well as chronic renal failure and may potentially
be reversed with renal transplantation.
In acute uremia, the neurologic picture is characterized by
agitation, confusion, tremor, fasciculations, myoclonus, and
coma (92). The EEG may demonstrate irregular low-voltage
activity with slowing of the posterior basic rhythm and occa-
sional theta bursts. Other cases show prolonged bursts of bilat-
eral synchronous mixed slow and sharp activity or frank spikes.
Bilateral spike discharges may or may not be associated with
widespread myoclonic jerking. Grand mal and partial seizures
as well as infrequent nonconvulsive status epilepticus may
develop (93,94). About 10% to 30% of patients with renal
insufficiency develop epileptic seizures (95–97). Seizures are
usually due to water–electrolyte imbalance (98,99), although
they may also occur in the setting of medication side effects,
infection, malignant hypertension, or dialysis disequilibrium
(95,96).
Figure 21.5 Left: Encephalopathy due to chronic renal failure (age 50 years). Note intermittent runs of delta slowing.
Right: The same patient in sleep. Trains of delta waves with superimposed spindles, not exceeding the normal limits
of variability. The tendency toward normalization in sleep is found in most forms of metabolic encephalopathy.
Chapter 21 ■ Metabolic Disorders and EEG 399
In chronic uremia, patients may experience decreased
energy, poor memory, irritability, and agitation (100). The
EEG most typically demonstrates a slow and disorganized
background, with more pronounced slowing associated with
worsening symptoms of uremia and a prior history of malig-
nant hypertension (101–106) (Fig. 21.5). While Hughes noted
the closest parallel between increased urea concentrations and
more marked background slowing, this association was mini-
mal and of limited clinical significance (100). Rarely, theta dis-
charges, sharp waves, and polyspike and wave complexes are
noted (103,104). Photic stimulation may result in photomy-
oclonic or photoparoxysmal responses (103). Notably, children
with chronic uremia similarly demonstrate an increase in theta
and delta and a decrease in beta activity with spectral EEG
analysis (107), while only 20% of children demonstrate low-
amplitude spikes or bilateral frontocentral spike and wave dis-
charges (108).
Patients with chronic renal failure also experience alterations
in their sleep, with a decrease in deep, rapid eye movement
(REM), and total sleep times, resulting in increased difficulty
classifying sleep stages (109,110). Long bursts of high-voltage
12 to 13/sec waves with enhanced vertex sharp activity in
drowsiness, lack of spindles (14/sec) in stage 2 sleep, and pro-
longed high-voltage slow bursts with awakening have been
reported (111). In REM sleep, ocular movements are enhanced;
they may even be continuous and associated with blinking and
brief myoclonic motions. However, with hemodialysis, deeper
stages of sleep are attained (110).
400 Part III ■ Clinical EEG: General
Dialysis helps to maintain patients with chronic renal failure,
although it may result in transient worsening of EEG patterns,
with an increase in background slowing, bursts of frontal slow
waves, and rarely, spike and wave complexes (100,101,112).
However, with successive dialysis treatments, a slow, delayed
increase in background frequency is noted (105). Occasionally,
dialysis disequilibrium syndrome may occur, typically at the
onset of dialysis therapy or with rapid and highly effective treat-
ment (95). The symptoms may include irritability, nausea,
headache, blurred vision, asterixis, confusion, and generalized
convulsions, which may occur within 24 hours of the dialysis
session. The entity is felt to result from a rapid systemic clear-
ance of urea, causing cerebral edema due to slower cerebral
clearance. The risk for dialysis disequilibrium is diminished with
the use of aluminum-free dialysate. With chronic dialysis treat-
ment, dialysis encephalopathy may also occur, characterized by
speech and memory difficulties, psychiatric changes, involun-
tary movements, seizures, and death (100,113). The EEG fre-
quently reveals a slow and disorganized background rhythm,
frontal intermittent rhythmic delta activity, or paroxysmal fron-
tocentral spike and wave discharges (100,103,113,114).
Fortunately, renal transplantation may reverse the EEG
background slowing in as little as 2 weeks postsurgery (106).
While there have been various reports of post-transplant
seizures, more recent studies suggest an incidence of 2.3% in
adults (115) and none in children (116), presumably due to
improved post-transplant care. Seizures were more commonly
attributed to immunosuppressive agents, but also due to infec-
tion, hypertension, and biochemical alterations (116).
ECLAMPSIA GRAVIDARUM
Eclampsia gravidarum refers to the development of one or
more generalized convulsions and/or coma in pregnant or post-
partum patients with preeclampsia but in the absence of other
neurologic conditions (117). Preeclampsia is a syndrome char-
acterized by hypertension and proteinuria that typically occurs
after 20 weeks gestation. A number of target organs are affected,
including the brain, liver, kidney, and placenta (118). The
underlying factors involved in the pathogenesis of preeclampsia
are complex, involving a combination of maternal, fetal, and
placental factors. More specifically, abnormalities in the devel-
opment of placental vasculature early in pregnancy are thought
to lead to release of angiogenic factors into the systemic circu-
lation causing maternal endothelial dysfunction and ultimately
systemic vasoconstriction (119,120). Similarly, vasoconstriction
or vasospasm is seen in the cerebral blood vessels, which has
been demonstrated on cerebral angiography (121,122), and is
believed to be the cause of the observed pathologies, namely
focal necrosis, hemorrhage, and vasogenic edema (123,124).
Neuroimaging with brain MRI also shows areas of cerebral
ischemia, infarction, hemorrhage, or edema, in the watershed
regions, with a particular predilection for the parieto-occipital
lobes (125). Given these findings, loss of cerebral autoregulation
leading to hypertensive encephalopathy is thought to play a key
role in the pathogenesis of eclamptic convulsions (117).
There is little EEG documentation of the ictus (126). In the
acute stage of eclampsia, the posterior alpha rhythm is seldom
recorded, likely due to the preferential localization of the main
process in the occipital lobes (127). This is congruent with clin-
ical findings of cortical blindness (128,129) as well as the MRI
findings described above. In a group of 84 patients, Thomas et
al. found the interictal EEG to be abnormal in about 80% of
patients with eclampsia (130) and the abnormalities included
diffuse slowing (56.4%), focal slowing (25.6%), spike discharges
(14.1%), alpha coma (1.3%), and electrical silence (2.6%). A
study to compare cerebral magnetic resonance and interictal
EEG findings in preeclamptic versus eclamptic women by
Osmanagaoglu et al. found that EEG abnormalities occurred at
statistically higher rates in the eclamptic group than the
preeclamptic group (131). In the group with mild preeclampsia,
20% of patients demonstrated EEG abnormalities, including
diffuse slowing, focal slowing, or hemispheric slowing, while in
the group with severe preeclampsia, 36% of patients showed
these abnormalities as well as focal discharges. In the eclamptic
group, the percentage of patients showing EEG abnormalities
was significantly higher at 83%, and the abnormalities included
diffuse slowing, focal slowing, focal spike discharges, and multi-
focal spike discharges. Although focal MRI abnormalities were
noted in many cases, the EEG abnormalities tended to be diffuse.
The authors found no significant correlation between EEG
abnormalities and mean arterial blood pressure (130,132). Of
the 14 patients with abnormal EEGs, 94% of patients demon-
strated resolution of these abnormalities at 1-month follow-up.
Magnesium sulfate is frequently used in the treatment of
women with eclampsia and is typically administered during
labor and the 12 to 24-hour postpartum period (133).
Although a clinical benefit has been shown with treatment in
this setting, magnesium infusion does not alter the EEG abnor-
malities before, during, or after the infusion (132). A report by
Brophy described normalization of the EEG in eclampsia with
pyridoxine infusion (134); however, there is currently no clear
preventive therapy for preeclampsia (117).
Seizures occur acutely during eclampsia and chronically lin-
gering epileptic seizures known as “posteclamptic epilepsy” as a
consequence of eclampsia are rare (135).
ELECTROLYTE DISTURBANCES
Hypocalcemia
Hypocalcemia is frequently the consequence of hypoparathy-
roidism and manifests with tetany of peripheral origin with
carpopedal spasms and seizures due to CNS hyperexcitability. It
is important to emphasize that tetanic spasms are of peripheral
origin rather than cerebral origin as the EEG is normal in
tetany. In the CNS, hypocalcemia is highly epileptogenic, pro-
ducing marked EEG changes with slowing and generalized
bursts of spikes (Fig. 21.6). Corriol et al. found epileptic mani-
festations develop at calcium levels around 5 to 6 mg/dL in
humans and animal models (136). Seizures tend to be general-
ized, although partial seizures may also occur and typically

Figure 21.6 Hypocalcemia (ca. 5.2 mg/ 100 mL). Note marked diffuse
slowing along with scattered sharp activity.
respond poorly to anticonvulsants; however, both the seizures
and EEG abnormalities improve once normal serum calcium is
restored (137). The typical interictal EEG in the setting of
hypocalcemia shows high-amplitude delta activity and may be
accompanied by focal spike and spike-wave discharges, all of
which may be enhanced by hyperventilation (137). However,
no clear correlation was found among calcium serum level,
onset of seizures, and EEG changes in other studies (138).
Neonatal hypocalcemia produces acute epileptic seizures that
tend to be focal, with shifting focality and severe transient EEG
changes, such as focal rhythmic, high-voltage, frontocentral dis-
charges that generalize rapidly and often lasting several seconds
(139). A case reported by Kossoff showed seizures consisting of
3- to 4-Hz spike and wave discharges that resolved after intra-
venous calcium administration (140). Rachitic tetany (common
infantile tetany) is associated with hypocalcemia, carpopedal
spasms, laryngospasm, convulsions, and the clinical signs of rick-
ets; the EEG changes are usually mild with no spike activity (141).
In pseudohypoparathyroidism with short stature, osteo-
chondral dysplasia, short fourth finger, hypocalcemia, and
tetany, the EEG is most often abnormal as in cases of hypocal-
cemia (142,143). No significant EEG changes have been found
in pseudopseudohypoparathyroidism.
The EEG was found to be normal in five of six patients with
Fahr disease; the remaining patient showed mild slowing (144).
Hypercalcemia
Hypercalcemia may be due to hyperplasia or neoplasm of the
parathyroid glands, excessive parathyroid activity caused by
renal failure or skeletal decalcification, vitamin D intoxication,
paraneoplastic syndrome, or malignancy with bony metastases.
Hypocalcemia is associated with CNS hyperexcitability,
whereas hypercalcemia is associated with depression of the CNS
and epileptic activity rarely occurs. Clinically, irritability,
lethargy, and malaise are seen with mild elevations of serum cal-
cium, but as levels continue to rise, patients progress to coma.
When calcium levels reach about 13 mg/dL, EEG changes start to
appear (145). Neurologic deficits and mental impairment are
accompanied by EEG abnormalities consisting mainly of bursts
of frontal-dominant delta frequency activity and less frequently
Chapter 21 ■ Metabolic Disorders and EEG 401
triphasic waves (143,146–148), with or without diffuse slowing of
the basic rhythm (147). In general, epileptiform discharges are
not noted except for a case described by Huott et al. in which
massive spike activity was seen independently in the right and left
occipital regions in the setting of severe hypercalcemia (16.6
mg/dL), as well as two generalized seizures, and a right homony-
mous hemianopsia (149). The presence of occipital spike-slow-
wave complexes raises the suspicion for a vasospastic effect in the
posterior circulation associated with high calcium levels (150).
Hyponatremia and Water
Intoxication
Hyponatremia and acute water intoxication lead to severe EEG
abnormalities with pronounced generalized slowing. Typically,
the rate or sodium change is more important than the absolute
degree of hyponatremia as the determining factor in the devel-
opment of mental status changes, EEG abnormalities, and
seizures. Zwang and Cohn reported a case of acute compulsive
water intake resulting in a serum sodium level of 116.4 mg/dL,
causing convulsive status epilepticus, papilledema, and an EEG
with diffuse 0.5- to 2-Hz activity (151). Despite rapid normal-
ization of the electrolytes, EEG recovery was slow. Okura et al.
also reported a case of convulsive status epilepticus due to poly-
dipsia, where the EEG demonstrated delta waves with spike-
wave complexes, but the EEG improved quickly as his
hyponatremia was corrected (152). Other abnormalities have
been described such as bursts of high-voltage rhythmic delta
activity, central high-voltage 6- to 7-Hz activity with stimula-
tion-induced paroxysms of delta waves (153), periodic delta
waves (154), triphasic waves and periodic lateralized epilepti-
form discharges (155,156), and absence status epilepticus (157).
Hypernatremia
While hyponatremia is likely to be the underlying etiology of
seizures, hypernatremia may result as a consequence of seizures
(158). Hypernatremia is associated with reduction of cerebral
volume as water shifts from the intracellular space to the extra-
cellular space, and again the rate of electrolyte change is more
important than the absolute degree of hypernatremia as the
determining factor in the development of clinical symptoms. In
general, seizures are not seen except in cases of rapid correction
of hypernatremia. The EEG usually demonstrates mild nonspe-
cific slowing (159).
Hypomagnesemia
Hypomagnesemia causes cardiac arrhythmias, hyperexcitability
of the peripheral nervous system (tetany), and hyperexcitability
of the CNS (seizures, mental status changes, focal deficits). It is
often associated with hypocalcemia since magnesium defi-
ciency inhibits calcium mobilization via inhibition of secretion
and action of the pararathyroid hormone, causing secondary
hypocalcemia (160). The effects of hypomagnesemia are
thought to be potentiated by hypokalemia as well (160). Low
serum magnesium levels can result from excessive diuretic use,
renal disease, inadequate intake, or impaired gastrointestinal
absorption, as in cases of malabsorption syndromes and
402 Part III ■ Clinical EEG: General
postgastrointestinal surgeries. Magnesium serves as a neuronal
membrane stabilizer, and it is indicated in the treatment of
eclampsia (see the section “Eclampsia Gravidarum”). Unlike the
previously mentioned electrolyte disturbances, detailed studies
of EEG changes with hypomagnesemia are lacking. Clinical
symptoms, including epileptic seizures, typically do not develop
until the magnesium level falls below 1.2 mg/dL (161).
Although hypomagnesemia is a systemic disturbance, patients
may present with focal neurologic deficits, as was reported in
the case of a young child who presented with hemiparesis and
aphasia in the setting of severe hypomagnesemia and hypocal-
cemia. His EEG demonstrated background slowing in the
affected hemisphere (162), which resolved along with his clini-
cal symptoms when his magnesium and calcium levels were
restored to the normal range.
VITAMIN DEFICIENCIES
Vitamin B1
Thiamine (vitamin B1) deficiency may give rise to manifesta-
tions in the cardiovascular and nervous systems. There are two
forms: (i) dry beriberi, which consists of sensorimotor neu-
ropathy and Wernicke–Korsakoff syndrome, and (ii) wet
beriberi, which consists of edema and congestive heart failure,
but little CNS manifestations. The clinical picture of Wernicke’s
encephalopathy, with or without Korsakoff syndrome, is fre-
quently encountered in alcoholics, with predominant oculomo-
tor and cerebellar symptomatology (163), although it can also
be seen in other conditions, such as hyperemesis, dialysis, and
postgastrointestinal surgery.
EEG changes range from mild to severe and, in general, paral-
lel the severity of the neurologic deficits (164). Initial alpha slow-
ing is gradually followed by the appearance of theta and delta
frequencies (165) and by diffuse sharp and slow wave complexes
in severely ill patients (166). In a group of 50 adult patients, 16
patients with neurologic manifestations demonstrated epilepti-
form activity on their EEGs (167). Thiamine deficiency in infants
is rare in developed countries and occurs mainly in breast-fed
infants of mothers who have inadequate thiamine intake or in
infants fed with thiamine-deficient formula (168). In a group of
20 Israeli children who developed infantile thiamine deficiency,
seven patients subsequently developed epilepsy (169). Upon ini-
tial presentation, all patients were noted to have episodes of
apnea, increased tone, and clinical seizures that were tonic,
myoclonic, or focal in nature. Initial EEG recording was available
in six of the seven patients and demonstrated diffuse background
slowing, with or without lateralized epileptiform activity such as
sharp waves or spikes, and in one case, a focal electrographic
seizure in the frontal region was recorded. The seizures resolved
with thiamine repletion; however, after a seizure-free period of 1
to 9 months, all children had recurrent focal motor seizures,
complex partial seizures, or myoclonic seizures.
Vitamin B3
Vitamin B3 (niacin) deficiency results in pellagra, which is char-
acterized clinically by dermatitis, diarrhea, depression, and
dementia. Electrographically it is associated with diffuse EEG
slowing with theta and delta activity that is reversible with
niacin supplementation (170).
Vitamin B6
Vitamin B6 (pyridoxine) deficiency on a nutritional basis has been
recognized as a rare cause of severe and even fatal convulsions in
neonates and infants (171,172). Pyridoxine dependency develops
during fetal life as a genetic disorder and causes both intrauterine
and postnatal seizures (173–175). EEGs of affected patients show
generalized spike activity. Both the EEG abnormalities and clini-
cal seizures respond quickly to treatment with pyridoxine.
EEG abnormalities have also been reported in controlled
studies of B6 depletion by Kretsch et al. in which healthy young
women were placed on a diet containing less than 0.05 mg of B6
(176). Within 12 days, 2 of 11 study subjects exhibited abnor-
mal EEG patterns, including focal slowing and focal sharp wave
activity, which resolved with vitamin B6 repletion, but no clear
seizures were seen.
Vitamin B12
Vitamin B12 (cobalamin) deficiency gives rise to subacute com-
bined degeneration characterized by changes in the peripheral
and CNS. Patients present with large fiber peripheral neuropa-
thy and may exhibit hyperreflexia and mental status changes
such as delirium or dementia. Slow activity in the EEG may be
quite pronounced (177,178) and temporal spike and sharp wave
activity has been reported (178); but these findings are unrelated
to the presence of myelopathy or neuropathy (179). Pronounced
epileptiform EEG abnormalities have been reported in the early-
onset cobalamin C/D deficiency: an inborn error of intracellular
cobalamin metabolism with high plasma levels of methyl-
malonic acid, homocystine, and homocysteine (180). The EEG
demonstrates massive focal or generalized spikes and the cere-
bral MRI shows diffuse loss of white matter.
ENDOCRINE DISORDERS
Adrenal Cortex
Adrenocortical insufficiency (Addison disease) can be associ-
ated with pronounced background slowing with delta and theta
activity, but such changes are found in severe cases only
(181,182). In milder forms, the EEG shows little or no abnor-
mality. In severe cases, loss of reactivity to eye opening, increase
sensitivity to hyperventilation, and decreased beta activity also
can be seen (183). Cortisone can dramatically improve the clin-
ical and electroencephalographic picture, and it has been sug-
gested that the EEG abnormalities are caused mainly by altered
glucose metabolism.
In adrenal cortical hyperfunction (Cushing disease), EEG
abnormalities are less prominent than in Addisonian states. Slow
as well as fast frequencies may be excessively developed (184).
No specific metabolic disturbance can be attributed to the EEG
changes. There are also reports that therapeutic dosages of
adrenocorticotropic hormone (ACTH) or cortisone may pro-
duce mild-to-moderate EEG changes along with frequent men-
tal changes and enhanced seizure susceptibility (185). In adult
epileptics with abnormal EEG tracings, the EEG changes are

enhanced by intravenous cortisone (186), in contrast to the ben-
eficial effect of ACTH and cortisone in hypsarrhythmic infants.
Pheochromocytomas may cause extremely high blood pres-
sure values, but in these critical states, the EEG may remain
normal (187) unless hypertensive encephalopathy develops.
Pituitary Gland
Postpartum necrosis of the anterior lobe of the pituitary gland
(Sheehan syndrome) is a common cause of severe hypopitu-
itarism. The EEG shows massive diffuse intermittent or contin-
uous theta and delta activity, especially in association with
impaired consciousness, and these findings most likely reflect
the secondary depression of adrenocortical function (188).
Thyroid Gland
Hyperthyroidism
Hyperthyroidism is commonly associated with acceleration of
alpha rhythm frequency (Fig. 21.7). Such observations can be
traced back to Rubin et al. (189), Lindsley and Rubinstein (190),
and Ross and Schwab (191). Enhancement of fast activity was
reported by Gibbs and Gibbs (192), Vague et al. (193,194), and
Condon et al. (5). Wilson et al. found the highest incidence of
abnormalities in premenopausal females (195). According to
Vague et al., rolandic mu rhythm may be augmented by hyper-
thyroidism (194).
Paroxysmal bursts and clinical seizures, such as complex par-
tial seizures, focal motor seizures, and generalized seizures, have
been reported by Skanse and Nyman (196), Jabbari and Huott
(197), and Miyazaki (198). Administration of thyroxin pro-
duces epileptic activity. Epileptic patients with thyroid dysfunc-
Figure 21.7 Hyperthyroidism in a 52-year-old woman. Note fast basic
rhythm of 14 to 16 Hz with good blocking effect. Faster activity over
central regions does not show the blocking response.
Chapter 21 ■ Metabolic Disorders and EEG 403
tion may respond poorly to anticonvulsants unless the underly-
ing endocrine disturbance is properly treated. In thyrotoxic cri-
sis, prominent EEG abnormalities have been noted with
encephalopathy prompted by severe infection and surgical
intervention. Cranial nerve signs, acute myeloencephalopathy,
and choreic movements may also occur. In these cases, the EEG
may be dominated by marked slowing with superimposed fast
activity (199), triphasic waves (183), or bursts of anterior delta
frequency activity (200).
Hashimoto Encephalopathy
Hashimoto encephalopathy is a clinical syndrome of
encephalopathy, elevated serum levels of antithyroid antibodies,
and may present in two subtypes—a vasculitic subtype with
stroke-like episodes and a diffuse progressive type with dementia
and psychiatric symptoms (201,202). Both forms may be accom-
panied by depressed mental status, myoclonus, and tremor (202).
Epileptic seizures (partial and generalized), complex partial sta-
tus epilepticus, and epilepsia partialis continua have also been
reported (203,204). In Hashimoto’s thyroiditis, diffuse EEG
abnormalities can be seen such as background slowing (205,206).
Hypothyroidism
Hypothyroidism may present with a variety of clinical condi-
tionals such as congenital myxedema, cretinism, and myxedema
in adulthood with complications such as myxedema coma. In
myxedematous infants, a delay of the EEG development and
especially of the sleep spindle development has been reported
by Schultz et al. (207), and low voltage and excessive slowing
have been noted by Niemanan (208). In adults, an excess of
low-voltage slow activity has been reported (191) and reactivity
may be poor or absent (209). Epileptic seizures may precede
myxedema coma (210), and various EEG patterns have been
noted: triphasic waves and generalized periodic sharp waves
resembling Creutzfeldt–Jakob disease (211,212).
Other Hormonal Disturbances
The physiologic influence of estrogens on the EEG has been
demonstrated by some investigators. Mild-to-moderate slowing
of the alpha rhythm was found at the end of pregnancy (213).
However, subsequent studies could not replicate these findings,
and the authors hypothesize that the clinically observed changes
may be reflective of a pregnancy-related disorder (214). Although
an initial report by Dusser de Barenne and Biggs in 1942 found
slowing of alpha 1 to 3 days prior to menstruation (215), subse-
quent studies have not yielded consistent results (216).
ACUTE INTERMITTENT PORPHYRIA
In patients with abdominal pain and polyneuropathy due to
acute porphyria, the EEG is normal. However, when delirium
and mental impairment develop, the EEG shows theta and delta
frequencies (217), and occasionally asymmetrical slowing and
spike or spike-wave activity can be seen (218). Pronounced
slowing is found in central forms of porphyria (Fig. 21.8).
Seizures can be seen in approximately 3% to 5% of the
patients with acute intermittent porphyria (219) and their
404 Part III ■ Clinical EEG: General

indisputable diffuse EEG slowing. A variety of medical condi-

tions (such as pneumonia or influenza in aged patients) and
postsurgical conditions (excluding neurosurgical cases) play a
major role in the genesis of mixed-type encephalopathy. The
length of time of anesthesia augments the problem in postsur-
gical cases; thus, introjecting a pharmacologic element. This
mixed type of encephalopathy usually carries a good progno-
sis. A special form of mixed-type encephalopathies occurs in
30% of severe burns; in these cases, however, the prognosis is
guarded (222).

CHRONIC RESPIRATORY (PULMONARY

INSUFFICIENCY)
Chronic bronchial disease and pulmonary emphysema, along
with right cardiac failure (cor pulmonale), may lead to a
chronic encephalopathy clinically characterized by headache,
papilledema, elevated cerebral spinal fluid pressure, and twitch-
ing of extremities (8,223). The EEG shows considerable diffuse
delta and theta activity; occasionally, anterior delta waves may
show the configuration of triphasic waves. Appropriate treat-
ment may quickly reverse these changes, but oxygen adminis-
tration may be dangerous.
Chronic alveolar hypoventilation in conjunction with obe-
sity gives rise to the Pickwickian syndrome with excessive som-
nolence (224,225) (Fig. 21.9). Most of these patients also suffer
from sleep apnea, especially from the obstructive type (see
Chapter 41). The routine EEG shows abnormal amounts of
diffuse slow activity (Fig. 21.10).

Figure 21.8 A: Cerebral form of acute intermittent porphyria in a

24-year-old woman with slow mentation, personality change, and
grand mal seizures. B: Same patient, improving although still lethargic.
Note runs of rhythmical slowing over left frontotemporal region.
This is an example of occasional focal characteristics in metabolic
encephalopathies.

treatment can be problematic because of the well-known

induction of porphyria with the use of antiepileptic medica-
tions (220). However, gabapentin is considered an exception
and thus is the treatment of choice (221).

MIXED-TYPE ENCEPHALOPATHY
This polyetiologic condition (222) stands apart from all of the
well-defined metabolic encephalopathies. It is a common con-
dition, mostly acute, characterized simply by change of menta- Figure 21.9 Abnormally slow and disorganized waking record in a
tion (delirium) with little or no neurologic deficits but with patient with pickwickian syndrome.

Figure 21.10 Burst of diffuse delta activity during wakefulness in a patient with severe chronic obstructive pul-
monary disease.
REFERENCES
1. Berger H. Das Elektrenkephalogramm des Menschen. Arch
Psychiatry Nervenkr. 1937;106:165–187.
2. Foley JM, Watson CW, Adams RD. Significance of the electroen-
cephalographic changes in hepatic coma. Trans Am Neurol Assoc.
1950;51:161–165.
3. Bickford RG, Butt HR. Hepatic coma: the electroencephalographic
pattern. J Clin Invest. 1955;34(6):790–799.
4. Cadilhac J, Ribstein M, Jean R. EEG et troubles métaboliques. Rev
Neurol (Paris). 1959;100:270–296.
5. Condon JV, Becka DR, Gibbs FA. Electroencephalographic
abnormalities in endocrine disease. N Engl J Med. 1954;251(16):
638–641.
6. Krump JE, ed. Die Klinische Bedeutung des Elecktroenzephalogrammes
bei Vergiftungen, Endotoxikosen und Endokrinopathien. Proceed
Deutsche Internisten-Tagung; Berlin: VEB Verlag Volk und
Gesundheit; 1955.
7. Thiebaut F, Rohmer F, Wackenheim A. Electroencephalographical
study of endocrine diseases. Electroencephalogr Clin Neurophysiol.
1958;10(1):1–30.
8. Glaser GH. Metabolic encephalopathy in hepatic, renal and
pulmonary disorders. Postgrad Med. 1960;27:611–619.
Chapter 21 ■ Metabolic Disorders and EEG 405
9. MacGillivray BB. The EEG in liver disease. In: Remond A, ed.
Handbook of Electroencephalography and Clinical Neurophysiology.
Amsterdam: Elsevier; 1976:26–50.
10. Moruzzi G. Action de l’hypoglycémie insulinique sur l’activité
spontanée et provoquée de l’écorce cérébrale. C R Soc Biol.
1938;128:1181–1184.
11. Himwich HE, Frostig JP, Hoagland H, et al. Clinical electroen-
cephalographic and biochemical changes during insulin hypo-
glycemia. Proc Soc Exp Biol. 1939;40:401–402.
12. Gellhorn E, Kessler M. The effect of hypoglycemia on the elec-
troencephalogram at varying degrees of oxygenation of the blood.
Am J Physiol. 1942;136:1–6.
13. Davis PA. Effect on the electroencephalogram of changing blood
sugar level. Arch Neurol Psychiatry. 1943;49:186–194.
14. Hoagland H, Cameron DE, Rubin MA. The electroencephalogram
of schizophrenics during insulin treatments. Am J Psychiatry.
1937;94:183–208.
15. Gibbs FA, Williams D, Gibbs EL. Modification of the cortical
frequency spectrum by changes in CO2, blood sugar and oxygen.
J Neurophysiol. 1940;3:49–58.
16. Engel GL, Romano J, Ferris EG, et al. A simple method of deter-
mining the frequency spectrum of the electroencephalograph.
Observations on physiological variations in glucose, oxygen,
406 Part III ■ Clinical EEG: General
posture and acid base balance on the normal electroencephalo-
gram. Arch Neurol Psychiatry. 1944;51:134–146.
17. Weinland WL, Weinland W. Űber elektroencephalographische
Beobachtungen beim Insulinschock. Arch Psychiatr Nervenkr.
1949;18:24–44.
18. Lafon R, Passouant P, Billet M, et al. Étude électroencéphalo-
graphique du coma insulinique. Ann Méd Psychol. 1950;2:
251–254.
19. Gibbs FA, Murray EL. Hypoglycemic convulsions. Electro-
encephalgr Clin Neurophysiol. 1954;6:674.
20. Hetzel H, Niedermeyer E. Der Gehirnhydrolysat-Weckeffekt im
hypoglykaemischen Koma und sein hirnelektrisches. Bild Arch
Psychiatr Nervenkr. 1955;76:369–382.
21. Shagass C, Rowsell PW. Serial electroencephalographic and clini-
cal studies in a case of prolonged insulin coma. AMA Arch Neurol
Psychiatry. 1954;72(6):705–711.
22. Regan PF III, Browne-Mayers AN. Electroencephalography, fre-
quency analysis and consciousness; a correlation during insulin-
induced hypoglycemia. J Nerv Ment Dis. 1956;124(2):142–147.
23. Zeigler DK, Presthus J. Normal electroencephalogram at deep lev-
els of hypoglycemia. Electroencephalogr Clin Neurophysiol.
1957;9:523–526.
24. Von Braunmühl A. Insulinschock und Heilkrampf in der Psychiatrie.
Stuttgart: Wissenschaftliche Verlagsgesellschaft; 1947.
25. Boyle PJ, Kempers SF, O’Connor AM, et al. Brain glucose uptake
and unawareness of hypoglycemia in patients with insulin-
dependent diabetes mellitus. N Engl J Med. 1995;333(26):
1726–1731.
26. Tribl G, Howorka K, Heger G, et al. EEG topography during
insulin-induced hypoglycemia in patients with insulin-dependent
diabetes mellitus. Eur Neurol. 1996;36(5):303–309.
27. Tamburrano G, Lala A, Locuratolo N, et al. Electro -
encephalography and visually evoked potentials during
moderate hypoglycemia. J Clin Endocrinol Metab. 1988;66(6):
1301–1306.
28. Bjorgaas M, Sand T, Vik T, et al. Quantitative EEG during con-
trolled hypoglycaemia in diabetic and non-diabetic children.
Diabet Med. 1998;15(1):30–37.
29. Sperling MR. Hypoglycemic activation of focal abnormalities in
the EEG of patients considered for temporal lobectomy.
Electroencephalogr Clin Neurophysiol. 1984;58(6):506–512.
30. Deutsch E, Sohmer H, Weidenfeld J, et al. Auditory nerve-brain
stem evoked potentials and EEG during severe hypoglycemia.
Electroencephalogr Clin Neurophysiol. 1983;55(6):714–716.
31. Kirchner A, Veliskova J, Velisek L. Differential effects of low glucose
concentrations on seizures and epileptiform activity in vivo and in
vitro. Eur J Neurosci. 2006;23(6):1512–1522.
32. Wang S, Hu HT, Wen SQ, et al. An insulinoma with clinical and
electroencephalographic features resembling complex partial
seizures. J Zhejiang Univ Sci B. 2008;9(6):496–499.
33. Dion MH, Cossette P, St-Hilaire JM, et al. Insulinoma misdiag-
nosed as intractable epilepsy. Neurology. 2004;62(8):1443–1445.
34. Engel R, Halberg F, Tichy FY, et al. Electrocerebral activity and
epileptic attacks at various blood sugar levels; with a case report.
Acta Neuroveg (Wien). 1954;9(1–4):147–167.
35. Dibenedetto RJ, Crocco JA, Soscia JL. Hyperglycemia nonketotic
coma. Arch Intern Med. 1965;116:74–82.
36. Daniels JC, Chokroverty S, Barron KD. Anacidotic hyperglycemia
and focal seizures. Arch Intern Med. 1969;124(6):701–706.
37. Maccario M, Messis CP, Vastola EF. Focal seizures as a manifesta-
tion of hyperglycemia without ketoacidosis. A report of seven
cases with review of the literature. Neurology. 1965;15:195–206.
38. Vastola EF, Maccario M, Homan R. Activation of epileptogenic foci
by hyperosmolality. Neurology. 1967;17(5):520–526.
39. Aquino A, Gabor AJ. Movement-induced seizures in nonketotic
hyperglycemia. Neurology. 1980;30(6):600–604.
40. Singh BM, Strobos RJ. Epilepsia partialis continua associated
with nonketotic hyperglycemia: clinical and biochemical profile of
21 patients. Ann Neurol. 1980;8(2):155–160.
41. Gabor AJ. Focal seizures induced by movement without sensory
feedback mechanisms. Electroencephalogr Clin Neurophysiol.
1974;36(4):403–408.
42. Duncan MB, Jabbari B, Rosenberg ML. Gaze-evoked visual
seizures in nonketotic hyperglycemia. Epilepsia. 1991;32(2):
221–224.
43. Harden CL, Rosenbaum DH, Daras M. Hyperglycemia presenting
with occipital seizures. Epilepsia. 1991;32(2):215–220.
44. Huang CW, Hsieh YJ, Pai MC, et al. Nonketotic hyperglycemia-
related epilepsia partialis continua with ictal unilateral parietal
hyperperfusion. Epilepsia. 2005;46(11):1843–1844.
45. Manford M, Fuller GN, Wade JP. “Silent diabetes”: non-ketotic
hyperglycaemia presenting as aphasic status epilepticus. J Neurol
Neurosurg Psychiatry. 1995;59(1):99–100.
46. Knell AJ, Davidson AR, Williams R, et al. Dopamine and serotonin
metabolism in hepatic encephalopathy. Br Med J. 1974;1(5907):
549–551.
47. Shawcross D, Jalan R. The pathophysiologic basis of hepatic
encephalopathy: central role for ammonia and inflammation. Cell
Mol Life Sci. 2005;62(19–20):2295–2304.
48. Ahboucha S, Butterworth RF. Pathophysiology of hepatic
encephalopathy: a new look at GABA from the molecular stand-
point. Metab Brain Dis. 2004;19(3–4):331–343.
49. Kugler CF, Petter J, Taghavy A, et al. Dynamics of cognitive brain
dysfunction in patients with cirrhotic liver disease: an event-
related P300 potential perspective. Electroencephalogr Clin
Neurophysiol. 1994;91(1):33–41.
50. Haussinger D, Schliess F. Pathogenetic mechanisms of hepatic
encephalopathy. Gut. 2008;57(8):1156–1165.
51. Senzolo M, Amodio P, D’Aloiso MC, et al. Neuropsychological and
neurophysiological evaluation in cirrhotic patients with minimal
hepatic encephalopathy undergoing liver transplantation.
Transplant Proc. 2005;37(2):1104–1107.
52. Amodio P, Del Piccolo F, Petteno E, et al. Prevalence and prognos-
tic value of quantified electroencephalogram (EEG) alterations in
cirrhotic patients. J Hepatol. 2001;35(1):37–45.
53. Amodio P, Gatta A. Neurophysiological investigation of hepatic
encephalopathy. Metab Brain Dis. 2005;20(4):369–379.
54. Friedlander WJ. Electroencephalographic changes in hyperam-
monemia. Electroencephalogr Clin Neurophysiol. 1956;8(3):
513–516.
55. Laidlaw J. The application in general medical conditions of a visual
method of assessing and representing generalized electroen-
cephalographic abnormalities. J Neurol Neurosurg Psychiatry.
1959;22(1):69–76.
56. Laidlaw J, Read AE, Sherlock S. Morphine tolerance in hepatic cir-
rhosis. Gastroenterology. 1961;40:389–396.
57. Silverman D. Some observations on the EEG in hepatic coma.
Electroencephalogr Clin Neurophysiol. 1962;14:53–59.
58. Trewby PN, Casemore C, Williams R. Continuous bipolar record-
ing of the EEG in patients with fulminant hepatic failure.
Electroencephalogr Clin Neurophysiol. 1978;45(1):107–110.
59. Girier B, Clouzeau J, Navelet Y, et al. Fulminant hepatic failure
(FHF) in children. EEG prognosis. Electroencephalogr Clin
Neurophysiol. 1989;73:48P–49P.

60. Sheridan PH, Sato S. Triphasic waves of metabolic encephalopathy
versus spike-wave stupor. J Neurol Neurosurg Psychiatry. 1986;
49(1):108–109.
61. Rumpl E, Hackl JM, Gerstenbrand F, et al. Zum EEG im
Leberkoma. Z EEG-EMG. 1979;10:88–94.
62. Simsarian JP, Harner RN. Diagnosis of metabolic encephalopathy:
significance of triphasic waves in the electroencephalogram.
Neurology (Minneapolis). 1972;22:456.
63. Fisch BJ, Klass DW. The diagnostic specificity of triphasic wave
patterns. Electroencephalogr Clin Neurophysiol. 1988;70(1):1–8.
64. Aguglia U, Gambardella A, Oliveri RL, et al. Nonmetabolic causes
of triphasic waves: a reappraisal. Clin Electroencephalogr.
1990;21(3):120–125.
65. Blatt I, Brenner RP. Triphasic waves in a psychiatric population: a
retrospective study. J Clin Neurophysiol. 1996;13(4):324–329.
66. Kenangil G, Orken DN, Yalcin D, et al. Triphasic EEG pattern in
bilateral paramedian thalamic infarction. Clin EEG Neurosci.
2008;39(4):185–190.
67. Eriksson B, Wictor L. EEG with triphasic waves in Borrelia
burgdorferi meningoencephalitis. Acta Neurol Scand. 2007;116(2):
133–136.
68. Karnaze DS, Bickford RG. Triphasic waves: a reassessment of their
significance. Electroencephalogr Clin Neurophysiol. 1984;57(3):
193–198.
69. Bahamon-Dussan JE, Celesia GG, Grigg-Damberger MM.
Prognostic significance of EEG triphasic waves in patients with
altered state of consciousness. J Clin Neurophysiol. 1989;6(4):
313–319.
70. Cohn R, Castell DO. The effect of acute hyperammonemia on the
electroencephalogram. J Lab Clin Med. 1966;68(2):195–205.
71. Kennedy J, Parbhoo SP, MacGillivray B, et al. Effect of extracorpo-
real liver perfusion on the electroencephalogram of patients in
coma due to acute liver failure. Q J Med. 1973;42(167):549–561.
72. Kurtz D, Zenglein JP, Imler M, et al. Night sleep in porto-caval
encephalopathy. Electroencephalogr Clin Neurophysiol. 1972;33(2):
167–178.
73. MacGillivray BB. An EEG monitor incorporating simple pattern
recognition. J Physiol. 1969;201:65–67.
74. Van der Rijt CC, Schalm SW, De Groot GH, et al. Objective meas-
urement of hepatic encephalopathy by means of automated EEG
analysis. Electroencephalogr Clin Neurophysiol. 1984;57(5):
423–426.
75. Yaar I, Shapiro MB, Pottala EW. Spectral analysis of the EEG in
hepatic encephalopathy treated with levodopa. Electroencephalogr
Clin Neurophysiol. 1981;52(6):617–625.
76. Amodio P, Pellegrini A, Ubiali E, et al. The EEG assessment of low-
grade hepatic encephalopathy: comparison of an artificial neural
network-expert system (ANNES) based evaluation with visual
EEG readings and EEG spectral analysis. Clin Neurophysiol.
2006;117(10):2243–2251.
77. Lehmkuhl P, Kaukemuller J, Pohl S, et al. EEG monitoring
of intensive care patients after liver transplantation.
Electroencephalagr Clin Neurophysiol. 1988;69:17P.
78. Wszolek ZK, Aksamit AJ, Ellingson RJ, et al. Epileptiform elec-
troencephalographic abnormalities in liver transplant recipients.
Ann Neurol. 1991;30(1):37–41.
79. Bauer G, Niedermeyer E. Acute convulsions. Clin Electroencephalogr.
1979;10(3):127–144.
80. Gastaut H, Rohmer F, Cossette A, et al. Introduction to the study
of functional generalized epilepsies. In: Gastaut H, Jasper HH,
Bancaud J, et al., eds. The Physiopathogenesis of the Epilepsies.
Springfield, IL: Charles C Thomas; 1969:15–25.
Chapter 21 ■ Metabolic Disorders and EEG 407
81. Plum F, Posner JB. Diagnosis of Stupor and Coma. Philadelphia:
F. A. Davis; 1966.
82. Adams RD, Foley JM. Neurological complications of liver disease.
Publ Assoc Res Neuro Ment Dis. 1953;32:198–237.
83. Hughes JR. Principles in the differentiation of atypical spike-
waves and triphasic waves. Am J EEG Technol. 1990;30:309–312.
84. Klass DW. Principles in the differentiation of atypical spike-waves
and triphasic waves. Am J EEG Technol. 1990;30:313–314.
85. Ghigo J. Principles in the differentiation of atypical spike-waves
and triphasic waves: a technologist’s view. Am J EEG Technol.
1990;30:315–316.
86. Ficker DM, Westmoreland BF, Sharbrough FW. Epileptiform
abnormalities in hepatic encephalopathy. J Clin Neurophysiol.
1997;14(3):230–234.
87. Eleftheriadis N, Fourla E, Eleftheriadis D, et al. Status epilepticus
as a manifestation of hepatic encephalopathy. Acta Neurol Scand.
2003;107(2):142–144.
88. Victor M. Neurologic changes in liver disease. In: Plum F, ed.
Brain Dysfunction in Metabolic Disorders. New York: Raven Press;
1974:1–12.
89. Victor M, Adams RD, Cole M. The acquired (non-Wilsonian)
type of chronic hepatocerebral degeneration. Medicine. 1965;
44(5):345–396.
90. Nevsimalova S, Marecek Z, Roth B. An EEG study of Wilson’s dis-
ease. Findings in patients and heterozygous relatives.
Electroencephalogr Clin Neurophysiol. 1986;64(3):191–198.
91. Giagheddu M, Tamburini G, Piga M, et al. Comparison of MRI,
EEG, EPs and ECD-SPECT in Wilson’s disease. Acta Neurol
Scand. 2001;103(2):71–81.
92. Cadilhac J. The EEG in renal insufficiency. In: Remond A,
ed. Handbook of Electroencephalography and Clinical
Neurophysiology. Amsterdam: Elsevier; 1976:351–369.
93. Grill MF, Maganti R. Cephalosporin-induced neurotoxicity: clin-
ical manifestations, potential pathogenic mechanisms, and the
role of electroencephalographic monitoring. Ann Pharmacother.
2008;42(12):1843–1850.
94. Chuang CL, Chen KP, Kwan SY, et al. Creutzfeldt-Jakob-like EEG
in a patient with end-stage renal failure. Nephrol Dial Transplant.
2004;19(1):252–254.
95. Lacerda G, Krummel T, Sabourdy C, et al. Optimizing therapy of
seizures in patients with renal or hepatic dysfunction. Neurology.
2006;67(12 suppl 4):S28–S33.
96. Boggs JG. Seizures in medically complex patients. Epilepsia.
1997;38(suppl 4):S55–S59.
97. McEnery PT, Nathan J, Bates SR, et al. Convulsions in children
undergoing renal transplantation. J Pediatr. 1989;115(4):
532–536.
98. Glaser GH. Brain dysfunction in uremia. In: Plum F, ed. Brain
Dysfunction in Metabolic Disorders. New York: Raven Press;
1974:3173–3197.
99. Zuckermann EG, Glaser GH. Urea-induced myoclonic seizures.
An experimental study of site of action and mechanism. Arch
Neurol. 1972;27(1):14–28.
100. Hughes JR. Correlations between EEG and chemical changes in
uremia. Electroencephalogr Clin Neurophysiol. 1980;48(5):583–594.
101. Lavy S, Aviram A. EEG changes in uraemic patients undergoing
regular haemodialysis. Electroencephalogr Clin Neurophysiol.
1969;27(2):217.
102. Spehr W, Sartorius H, Berglund K, et al. EEG and haemodialysis. A
structural survey of EEG spectral analysis, Hjorth’s EEG descrip-
tors, blood variables and psychological data. Electroencephalogr
Clin Neurophysiol. 1977;43(6):787–797.
408 Part III ■ Clinical EEG: General
103. Markand ON. Electroencephalography in diffuse
encephalopathies. J Clin Neurophysiol. 1984;1(4):357–407.
104. Zysno E, Reichenmiller HE, Durr F. EEG studies in uraemic
patients during extracorporeal haemodialysis or intermittent
peritoneal dialysis. Electroencephalogr Clin Neurophysiol.
1967;23(4):384.
105. Bourne JR, Ward JW, Teschan PE, et al. Quantitative assessment
of the electroencephalogram in renal disease. Electroencephalogr
Clin Neurophysiol. 1975;39(4):377–388.
106. Kiley JE, Woodruff MW, Pratt KI. Evaluation of encephalopathy
by EEG frequency analysis in chronic dialysis patients. Clin
Nephrol. 1976;5(6):245–250.
107. Balzar E, Saletu B, Khoss A, et al. Quantitative EEG: investigation
in children with end stage renal disease before and after
haemodialysis. Clin Electroencephalogr. 1986;17(4):195–202.
108. Fusco L, Picca S, Rizzoni G, et al. Long-term EEG monitoring in
uremic children on chronic dialysis treatment. Eur Neurol.
1991;31(4):193–198.
109. Zysno EA. The EEG in uraemia. Electroencephalogr Clin
Neurophysiol. 1970;29(2):212–213.
110. Reichenmiller HE, Reinhard U, Durr F. Sleep EEG and uraemia.
Electroencephalogr Clin Neurophysiol. 1971;30(3):263–264.
111. Jacob JC, Gloor P, Elwan OH, et al. Electroencephalographic
changes in chronic renal failure. Neurology. 1965;15:419–429.
112. Kerr DN, Osselton JW. The EEG as a monitor of cerebral distur-
bance during fast and slow haemodialysis of patients in chronic
renal failure. Electroencephalogr Clin Neurophysiol. 1967;
23(5):488.
113. Noriega-Sanchez A, Martinez-Maldonado M, Haiffe RM. Clinical
and electroencephalographic changes in progressive uremic
encephalopathy. Neurology. 1978;28(7):667–669.
114. Chokroverty S, Gandhi V. Electroencephalograms in patients
with progressive dialytic encephalopathy. Clin Electroencephalogr.
1982;13(2):122–127.
115. Yardimci N, Colak T, Sevmis S, et al. Neurologic complications
after renal transplant. Exp Clin Transplant. 2008;6(3):224–228.
116. Hamiwka LD, Midgley JP, Hamiwka LA. Seizures in children after
kidney transplantation: has the risk changed and can we predict
who is at greatest risk? Pediatr Transplant. 2008;12(5):527–530.
117. Sibai BM. Diagnosis, prevention, and management of eclampsia.
Obstet Gynecol. 2005;105(2):402–410.
118. Lain KY, Roberts JM. Contemporary concepts of the pathogene-
sis and management of preeclampsia. JAMA. 2002;287(24):
3183–3186.
119. Roberts JM, Taylor RN, Goldfien A. Clinical and biochemical evi-
dence of endothelial cell dysfunction in the pregnancy syndrome
preeclampsia. Am J Hypertens. 1991;4(8):700–708.
120. Ferris TF. Pregnancy, preeclampsia, and the endothelial cell.
N Engl J Med. 1991;325(20):1439–1440.
121. Will AD, Lewis KL, Hinshaw DB Jr, et al. Cerebral vasoconstric-
tion in toxemia. Neurology. 1987;37(9):1555–1557.
122. Trommer BL, Homer D, Mikhael MA. Cerebral vasospasm and
eclampsia. Stroke. 1988;19(3):326–329.
123. Sheehan HL, Lynch JB. Pathology of Toxemia of Pregnancy.
Edinburgh: Churchill Livingstone; 1973.
124. Dahmus MA, Barton JR, Sibai BM. Cerebral imaging in eclamp-
sia: magnetic resonance imaging versus computed tomography.
Am J Obstet Gynecol. 1992;167(4 Pt 1):935–941.
125. Schwartz RB, Feske SK, Polak JF, et al. Preeclampsia-eclampsia:
clinical and neuroradiographic correlates and insights into the
pathogenesis of hypertensive encephalopathy. Radiology.
2000;217(2):371–376.
126. Guieu JD, Milbled G, Pruvot P. Unusual EEG events during a case
of puerperal eclampsia. Electroencephalogr Clin Neurophysiol.
1976;40:549(abst).
127. Calloni MV, Porazzi D, Rovetta P. Cerebral dysfunction in
eclampsia: a longitudinal EEG study. Electroencephalogr Clin
Neurophysiol. 1981;51:79P(abst).
128. Grimes DA, Ekbladh LE, McCartney WH. Cortical blindness in
preeclampsia. Int J Gynaecol Obstet. 1980;17(6):601–603.
129. Gurjinder PS, Dhand UK, Chopra JS. Balint’s syndrome follow-
ing eclampsia. Clin Neurol Neurosurg. 1989;91(2):161–165.
130. Thomas SV, Somanathan N, Radhakumari R. Interictal EEG
changes in eclampsia. Electroencephalogr Clin Neurophysiol.
1995;94(4):271–275.
131. Osmanagaoglu MA, Dinc G, Osmanagaoglu S, et al. Comparison
of cerebral magnetic resonance and electroencephalogram find-
ings in pre-eclamptic and eclamptic women. Aust N Z J Obstet
Gynaecol. 2005;45(5):384–390.
132. Sibai BM, Spinnato JA, Watson DL, et al. Effect of magnesium
sulfate on electroencephalographic findings in preeclampsia-
eclampsia. Obstet Gynecol. 1984;64(2):261–266.
133. Sibai BM. Diagnosis and management of gestational hyperten-
sion and preeclampsia. Obstet Gynecol. 2003;102(1):181–192.
134. Brophy E, Brophy MH. Pyridoxal phosphate normalization of the
EEG in eclampsia. Electroencephalogr Clin Neurophysiol. 1991;
79:36P(abst).
135. Ledermair O, Niedermeyer E. Posteklamptische Epilepsie.
Geburtshilfe Frauenheilkd. 1956;16(8):679–685.
136. Corriol J, Papy J, Rohner JJ, et al. Electroclinical correlations
established during tetanic manifestations induced by parathyroid
removal in the dog. In: Gastaut H, Jasper HH, Bancaud J, et al.
eds. The Physiopathogenesis of the Epilepsies. Springfield, IL:
Charles C Thomas; 1969:3128–3140.
137. Glaser GH, Levy L. Seizures and idiopathic hypoparathyroidism.
A clinical-electroencephalographic study. Epilepsia. 1960;1:
454–465.
138. Goldberg HH. The electroencephalogram in hypocalcemic syn-
dromes. Electroencephalogr Clin Neurophysiol. 1959;11:398.
139. Lynch BJ, Rust RS. Natural history and outcome of neonatal
hypocalcemic and hypomagnesemic seizures. Pediatr Neurol.
1994;11(1):23–27.
140. Kossoff EH, Silvia MT, Maret A, et al. Neonatal hypocalcemic
seizures: case report and literature review. J Child Neurol.
2002;17(3):236–239.
141. Kurtz D. The EEG in parathyroid dysfunction. In: Remond A, ed.
Handbook of Electroencephalography and Clinical Neurophysiology.
Amsterdam: Elsevier; 1976:77–87.
142. Guberman A, Jaworski ZF. Pseudohypoparathyroidism and
epilepsy: diagnostic value of computerized cranial tomography.
Epilepsia. 1979;20(5):541–553.
143. Swash M, Rowan AJ. Electroencephalographic criteria of
hypocalcemia and hypercalcemia. Arch Neurol. 1972;26(3):
218–228.
144. Devlesc Howard AB, Manyam BV. Electroencephalography in
Fahr’s disease. Electroencephalogr Clin Neurophysiol. 1989;73:
68P(abst).
145. Spatz R, Kugler J, Angstwurm H. The EEG in hypercalcemia
(author’s transl). EEG EMG Z Elektroenzephalogr Elektromyogr
Verwandte Geb. 1977;8(2):70–76.
146. Allen EM, Singer FR, Melamed D. Electroencephalographic
abnormalities in hypercalcemia. Neurology. 1970;20(1):15–22.
147. Moure JM. The electroencephalogram in hypercalcemia. Arch
Neurol. 1967;17(1):34–51.

148. Strickland NJ, Bold AM, Medd WE. Bronchial carcinoma with
hypercalcaemia simulating cerebral metastases. Br Med J.
1967;3(5565):590–592.
149. Huott AD, Madison DS, Niedermeyer E. Occipital lobe epilepsy.
A clinical and electroencephalographic study. Eur Neurol.
1974;11(6):325–339.
150. Kaplan PW. Reversible hypercalcemic cerebral vasoconstriction
with seizures and blindness: a paradigm for eclampsia? Clin
Electroencephalogr. 1998;29(3):120–123.
151. Zwang HJ, Cohn D. Electroencephalographic changes in acute
water intoxication. Clin Electroencephalogr. 1981;12(1):35–40.
152. Okura M, Okada K, Nagamine I, et al. Electroencephalographic
changes during and after water intoxication. Jpn J Psychiatry
Neurol. 1990;44(4):729–734.
153. Crawford JD, Dodge P. Complications of fluid therapy in neuro-
logic disease; water intoxication and hypertonic dehydration.
Pediatr Clin North Am. 1964;11:1029–1052.
154. Nakayama Y, Tanaka A, Naritomi K, et al. Hyponatremia-induced
metabolic encephalopathy caused by Rathke’s cleft cyst: a case
report. Clin Neurol Neurosurg. 1999;101(2):114–117.
155. Itoh N, Matsui N, Matsui S. Periodic lateralized epileptiform dis-
charges in EEG during recovery from hyponatremia: a case
report. Clin Electroencephalogr. 1994;25(4):164–169.
156. Maruyama T, Tabata K, Nakagawa S, et al. A case of acute water
intoxication showing triphasic waves on EEG. Rinsho
Shinkeigaku. 1991;31(5):523–527.
157. Azuma H, Akechi T, Furukawa TA. Absence status associated
with focal activity and polydipsia-induced hyponatremia.
Neuropsychiatr Dis Treat. 2008;4(2):495–498.
158. Castilla-Guerra L, del Carmen Fernandez-Moreno M, Lopez-
Chozas JM, et al. Electrolytes disturbances and seizures. Epilepsia.
2006;47(12):1990–1998.
159. Van der Helm-van Mil AH, Van Vugt JP, Lammers GJ, et al.
Hypernatremia from a hunger strike as a cause of osmotic myeli-
nolysis. Neurology. 2005;64(3):574–575.
160. Shils ME. Magnesium, calcium, and parathyroid hormone inter-
actions. Ann N Y Acad Sci. 1980;355:165–180.
161. Riggs JE. Neurologic manifestations of electrolyte disturbances.
Neurol Clin. 2002;20(1):227–239, vii.
162. Leicher CR, Mezoff AG, Hyams JS. Focal cerebral deficits in
severe hypomagnesemia. Pediatr Neurol. 1991;7(5):380–381.
163. Victor M, Adams RD, Collins GH. The Wernicke-Korsakoff
Syndrome. Philadelphia, PA: F.A. Davis; 1971.
164. Dreyfus PM, Victor M. Effects of thiamine deficiency on the cere-
bral nervous system. Am J Clin Nutr. 1961;9:414–425.
165. Frantzen E. Wernicke’s encephalopathy. 3 Cases occurring in con-
nection with severe malnutrition. Acta Neurol Scand. 1966;42(4):
426–441.
166. Fournet A, Lanternier J. Constatations electroencephalo-
graphiques dans 17 cas d’encephalopathie de Gayet-Wernicke.
Rev Neurol (Paris). 1956;94(5):644–645.
167. Keyser A, De Bruijn SF. Epileptic manifestations and vitamin B1
deficiency. Eur Neurol. 1991;31(3):121–125.
168. Fattal-Valevski A, Kesler A, Sela BA, et al. Outbreak of life-threat-
ening thiamine deficiency in infants in Israel caused by a defec-
tive soy-based formula. Pediatrics. 2005;115(2):e233–e238.
169. Fattal-Valevski A, Bloch-Mimouni A, Kivity S, et al. Epilepsy in
children with infantile thiamine deficiency. Neurology. 2009;
73(11):828–833.
170. Srikantia SG, Reddy MV, Krishnaswamy K. Electro -
encephalographic patterns in pellagra. Electroencephalogr Clin
Neurophysiol. 1968;25(4):386–388.
Chapter 21 ■ Metabolic Disorders and EEG 409
171. Coursin DB. Convulsive seizures in infants with pyridoxine-defi-
cient diet. J Am Med Assoc. 1954;154(5):406–408.
172. Reilly RH, Killam KF, Jenney EH, et al. Convulsant effects of iso-
niazid. J Am Med Assoc. 1953;152(14):1317–1321.
173. Bejsovec M, Kulenda Z, Ponca E. Familial intrauterine convul-
sions in pyridoxine dependency. Arch Dis Child. 1967;
42(222):201–207.
174. Hunt AD Jr, Stokes J Jr, Mc CW, et al. Pyridoxine dependency:
report of a case of intractable convulsions in an infant controlled
by pyridoxine. Pediatrics. 1954;13(2):140–145.
175. Waldinger C, Berg RB. Signs of pyridoxine dependency manifest
at birth in siblings. Pediatrics. 1963;32:161–168.
176. Kretsch MJ, Sauberlich HE, Newbrun E. Electroencephalographic
changes and periodontal status during short-term vitamin B-6
depletion of young, nonpregnant women. Am J Clin Nutr.
1991;53(5):1266–1274.
177. Diefenbach WC, Beyers MR, Meyer LM. The encephalographic
and psychologic changes in pernicious anemia. Acta Haematol.
1953;9(4):201–208.
178. Walton JN, Kiloh LG, Osselton JW, et al. The electroencephalo-
gram in pernicious anaemia and subacute combined degenera-
tion of the cord. Electroencephalogr Clin Neurophysiol. 1954
Feb;6(1):45-64.
179. Andersen K, Stigsby B. Quantitative EEG analysis in patients with
pernicious anemia. Electroencephalogr Clin Neurophysiol.
1985;61:33P(abst).
180. Biancheri R, Cerone R, Rossi A, et al. Early-onset cobalamin C/D
deficiency: epilepsy and electroencephalographic features.
Epilepsia. 2002;43(6):616–622.
181. Engel GL, Margolin SG. Neuropsychiatric disturbance in internal
disease: metabolic factors and electroencephalographic correla-
tionss. Arch Intern Med. 1942;70:236–259.
182. Hoffman WC, Lewis R, Thorn GW. The electroencephalogram in
Addison’s disease. Bull Johns Hopkins Hosp. 1942;70:335–361.
183. Vas GA, Cracco JB. Diffuse encephalopathies. In: Daly DD, Pedley
TA, eds. Current Practice of Clinical Electroencephalography. 2nd
ed. New York: Raven Press; 1990:371–399.
184. Garcia-Austt E, Torrents E, Mussio-Fournie F. The electroen-
cephalogram in Cushings disease. Electroencephalogr Clin
Neurophysiol. 1950;2:103.
185. Pine I, Engel FL, Schwartz TB. The electroencephalogram in
ACTH and cortisone treated patients. Electroencephalogr Clin
Neurophysiol. 1951;3(3):301–310.
186. Glaser GH, Komfeld DS, Knight RP. Intravenous hydrocortisone,
corticotropin and the electroencephalograms. Arch Neurol
Psychiatry. 1955;73:338–344.
187. Raab W, Smithwick RH. Pheochromocytoma with hypothalamic
manifestations and excessive hypermetabolism; a case report. J
Clin Endocrinol Metab. 1949;9(8):782–790.
188. Kennedy JM, Thomson AP, Whitfield IC. Coma and electro-
encephalographic changes in hypopituitarism. Lancet. 1955;
269(6896):907–908.
189. Lindsley DB, Rubenstein BB. Relationships between brain poten-
tials and some other physiological variables. Proc Soc Exp Biol
(New York). 1937;35:558–563.
190. Rubin MA, Cohen HC, Hoagland H. The effect of artificially
raised metabolic rate and the electroencephalogram of schizo-
phrenic patients. Endocrinology. 1937;25:536–540.
191. Ross DA, Schwab RJ. The cortical alpha rhythm in thyroid disor-
ders. Endocrinology. 1939;25:75–79.
192. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. 1st ed.
Cambridge, MA: Addison-Wesley; 1941.
410 Part III ■ Clinical EEG: General
193. Vague J, Gastaut H, Favier G, Rey J. Electroencephalography in
endocrine pathology. Mars Med. 1952;89(1):1–17.
194. Vague J, Gastaut H, Codaccioni JL, et al. EEG in thyroid diseases.
Ann Endocrinol (Paris). 1957;18(6):996–1008.
195. Wilson WP, Johnson JE, Feist FW. Thyroid hormone and brain
function. II. Changes in photically elicited EEG responses follow-
ing the administration of triiodothyronine to normal subjects.
Electroencephalogr Clin Neurophysiol. 1964;16:329–331.
196. Skanse B, Nyman GE. Thyrotoxicosis as a cause of cerebral dys-
rhythmia and convulsive seizures. Acta Endocrinol (Copenh).
1956;22(3):246–263.
197. Jabbari B, Huott AD. Seizures in thyrotoxicosis. Epilepsia.
1980;21(1):91–96.
198. Miyazaki Y, Fukuoka K, Murase T, et al. Abnormal MRI and EEG
findings in thyroid storm resulting from Graves’disease. Thyroid.
2008;18(10):1131–1132.
199. Cadilhac J. The EEG in thyroid dysfunction. In: Remond A, ed.
Handbook of Electroencephalography and Clinical Neurophysiology,
Vol 15C: Metabolic, Endocrine and Toxic Diseases. Amsterdam:
Elsevier; 1976:70–76.
200. Berlit P, Rakicky J. The EEG in metabolic and endocrine
encephalopathies. Clin Neurophysiol. 1991;78:36P(abst).
201. Chong JY, Rowland LP, Utiger RD. Hashimoto encephalopathy:
syndrome or myth? Arch Neurol. 2003;60(2):164–171.
202. Kothbauer-Margreiter I, Sturzenegger M, Komor J, et al.
Encephalopathy associated with Hashimoto thyroiditis: diagnosis
and treatment. J Neurol. 1996;243(8):585–593.
203. Tsai MH, Lee LH, Chen SD, et al. Complex partial status epilepti-
cus as a manifestation of Hashimoto’s encephalopathy. Seizure.
2007;16(8):713–716.
204. Aydin-Ozemir Z, Tuzun E, Baykan B, et al. Autoimmune thyroid
encephalopathy presenting with epilepsia partialis continua. Clin
EEG Neurosci. 2006;37(3):204–209.
205. Shaw PJ, Walls TJ, Newman PK, et al. Hashimoto’s encephalopa-
thy: a steroid-responsive disorder associated with high anti-thy-
roid antibody titers—report of 5 cases. Neurology. 1991;41(2 Pt 1):
228–233.
206. Henchey R, Cibula J, Helveston W, et al. Electroencephalographic
findings in Hashimoto’s encephalopathy. Neurology. 1995;45(5):
977–981.
207. Schultz MA Schulte FJ, Akiyama Y, et al. Development of EEG
sleep phenomena in hypothyroid infants. Electroencephalogr Clin
Neurophysiol. 1968;25:351–358.
208. Nieman EA. The electroencephalogram in congenital hypothy-
roidism: a study of 10 cases. J Neurol Neurosurg Psychiatry.
1961;24:50–57.
209. Scarpalezos S, Lygidakis C, Papageorgiou C, et al. Neural and
muscular manifestations of hypothyroidism. Arch Neurol. 1973;
29(3):140–144.
210. Forester CF. Coma in myxedema. Report of a case and review of
the world literature. Arch Intern Med. 1963;111:734–743.
211. River Y, Zelig O. Triphasic waves in myxedema coma. Clin
Electroencephalogr. 1993;24(3):146–150.
212. Wynn D, Lagerlund T, Mokri B, et al. Periodic complexes in
hypothyroidism masquerading as Jakob-Creutzfeldt disease: a
case report. Electroencephalogr Clin Neurophysiol. 1989;72:
31P(abst).
213. Gibbs FA, Reid DE. The EEG in pregnancy. Am J Obstet Gynecol.
1942;44:672–675.
214. Keunen RW, Vliegen JH, van der Pol DA, et al. The electroen-
cephalogram during normal third trimester pregnancy and six
months postpartum. Br J Obstet Gynaecol. 1997;104(2):
256–258.
215. Dusser de Barenne D, Gibbs FA. Variations in the electroen-
cephalogram during the menstrual cycle. Am J Obstet Gynecol.
1942;44:687–690.
216. Holmes GL, Donaldson JO. Effect of sexual hormones on the
electroencephalogram and seizures. J Clin Neurophysiol. 1987;
4(1):1–22.
217. Goldberg A. Acute intermittent porphyria: a study of 50 cases. Q
J Med. 1959;28(110):183–209.
218. Dow RS. The electroencephalographic changes in acute intermit-
tent porphyria. Electroencephalogr Clin Neurophysiol. 1961;13:
425–437.
219. Bylesjo I, Forsgren L, Lithner F, Boman K. Epidemiology and
clinical characteristics of seizures in patients with acute intermit-
tent porphyria. Epilepsia. 1996;37(3):230–235.
220. Birchfield RI, Cowger ML. Acute intermittent porphyria with
seizures. Anticonvulsant medication-induced metabolic changes.
Am J Dis Child. 1966;112(6):561–565.
221. Krauss GL, Simmons-O’Brien E, Campbell M. Successful treat-
ment of seizures and porphyria with gabapentin. Neurology.
1995;45(3 Pt 1):594–595.
222. Niedermeyer E, Ribeiro M, Hertz S. Mixed-type encephalopathies:
preliminary considerations. Clin Electroencephalogr. 1999;30(1):
12–15.
223. Austen FK, Carmichael MW, Adams RD. Neurologic manifesta-
tions of chronic pulmonary insufficiency. N Engl J Med.
1957;257(13):579–590.
224. Gastaut H, Tassinari CA, Duron B. Polygraphic study of diurnal
and nocturnal (hypnic and respiratory) episodal manifestations
of Pickwick syndrome. Rev Neurol (Paris). 1965;112(6):
568–579.
225. Jung R, Kuhlo W. Neurophysiological studies of abnormal
night sleep in the Pickwickian syndrome. In: Akert K, Bally C,
Schade JP, eds. Sleep Mechanisms. Amsterdam: Elsevier; 1965:
140–159.

EEG and Craniocerebral Trauma
A
craniocerebral trauma raises the important question of
the degree of cerebral disturbance. This question may
be answered easily in cases of commotio cerebri or
slight concussion of the brain; this is a reversible impairment of
consciousness of brief duration with no clinical evidence of any
gross structural change in the brain substance. The question of
brain function becomes more important in cases of cerebral
contusion after more severe injury with prolonged unconscious-
ness and signs of brainstem dysfunction. Brainstem damage may
result from primary brainstem injury (1,2) or from secondary
brainstem lesion due to downward transtentorial herniation (3).
It has been shown that brainstem injury does not exist in isola-
tion; it is merely one aspect of diffuse brain damage (4).
The clinical parameter of the grade of disintegration of brain
function and impairment of the brainstem is demonstrated by the
development of an acute secondary midbrain and bulbar brain
syndrome (BBS). The symptoms of the well-known rostrocaudal
deterioration, first described by Mc-Nealy and Plum (5) and Plum
and Posner (6) and modified by Gerstenbrand and Lücking (7),
allow a clear clinical statement about the depth of posttraumatic
coma. In addition to the neurologic examination, cranial com-
puted tomography (CT) and magnetic resonance imaging (MRI)
easily identify a space-occupying lesion and sometimes demon-
strates the displacement of the brainstem (8,9). These methods,
however, fail to give any information about cerebral activity.
Therefore, the electroencephalogram (EEG) is still important in
the diagnosis of traumatic cerebral lesions, especially in the assess-
ment of the degree of cortical activity, which shows reasonably
good correlation with the depth of posttraumatic coma (10–16).
Chronic stages of craniocerebral trauma demonstrate a
diminished electroencephalographic–neurologic correlation
(17–20). However, there is an approximate correlation between
the EEG and clinical improvement, especially in patients who
have had systematic follow-up studies of EEG and clinical
examination (21–24). Occasionally, the EEG may return to nor-
mal when neurologic or psychiatric abnormalities persist, a dis-
parity that indicates a bad prognosis (18,25,26). By contrast, in
some patients with normal clinical findings, an abnormal EEG
may be the forerunner of an intracranial complication such as
posttraumatic epilepsy (21).
EARLIER EEG WORK IN
CRANIOCEREBRAL TRAUMA
Williams (27) studied a few patients after mild craniocerebral
trauma and observed that normal records can be obtained
within a few hours of injury. Further electroencephalographic
CHAPTER
ERIK RUMPL 22
studies in brain injuries were done by Dow et al. (28), who
included patients with more severe injuries and repeated studies
during the recovery of the patients. Dawson et al. (29) continued
to observe patients and studied the evolution of abnormalities
with repeated records. Extensive work in this field was done by
Meyer-Mickeleit (19), Schneider and Hubach (30), Courjon and
Scherzer (21), and Koufen and Dichgans (23).
The impairment of consciousness to any degree was usually
found to be accompanied by an abnormal EEG (19,28–32).
Most of these authors directed their attention to the frequency
of the basic activity and to focal, general, or epileptic abnormal-
ities, but no further distinction of the EEG pattern was made.
High-voltage delta activity was regarded as the electroen-
cephalographic correlate of coma, which in turn was considered
an equivalent of deep sleep (33–35). Further studies in coma
(36–38) made it clear that states of coma were associated with a
variety of EEG patterns, often with the criteria of sleep
(12,14,39). The manifestation of sleep patterns appeared to be
of prognostic significance, supported by the results of overnight
or prolonged EEG recordings (40–42). The role of reactivity in
the assessment of the depth of coma was first pointed out by
Fischgold et al. (43) and is an important prognostic indicator
for good outcome (44). In addition to these findings, asymme-
try in reactivity was considered a reliable lateralizing sign (45).
Important reviews concerning posttraumatic comatose
states have given detailed insight into the EEG in altered states
of consciousness (10,46,47). Early and late changes of the EEG
after craniocerebral trauma were discussed extensively by
Courjon and Scherzer (21) and Stockard et al. (48).
EXPERIMENTAL WORK
Williams and Denny-Brown (49) found an immediate reduc-
tion in amplitude (sometimes approaching isoelectricity) in
the EEG in cats subjected to experimental head concussion.
After a short period of delta activity, the EEG returned to
normal within a period of seconds to a few minutes. Dow et al.
(28) studied the physical factors responsible for the injury in
humans. These factors appeared to be of the same order as
those necessary to produce experimental concussion in ani-
mals (50). Interestingly, the threshold velocity that provoked
concussion at the moment of impact in animals also marked
the limit between the average velocity in patients with border-
line and abnormal EEG (28). Other than this finding, patho-
logic observations made in humans could not be convincingly
duplicated in experimental studies. This might be due to the
smaller mass of brain in animals and the restriction in impact
411
412 Part III ■ Clinical EEG: General
techniques (51,52). Even in standardized animal experiments,
the analysis of the EEG did not provide adequate indices of
injury severity, and there was no correlation found with the
duration of coma (53).
Following the convulsive theory of slight concussion (54),
the energy imported to the brain may generate turbulent rota-
tory and other movements of the cerebral hemispheres provok-
ing a convulsive disorder, which might present a mild form of
cerebral concussion with a sudden short loss of consciousness,
paralysis of reflex activity, and loss of memory (commotio cere-
bri). This theory may explain the observation that the EEG
returns to normal within 24 hours in these cases.
Some closer correlations to EEG changes in comatose states
in humans were found in other animal experiments producing
localized destructive brain lesions or using local electrical stim-
ulation. Local lesions in the white matter and in thalamic struc-
tures were immediately followed by the appearance of delta
waves, which were accompanied by spindles in the case of thal-
amic lesions. Slow activity developed gradually in mesen-
cephalic reticular formation lesions; it was absent in cortical
lesions (55). EEG changes similar to those seen in thalamic
destructive lesions were found after stimulation of thalamic
nuclei (56,57). Stimulation at the level of the mesencephalic
reticulum induced behavior and EEG arousal in sleeping in oth-
erwise intact animals, whereas destructive lesions at mesen-
cephalic levels caused coma and EEG activity reminiscent of
different types of sleep (58). All these mechanisms may be
involved in posttraumatic comatose states as long as delta waves
and spindles are generated by otherwise relatively normal corti-
cal neurons.
EEG CHANGES IN THE ACUTE STAGES
OF POSTTRAUMATIC COMA
Comatose posttraumatic states may be associated with a multi-
tude of abnormal waveforms other than delta (12,14,36,38).
These abnormal waveforms, especially the different patterns of
sleep, depend above all on the patient’s level of consciousness
(46). Therefore, a detailed neurologic description of comatose
patients is necessary to define coma (59) in order to establish a
sufficient correlation between different EEG changes and differ-
ent stages of coma.
CLASSIFICATION OF EEG
Different parameters have been used in the breakdown of EEG
signs in comatose states (10–12,14,21,36,60–62). To increase the
value of the EEG in posttraumatic coma, all these parameters
were integrated and correlated with the stages of rostrocaudal
deterioration (13).
The EEG abnormalities may be classified into five grades (13):
grade 1, predominant alpha and little theta activity; grade 2,
predominant theta and little delta activity; grade 3, predominant
high-voltage rhythmic and arrhythmic delta and subdelta acti-
vity; grade 4, diffuse, mostly low-voltage delta and subdelta
activity and low-voltage activity only recognizable with increased
amplification (3.5 V/mm); and grade 5, isoelectric record.
Other signs apart from these grades can also be noted in the
EEG: (i) superimposed fast activity (6 to 18 c/sec localized over
the frontal region or diffusely spread); (ii) normal-looking
sleep records; (iii) diffuse slowing accompanied by “typical
sleep potentials” (spindles, vertex sharp waves, K complexes);
(iv) diffuse slowing accompanied by altered “sleep potentials,”
listed under “atypical sleep potentials”; (v) spontaneously
alternating EEG patterns with a rapid succession of low-
voltage delta activity and high-voltage slow waves (this activity
can be classified as delta bursts [0.5 to 2 seconds] and short [2
and 5 seconds] and long [more than 5 seconds in duration]
runs of delta with frequencies of 0.5 to 4 c/sec); (vi) “lateral-
ized signs” such as local slow activity (“local slowing”), unilat-
eral predominant slow activity, unilateral low-voltage output,
and unilateral depression of superimposed fast activity listed
under the heading of “unilateral predominant slowing,” asym-
metries of “typical or atypical sleep potentials,” and asymme-
tries of response to external stimuli; (vii) four types of reaction
to external stimuli: (1) appearance of widespread 1 to 7 c/sec
activity, (2) blocking of slow activity, (3) appearance of alter-
nating EEG patterns, and (4) appearance or disappearance of
“typical or atypical sleep potentials”; and (viii) epileptiform
patterns: (1) focal or generalized runs of spikes and sharp spike
waves, (2) periodic lateralized epileptiform discharges
(PLEDs), and (3) triphasic waves.
Reactivity may be immediate or delayed. Delayed reactiv-
ity is a slow EEG response inducing a new pattern under the
influence of repetitive stimulation (10). Fischgold and Mathis
(36) emphasize the usefulness of auditory and painful stimuli
in contrast to the insignificant effect of visual stimulation.
In lighter stages of coma, stimulation may also provoke
extracerebral changes such as muscle activity and respiratory
artifacts.
CLASSIFICATION OF COMA
Physiologically, the loss of consciousness implies that the
patient has suffered widespread dysfunction of cerebral hemi-
spheres or the brainstem or both. When there is a diffuse
swelling of the brain or bilateral hemispheric lesions, the cere-
bral hemispheres, the diencephalon, and the adjoining mid-
brain are displaced downward through the tentorial notch. In
this case, the herniation through the tentorium is symmetrical
(central tentorial herniation). If downward displacement con-
tinues, the contents of the posterior fossa will finally be pressed
through the foramen magnum. This brain shift evokes a num-
ber of characteristic clinical signs that have been described by
various authors (5–7). The most prominent clinical signs of
central herniation are listed in Table 22.1. According to
Gerstenbrand and Lücking (7) posttraumatic comatose states
can be broken down into six stages, including four stages of
midbrain syndrome (MBS) and two stages of BBS. With regard
to the terminology of Plum and Posner (6), MBS 1 and MBS 2
correspond to the early diencephalic stage, MBS 3 to the late
diencephalic stage, MBS 4 to the midbrain-upper pons stage,
BBS 1 to the lower pontine-upper medullary stage, and BBS 2
to the medullary stage.
 Chapter 22 ■ EEG and Craniocerebral Trauma 413

Tabl e 2 2 . 1

Clinical Signs in Posttraumatic Patients at the Different Stages of the Midbrain Syndrome (MBS)
and Bulbar Brain Syndrome (BBS)

MBS 1 MBS 2 MBS 3 MBS 4 BBS 1 BBS2

Spontaneous Nonstereotyped Nonstereotyped Decorticate Decerebrate Flaccidity Flaccidity
limb postures movements movements posturing posturing
in the arms in the arms
and legs
Extension of
the legs
Motor response Nonstereotyped Nonstereotyped Decorticate Decerebrate Decerebrate
to pain withdrawal of withdrawal of response response response
the limbs the arms No response
Extensor
response of
the legs
Eye position Roving Roving, more Immobile, Immobile, Immobile, Immobile,
movements irregular straight divergent divergent divergent
ahead
Pupil size and Normal Normal small Small Enlarged Large Large
reaction
Reacting Reacting Small range Small range Unreacting Unreacting
of contraction of contraction
Respiration Normal Cheyne-Stoke Cheyne-Stokes Regular Ataxic No
pattern hyperven- respiration
tilation
Rapid regular
hyperven-
tilation

From Rumpl E, Prugger M, Bauer G, et al. Incidence and prognostic value of spindles in post-traumatic coma. Electroencephalogr Clin Neurophysiol. 1983;56:420–429.

In cases with expanding lesions in the lateral middle fossa or
temporal lobe, the medial edges of the uncus and hippocampal
gyrus are commonly pushed toward the midline and over the free
lateral edge of the tentorium, compressing the third cranial nerve
(early uncal herniation). During uncal herniation, a unilateral
dilated pupil is the most consistent finding (6). Other neurologic
signs may be present at this early stage and may provide hints
concerning the original hemispheric lesion. Once the pupil
dilates fully (late uncal herniation), the patients become coma-
tose, and abnormal postures and motor responses evolve (6,63).
Eventually, with increasing intracranial pressure, lateralizing
signs disappear and the patients show the clinical signs of MBS 4
or BBS 1.
Neurologic signs different from this expected rostrocaudal
deterioration may be found in cases of primary brainstem
injury. Maciver et al. (2) state that a primary brainstem lesion
can be suspected if the neurologic signs do not fit the classical
stages of rostrocaudal deterioration (i.e., the relatively intact
oculomotor functions are in contrast to decerebrate posturing
and respiratory abnormalities). The CT scan is diagnostic and
eliminates a supratentorial lesion causing secondary brainstem
dysfunction. The MRI scan shows the extent of brainstem dam-
age in the further course.
RELATION OF EEG ABNORMALITIES
TO THE STAGE OF COMA
The EEG patterns in the different stages of MBS and BBS are
presented in Figure 22.1. The significance of this observation is
somewhat hampered by the fact that the figure presents EEG
data obtained within 1 week after brain injury, thereby neglect-
ing the special behavior of spindle activity (see the section
“Relation of EEG Abnormalities to Outcome”). Nevertheless,
the figure demonstrates well the principal relation of EEG to
the stages of coma. There is a decrease in the number of differ-
ent EEG patterns related to the stages of MBS and BBS, indicat-
ing the increase of intracranial pressure. Unfavorable prognosis
is shown by the disappearance of “typical or atypical sleep
414 Part III ■ Clinical EEG: General

Figure 22.1. The EEG patterns at the different stages of acute traumatic secondary midbrain and bulbar brain syndrome.
see A–E above. Note the reduction in the variety of EEG patterns in the course of increasing rostrocaudal deterioration. At
bulbar brain syndrome stage 2, the EEG is isoelectric. The records were listed under one or more of the following cate-
gories: A/ T, predominant alpha and little theta activity; T/ D, predominant theta and little delta activity; D/ SD, predominant
diffuse rhythmic and arrhythmic delta and subdelta activity; DB, delta bursts; DS, short runs of delta; DL, long runs of
delta; TSP, typical sleep potentials; ASP, atypical sleep potentials; S, spindles; LS, local slowing; US, unilateral predominant
slowing; SFA, superimposed fast activity; MBS, midbrain syndrome; BBS, bulbar brain syndrome; N, number of records.
(From Rumpl, E. Elektro-neurologische Korrelationen in den frühen Phasen des posttraumatischen Komas. I. Das EEG in
den verschiedenen Phasen des akuten traumatischen sekundären. Mittelhirn und Bulbärhirnsyndroms. Z. EEG-EMG.
1979;10:148–157; and Rumpl E, Lorenzi E, Hackl JM, et al. The EEG at different stages of acute secondary traumatic mid-
brain and bulbar brain syndromes. Electroencephalogr Clin Neurophysiol. 1979;46:487–497, somewhat modified.)

potentials,” alternating patterns, and loss of reactivity (10,12–
14,21,41). In accordance with the mildly diffuse cortical distur-
bance, the EEG is slightly or moderately abnormal in MBS 1.
Stimuli to diencephalic or rostromesencephalic structures
(64) may induce delta bursts or short runs of delta waves. Local
slow activity can be seen only in an otherwise mildly abnormal
EEG, while increasing diffuse slowing may overwhelm local
abnormalities (65). Borderline EEG patterns occur at MBS 1 but
not in deeper stages of MBS. In MBS 2, the amount of slow activ-
ity increases. This increasing abnormality may be due to a direct
cortical disturbance (64) or due to a more remote effect from
deeper structures (64,66). However, “sleep or sleep-like poten-
tials” (see Figs. 22.6A and 22.7A) indicate a relatively intact cortex
(14). In MBS 2, the EEG shows the widest range of different EEG
patterns. In MBS 3, a clear reduction in the number of EEG pat-
terns is noted. A decrease of spontaneous alternating patterns
and “sleep potentials” has been thought to indicate an increasing
disturbance of the diencephalic and mesencephalic systems (12).
Increasing rostrocaudal deterioration is further character-
ized by the increased change of typical or atypical sleep poten-
tials (14) and the shift of superimposed fast activity, diffusely
spread in lighter stages of MBS, to the frontal regions (67). The
EEG patterns are still more simplified in MBS 4 and BBS 1. The
absence, reduction, and deterioration of sleep potentials or
alternating patterns are suggestive of marked damage at the
diencephalic level (39,41). The telencephalon (10) must also be
involved, considering the diffuse brain edema in advanced
stages of MBS. In BBS 1, low-voltage delta and subdelta activity
appears, followed by low-voltage cerebral activity, often recog-
nizable only with increased amplification. No electrical cerebral
activity can be observed in patients with BBS 2 (13). Patients
with an isoelectric EEG or an EEG with repeated isoelectric
periods will die (68). The high prognostic value of an EEG clas-
sification with a great number of different EEG patterns was
supported by the development of an EEG score, which included
many of these phenomena weighted according to their per-
ceived prognostic value and scored as dichotomous variables—
present or absent (69).
Unreactive alpha activity or “alpha-like” rhythms (70) are
not seen in patients in BBS 1 after supratentorial lesions causing
 Chapter 22 ■ EEG and Craniocerebral Trauma 415

secondary brainstem involvement. This pattern usually appears

in polytraumatic patients at this stage and thereby strongly sug-
gests an additional hypoxic or anoxic cerebral lesion. The same
suggestion can be made with theta-pattern coma, which was
found to be a variant of alpha coma (71). Usually, both patterns
are predictors of fatal outcome. Rhythmic coma in children
especially in the alpha frequency range has generally a better
prognosis than in adults (72).
Primary brainstem injuries may produce clinical symp-
toms that are hardly separable from symptoms of secondary
brainstem injury caused by downward herniation. Although
primary brainstem damage rarely exists in pure form (4), a
primary brainstem lesion may be the principal cause of
coma. In these cases, the EEG can be a helpful diagnostic tool.
Patients in coma from caudally placed brainstem lesions may
have nonreactive alpha activity even in an unresponsive
decerebrate state (73). Sleep spindles, vertex sharp waves, and
K complexes suggest that the lesion is below the still intact
thalamocortical system (74). This suggestion has been con-
firmed by Steudel et al. (75), who described brainstem lesions
with evidence based on cranial CT and autopsy findings in all
their decerebrate patients with spindles and/or alpha activity
in the EEG. During the first week, these activities were
replaced by delta waves, indicating that EEG patterns similar
to the one generally seen in prolonged comatose states may
appear (46,76).
The reaction to external stimuli further characterizes the
depth of coma (10,21). In MBS 1, sensory stimulation may
briefly block the slow activity; sleep, easily recognizable at this
stage, may have a similar effect (48). In further stages of MBS,
reactivity consists of widespread delta activity, alternating pat-
terns, and typical or atypical sleep potentials (see Fig. 22.13A).
In BBS 1 and BBS 2, no reactivity can be observed.
Generalized burst suppression occurs in posttraumatic coma
only when cerebral anoxia was sustained with the trauma
(48,67). Burst suppression activity due to an overdose of sopo-
rific drugs tends to produce uniform intersuppression activity,
in contrast to the intersuppression activity in anoxic coma, in Figure 22.2. A: EEG from a 32-year-old woman in midbrain syndrome
which delta waves and intermittent spikes are observed fre- stage 2 and with clinical signs of a left hemisphere lesion. No clinical
quently (48). This type of activity carries a poor prognosis evidence of epileptic fits. Unilateral predominant slowing in the left
because it indicates a diffuse anoxic encephalopathy that inter- hemisphere. Sharp waves in left upper frontal region. B: EEG from the
feres with the original traumatic brain damage. same patient in midbrain syndrome stage 1, 1 day after first recording.
Metabolic derangements certainly have an important impact Right-sided clinical focal motor attacks. Continuous focal spikes in left
on coma and accompanying EEG abnormalities upper frontal-central region. (From Rumpl E. Elektro-neurologische
(10,13,14,39,77–79). The number of metabolic disturbances is Korrelationen in den frühen Phasen des posttraumatischen Komas. I.
obviously higher in patients in deep stages of posttraumatic Das EEG in den verschiedenen Phasen des akuten traumatischen
coma (13). Most of these patients suffer from multiple trauma sekundären. Mittelhirn und Bulbärhirnsyndroms. Z. EEG-EMG. 1979;
including the liver, kidneys, and lungs. The combination of 10:148–157.)
metabolic and traumatic encephalopathy carries the worst
prognosis. The EEG changes due to metabolic disturbance
interfere with the EEG changes because of the herniation itself, may have clinical seizures shortly before or shortly after the
thus blurring the rules concerning the decrease of different EEG discharges are found in the EEG (Fig. 22.2). Acute epileptic
patterns with advancing deterioration. Further influences on and paroxysmal EEG discharges may also accompany
the EEG are caused by the effects of sedative or anesthetic drugs nonconvulsive (subclinical) seizures or status epilepticus
frequently used in intensive care units. in severe brain injuries. In cases of nonconvulsive seizures
Early electroencephalographic seizure activity is only seen or status, a more severe clinical picture is simulated.
in cases of severe craniocerebral trauma, in which patients Seizures either are generalized from the start or more
416 Part III ■ Clinical EEG: General

often are focal with subsequent generalization (21). Focal

seizure activity consists of runs of spikes and sharp waves
and is more likely to occur in patients with underlying
intracranial hematoma. The EEG is of considerable localizing
value in these cases, and there has been only one reported
case of focal seizure activity occurring contralateral to the
posttraumatic hematoma (21).
Usually, the appearance of the electroencephalographic
seizure activity at these stages of coma carries a worse progno-
sis (21) and is associated with a mortality that is about three
times higher than in the absence of seizure activity (48).
Intracranial pressure may increase even in the presence of
nonconvulsive subclinical status epilepticus or repetitive elec-
troencephalographic seizures (80). Status epilepticus associ-
ated with generalized motor seizures may be the result of
severe cerebral trauma. Death occurs in more than 50% of Figure 22.3. Correlation of lateralizing signs in the EEG with the neuro-
these cases within the first few days (21). In children, post- logic or postmortem examinations. Dark columns demonstrate corre-
traumatic status epilepticus is associated with a far less omi- sponding neurologic and EEG signs of lateralization in midbrain
nous prognosis than in adults (81). Isolated spikes and slow syndrome (MBS) stages 1, 2, and 3. In midbrain syndrome stage 4 and
waves are rare after a recent injury (82), but they may occur in bulbar brain syndrome (BBS) stage 1 (listed in one column) results of
severe cases. Although prognosis in cases with early seizures is the autopsy have to be used in some cases to determine lateralization.
generally bad, acute convulsions are a more serious complica- (From Rumpl E, Lorenzi E, Hackl JM, et al. The EEG at different stages
tion in patients with brain tumor and metabolic of acute secondary traumatic midbrain and bulbar brain syndromes.
encephalopathies (83). Unfortunately, quantitative EEG has Electroencephalogr Clin Neurophysiol. 1979;46:487–497.)
difficulties detecting spikes, brief seizures, burst suppression,
PLEDs, triphasic waves, and intermittent theta and delta
activity. Therefore, the use of serial standard EEG recordings etative) state. Because there has been much doubt concerning
at least three times a week is still required in intensive care the prognostic significance of spindle activity in posttraumatic
patients (84). coma, special attention was paid to spindle activity in a study
PLEDs are rare events in posttraumatic coma; they are usu- based on 70 EEGs taken in the acute stage of coma (within the
ally seen in cases of subdural hematoma (85). More frequent first 2 days after brain injury) and 63 EEGs recorded in a pro-
PLEDs are seen in patients with acute unilateral lesions of vas- longed comatose state (days 3 through 12 after brain injury).
cular origin (86,87). Triphasic waves may also be produced by The incidence of spindles, form of spindles, and asymmetries of
subdural hematoma, but they are more characteristic of meta- spindle activity in the different outcome categories are listed in
bolic disturbances (88,89), especially when no asymmetry can Table 22.2 (91).
be detected.
The lateralizing signs in the EEG (other than epileptic dis-
charges) demonstrate a close correlation to the lateralizing
signs of the neurologic examination (Fig. 22.3). In MBS 4 and
BBS 1, the neurologic examination reveals no lateralizing
signs at all. At these stages, the EEG gives a hint of local cere-
bral lesion. The localized signs in the EEG may accompany all
local cerebral lesions including the three major types of
hematoma—subdural, epidural, and intracerebral (Fig. 22.4).
Asymmetry in reactivity (45) is helpful to ascertain uncertain
lateralizing signs in the acute stages of coma (see Chapter 23;
Fig. 23.9); in prolonged comatose states, asymmetry often
does not manifest itself unless the patient is stimulated (48).

RELATION OF EEG ABNORMALITIES

TO OUTCOME
According to Jennet and Bond (90), patient outcome may be
classified into four categories: death, severe disability, moderate
disability, and good recovery. In the results presented here, the
category of severe disability includes patients in an apallic (veg-
Figure 22.4. EEG from a 42-year-old man in midbrain syndrome stage
3, 1 day after surgical evacuation of a right epidural hematoma.
Unilateral slowing in right hemisphere and asymmetrical (and mildly
atypical) spindles point to the original lesion.
 Chapter 22 ■ EEG and Craniocerebral Trauma 417

Tabl e 2 2 . 2

Incidence of Spindles, Form of Spindles, and Asymmetries of Spindle Activity

in Different Outcome Categories a

Acute Coma (Total %) Prolonged Coma (Total %)

Outcome Asymmetry Asymmetry
Categories Nb Spindles Atypical Spindles N Spindles Atypical Spindles
Good 26 26 100 2 8 3 12 18 8 44 2 11 2 25
recovery
Moderate 10 9 90 3 33 3 33 7 4 57 2 50 1 25
disability
Severe 18 11 61 5 45 7 63 21 5 24 5 100 2 40
disability
Brain 16 9 56 4 44 8 88 17 2 12 1 50 2 100
death
Total 70 64 91 14 22 24 33 63 19 30 10 53 7 37
coma

a
The percentage of atypical spindles and symmetry of spindles remarkably increase with worsening of outcome. Spindles are seen in most cases in acute coma within
the first 2 days after injury but show clear reduction in prolonged coma (days 3 to 12 after injury). (From Rumpl E, Prugger M, Bauer G, et al. Incidence and prog-
nostic value of spindles in post-traumatic coma. Electroencephalogr Clin Neurophysiol. 1983;56:420–429.)
b
N, number of patients and EEGs.

In patients with good recovery, spindles are seen in all
EEGs recorded during the acute stage of coma (100%).
Atypical (8%) and asymmetrical spindles (12%) were rarely
observed in these cases. In prolonged coma, spindles occurred
in only 44% of patients with good outcome; there were atypi-
cal spindles in 11% and asymmetry in 25% of spindle EEGs.
The EEGs of patients with moderate disability outcome
showed spindles in 90% of all records during the acute stage
of coma, but atypical spindles (33%) and asymmetry (33%)
significantly increased. A remarkable increase of atypical spin-
dles (50%) was seen in prolonged coma. The EEGs of patients
who developed severe disability showed spindles in 61% of all
cases. Atypical spindles were found in 45%; asymmetry was
found in 63% of spindle EEGs. Patients with prolonged coma
and subsequent severe disability showed only atypical spin-
dles, and the percentage of spindles decreased to 24%.
Asymmetries were seen in 40%. Patients who suffered brain
death within 2 weeks after trauma showed spindles in 56% in
the acute stage of coma. Asymmetry of spindles was seen in
88% and atypical spindles were seen in 44% of spindle coma.
In prolonged coma, spindles were seen in only two cases
(12%), and both were asymmetrical.
From these findings, one can conclude that the observation
of spindles largely depends on the time of the EEG recording.
Variations of the timing of EEG recordings account for the
incidence of spindles varying between 14% and 67% in differ-
ent reports (75,92) and may also explain why the prognostic
value of spindles was doubted (93–95). Within 2 days after
brain injury, the vast majority of EEGs showed spindles
(12,45,75,93).
A spontaneous decrease of spindles in prolonged coma was
observed independent of the outcome (45,96). This observation
was further corroborated by Bricolo et al. (97), using com-
pressed spectral array in long-term EEG monitoring. These
authors found the disappearance of frequencies of the 12 to
14 c/sec range between days 3 and 5 after trauma to be one of
their most striking results. On the third day after trauma, a
change of sleep-like activity with spindles to a slow monoto-
nous pattern was also observed in children. There was no influ-
ence of this change on the final outcome (98). On the other
hand a sleep organization close to normal based on 24-hour
polysomnographic recordings in prolonged stages of posttrau-
matic coma was a reliable prognostic marker, both for survival
and for functional recovery (42).
GOOD OUTCOME CATEGORY
Typical spindles and reactivity are seen in the majority of
EEGs in acute coma, but spindles are clearly reduced in pro-
longed coma (Fig. 22.5). Reactivity of the alerting type is the
most prominent response, but, in rare cases, reactivity of the
blocking type or disappearance of spindles was the result of
auditory or painful stimulation. The variety of the different
EEG patterns is high and only slightly reduced in prolonged
coma. Hence, these EEG patterns strongly correspond to the
patterns found in MBS 2 (Fig. 22.1). Interestingly, eight
patients showed decerebrate posture, leading to the classifica-
tion of MBS 4. In contrast to these alarming neurologic find-
ings, the CT scan was normal or slightly abnormal in these
418 Part III ■ Clinical EEG: General
Figure 22.5. The EEG patterns in acute (top) and prolonged (bottom)
coma in patients with good recovery. Typical spindles (arrow) are seen
in the majority of EEGs in acute coma but are clearly reduced in pro-
longed coma. REAC, reactivity; other abbreviations as in Figure 21.1.
(From Rumpl E, Prugger M, Bauer G, et al. Incidence and prognostic
value of spindles in post-traumatic coma. Electroencephalogr Clin
Neurophysiol. 1983;56:420–429.)
cases. Therefore, a primary brainstem lesion was thought to be
the principal cause of coma. Typical spindles were noted in
patients with either primary or secondary brainstem lesions
(Figs. 22.6 and 22.7).
MODERATE DISABILITY OUTCOME
CATEGORY
The EEG patterns in acute and prolonged coma show a wide
range of variability, similar to that seen in the good outcome cat-
egory (Fig. 22.8). Typical spindles are reduced in prolonged coma,
but otherwise no significant change in other EEG patterns is seen.
Figure 22.6. A: EEG from a 23-year-old man in acute coma due to sec-
ondary brainstem dysfunction (classical midbrain syndrome stage 2
without lateralization). Easily recognizable typical spindles, symmet-
rical over both hemispheres, accompanied by K complexes and diffuse
slowing. Computed tomography (CT) scan: moderate brain edema
with scattered contusional areas in left frontoparietal and right tempo-
ral regions. No signs of compression at tentorial level. Blood within
the interhemispheric fissure. B: EEG from the same patient in pro-
longed coma. No spindles, superimposed fast activity, and change to
high-voltage short- and long-delta runs (alternating pattern). No later-
alizing signs in the EEG. Outcome: good recovery. (From Rumpl E,
Prugger M, Bauer G, et al. Incidence and prognostic value of spindles
in post-traumatic coma. Electroencephalogr Clin Neurophysiol.
1983;56:420–429.)
SEVERE DISABILITY OUTCOME CATEGORY
There is a decrease of spindles and reactivity and rarefaction of
the number of EEG patterns in acute and prolonged coma
(Fig. 22.9). No typical spindles are found in prolonged coma.
These alterations in EEG patterns are similar not only to those
found in MBS 3 and MBS 4 (Fig. 22.1) but also to those seen in
the transitional stage to the apallic syndrome (see the section
“EEG Changes in Prolonged Comatose States and the Apallic
Syndrome”).

Figure 22.7. A: EEG from a 20-year-old man in acute coma due to pri-
mary brainstem injury. Neurologic signs of atypical midbrain syn-
drome stage 4 (decerebrate posture). Typical spindles, delta bursts,
and superimposed fast activity accompanied by diffuse slowing. CT
scan: normal. B: EEG from the same patient in prolonged coma.
Predominant alpha/ theta activity, several delta waves, reactivity in
form of blocking slow waves, and no lateralizing EEG signs.
Characteristic EEG pattern for pontine lesions. Outcome: good recov-
ery. (From Rumpl E, Prugger M, Bauer G, et al. Incidence and prog-
nostic value of spindles in post-traumatic coma. Electroencephalogr Clin
Neurophysiol. 1983;56:420–429.)
BRAIN DEATH CATEGORY
All EEG patterns already reduced in acute coma showed further
reduction in prolonged coma (Fig. 22.10). Low-voltage output
EEGs were seen in three cases in acute coma. In prolonged coma,
the number of low-voltage output EEGs increased to seven, and
isoelectric EEG was found in four cases. Reactivity was poor in
acute and prolonged coma. Typical spindles were rarely seen;
they were symmetrical only in one case. In rare cases, remnants
of spindles were noted (Fig. 22.11). On the whole, these patterns
correlate strongly with the findings in BBS 1 and BBS 2,
respectively.
There is good reason to assume that spindle activity is of
great prognostic value in acute coma (i.e., within 2 days after
Chapter 22 ■ EEG and Craniocerebral Trauma 419
Figure 22.8. The EEG patterns in acute and prolonged coma in patients
with moderate disability outcome. Typical spindles (arrow) are fre-
quently seen in acute coma and show reduction in prolonged coma. No
dramatic change in other EEG patterns between acute and prolonged
coma. Reactivity is closely related to the capacity for spindling.
Observations hampered by small number of patients. REAC, reactivity;
other abbreviations as in Figure 21.1. (From Rumpl E, Prugger M, Bauer
G, et al. Incidence and prognostic value of spindles in post-traumatic
coma. Electroencephalogr Clin Neurophysiol. 1983;56:420–429.)
trauma). With worsening of outcome, there is a steady decrease
of spindle activity accompanied by an increase of atypical spin-
dles and asymmetry. In prolonged coma, spindles are less fre-
quent; atypical forms and asymmetries correspond with the
outcome, as in cases of acute coma. The overall incidence of
spindles in the first 2 days of coma is high (91%), but it shows
a remarkable decrease of 30% in the following days. The brain’s
capacity for spindling is closely related to its capacity for EEG
arousal; both are partially lost in late stages of MBS and BBS, as
well as in prolonged coma. The disappearance of spindles in
prolonged coma suggests that spindle activity cannot predict
the outcome in prolonged coma. On the other hand, if spindle
activity is present, asymmetry and distortion are of prognostic
value. Atypical spindles and asymmetries significantly increase
in patients with bad outcome. Generally, the decrease of spin-
dles not only indicates the deepening of coma, but more
420 Part III ■ Clinical EEG: General
Figure 22.9. The EEG patterns of patients with severe disability out-
come in acute and prolonged coma. Note the decrease of typical spin-
dles (arrow) in acute coma and the total loss of typical spindles in
prolonged coma. Also reactivity is less frequently seen in these cases.
Note the general decrease in the variety of EEG patterns. REAC, reactiv-
ity. A/ T, predominant alpha and little theta activity; T/ D, predominant
theta and little delta activity; D/ SD, predominant diffuse rhythmic and
arrhythmic delta and subdelta activity; DB, delta bursts; DS, short runs
of delta; DL, long runs of delta; TSP, typical sleep potentials; ASP, atyp-
ical sleep potentials; S, spindles; LS, local slowing; US, unilateral pre-
dominant slowing; SFA, superimposed fast activity; MBS, midbrain
syndrome; BBS, bulbar brain syndrome; N, number of records. (From
Rumpl E, Prugger M, Bauer G, et al. Incidence and prognostic value of
spindles in post-traumatic coma. Electroencephalogr Clin Neurophysiol.
1983;56:420–429.)
frequently heralds the transition to a prolonged comatose state.
In such a case, the loss of spindles is of less prognostic signifi-
cance when other EEG signs of favorable outcome, such as reac-
tivity, alternating patterns, high voltage, and symmetry of
activities, remain unchanged.
There is no significant influence of small contusional areas
in the CT scans on spindle activity (91). Larger hemispheric
lesions are accompanied by increasing asymmetry of spindle
activity. Furthermore, the neurologic findings show close
Figure 22.10. The EEG patterns in patients who suffered brain death
within 14 days after brain injury. In acute coma, typical spindles are
rarely seen (symmetrical only in one case). In prolonged coma, typical
but asymmetrical spindles are seen in one case only. There is similar
decrease in reactivity and limited variety of EEG patterns. Columns of
delta/ subdelta activity include three low-voltage EEGs in acute coma
and seven low-voltage EEGs and four isoelectric EEGs in prolonged
coma. REAC, reactivity; other abbreviations as in Figure 21.1. (From
Rumpl E, Prugger M, Bauer G, et al. Incidence and prognostic value of
spindles in post-traumatic coma. Electroencephalogr Clin Neurophysiol.
1983;56:420–429.)
correlation with symmetry and form of spindles as well as with
the results of CT scan. Exceptions are seen in cases of primary
brainstem injury. Despite decerebrate posturing, many patients
recover well (12,91); others die or survive in a state resembling
a posttraumatic locked-in syndrome (99). In these cases, one
might assume either reversible or irreversible impairment of
the midbrain or brainstem, respectively, but undisturbed thala-
mocortical connections subserve well-formed spindle activity.
On the other hand, there are no significant differences in spin-
dle activity in the different outcome categories when cases of
primary and secondary brainstem lesions are compared. One
must consider, however, that cases of isolated primary brain-
stem injury are rare (4) and that the hemispheres are also fre-
quently involved. This may be especially true in patients with
bad outcome.

Figure 22.11. A: EEG from a 17-year-old girl in acute coma due to sec-
ondary brainstem involvement (classical midbrain syndrome stage 4).
Typical (mildly abnormal) and asymmetrical spindles. Spindles are bet-
ter seen on the left. Slight reduction in amplitudes of slow waves on the
right. CT scan: areas of local brain contusions at basal ganglia level on
both sides. Small hemorrhagic lesion in the posterior part of the left
internal capsule. Signs of tentorial compression. Blood in the third ven-
tricle. B: EEG from the same patient in prolonged coma. Atypical
severely distorted spindles in the left frontal region. Spindles are hardly
recognizable and can only be detected from knowledge of the previous
EEG (A). Outcome: brain death. (From Rumpl E, Prugger M, Bauer G,
et al. Incidence and prognostic value of spindles in post-traumatic
coma. Electroencephalogr Clin Neurophysiol. 1983;56:420–429.)
Although there is increasing doubt about the effect of barbi-
turate anesthesia in the treatment of intracranial hypertension,
barbiturate anesthesia may be advisable in certain cases (100).
This therapy profoundly alters the EEG activity. The EEG by
itself is not sufficient to predict the level of barbiturate anesthe-
sia under clinical conditions in which adjuvant medications are
used (101). However, it is generally accepted that the appear-
ance of burst-suppression activity points to a therapeutic
plasma level of barbiturates (102). There is no doubt that the
EEG patterns mentioned above are widely suppressed by this
therapy; consequently, any prognostic value of the EEG is lost.
Therefore, obtaining early EEG recordings before treating with
barbiturates is advocated. This also pertains to cases treated
with large doses of other sedative or anesthetic drugs. In low
Chapter 22 ■ EEG and Craniocerebral Trauma 421
doses most of these drugs evoke activation of beta activity
(superimposed fast activity), accompanied by a steady increase
of theta and delta waves with increasing anesthesia depth (see
Chapter 56). Since the neurologic examination is also severely
hampered by the influence of high-dose sedative or anesthetic
drugs, many intensive care units prefer to reduce sedation sys-
tematically, to allow useful neurologic examination. This point
of time would be also the best moment for doing an EEG con-
trol examination besides continuous EEG monitoring.
EEG CHANGES IN PROLONGED COMATOSE
STATES AND THE APALLIC SYNDROME
In prolonged coma, neurologic findings are similar to those
seen at the acute stage; only small differences are observed
(103). The EEG changes in this period have been described in
detail see EEG CHANGES IN THE ACUTE STAGES OF POST-
TRAUMATIC COMA. There are, however, more characteristic
EEG patterns during the transitional stage from MBS to the
traumatic apallic syndrome (104). The clinical symptomatology
of the transitional stage is not limited to patients with evolution
into the full stage of an apallic (vegetative) syndrome but also
occurs in patients who recover with a marked organic brain
syndrome characterized by various severe neurologic signs of
multilocal brain damage.
CLASSIFICATION OF PROLONGED COMA
AND THE APALLIC SYNDROME
Transitional Stage to the Apallic Syndrome
During the prolonged evolution of a comatose state, the
patients are still in deep coma with closed eyes. At the end of
this stage, within 2 or 3 weeks, the survivors begin to open their
eyes, initially for very short periods and after arousing stimuli.
A steady increase of chewing, a decrease of decorticate or decer-
ebrate rigidity, an increase of extrapyramidal symptoms, and
normalization of the reflex eye movements can be observed.
Patients react to painful or acoustic stimuli with varying
degrees of extensor and flexor response in the limbs accompa-
nied by an altered respiratory pattern. The most important and
characteristic feature of this stage is the onset of overactivity of
the sympathetic nervous system, leading to extensive tachycar-
dia. This overactivity has been confirmed by the demonstration
of high catecholamine plasma levels (mainly of norepineph-
rine) (105). This sequence of events may begin in stages 2 to 4
of MBS but is most frequently seen in MBS 3 and 4 (106).
Full Stage of the Apallic Syndrome
The term apallic is used for a clinical syndrome (107) without
any implication of a particular cerebral lesion. The patients lie
with open eyes for longer periods, paying no attention to events
around them. They seem to be awake, but not aware. They do
not notice events around them. They may show no blink
response to menace and react with abnormal movements only
to painful stimulation. There are no emotional reactions. Chewing
or teeth grinding occurs spontaneously or may be evoked by
422 Part III ■ Clinical EEG: General
tactile perioral stimuli. The muscle tone is increased, and only
fragments of coordinated movements can be observed. The
sleep-waking rhythm seems to be controlled by exhaustion. In
the waking state, the overactivity of the sympathetic nervous
system continues; during sleep, a shift to a parasympathetic set
of responses may be observed.
State of Stupor and Confusion
In less severe cases, the patients show prolonged symptoms of
early MBS (usually MBS 1) and pass into a state of stupor and
confusion. The patients are usually drowsy, lying in a flexed atti-
tude. During the waking state, patients may be resistant to nurs-
ing care and examinations, disoriented, and sometimes noisy
and violent. Neurologic focal symptoms become more evident.
Most of these patients need sedative drugs. This condition may
last for days or even weeks and gradually passes. Symptoms of
local brain damage may persist and may cause neurologic and
mental deficits of varying severity.
RELATION OF EEG ABNORMALITIES TO THE
STAGES OF PROLONGED COMA AND THE
APALLIC SYNDROME
The EEG patterns of patients in the transitional stage of the
apallic syndrome are compared with an EEG recorded during
the acute MBS 3 in Figure 22.12. There is a clear decrease in
the variety of EEG patterns, especially of sleep or sleep-like
potentials and alternating patterns (76,108). A complete loss
of spindles is noted in these patients (Fig. 22.13). This evolu-
tion of the EEG patterns from waveforms of sleep into rather
Figure 22.12. The EEG patterns of 13 patients in the transitional stage
to the traumatic apallic syndrome evolving from midbrain syndrome
stage 3. Dark columns demonstrate the decrease of EEG patterns at the
transition stage compared with the EEG patterns at the acute stage of
midbrain syndrome (white columns). Note the complete loss of spindles
in these cases. Abbreviations as in Figure 22.1. (From Rumpl E. Elektro-
neurologische Korrelationen in den frühen Phasen des posttraumatis-
chen Komas. II. Das EEG im Übergang zum, und im Vollbild des
traumatischen apallischen Syndroms. Z. EEG-EMG. 1980;11:43–50.)
uniform, frequently high-voltage activity creates the impres-
sion of increasing rostrocaudal deterioration. This impression
may be supported by the decrease of reactivity; there may be
no reactivity to any mode of sensory stimulation (48,76). A
similar evolution of the EEG is seen in patients developing the
transition stage from MBS 2. In contrast to the unfavorable
outcome of patients with increasing intracranial pressure, the
loss of the activities mentioned above does not carry a poorer
prognosis when seen in patients at the transition stage
(46,76). Patients in MBS 4 show an already-simplified EEG
pattern with only slight further changes at the transition stage.
Loss of atypical sleep potentials may be observed, or reactivity
may be abolished (76).
The variety of EEG patterns is also reduced in the full stage of
the apallic syndrome, but this reduction is less marked than in
the transitional stage. The variety of EEG patterns increases as
soon as cyclic sleep and waking activity can be identified. This is
associated with the spindles and long runs of delta (Fig. 22.14).
On the other hand, the EEG of patients developing the full stage
of the apallic syndrome from MBS 4 remains rather uniform
and shows only a little difference between sleep and waking
activities. Periods of fast activity of low voltage, along with eye
closure, usually characterize sleep, whereas high-voltage slow
waves indicate the waking state (76,109). Similar alternating
periods of faster low-voltage activity and high-voltage slow
waves during were found in sleep in long-lasting nocturnal
records (110–112). The appearance of EEG sleep patterns and
circadian EEG variations is a favorable prognostic sign (112).
Completely organized sleep patterns or increasing organiza-
tion of nocturnal sleep in successive recordings indicates a more
favorable outcome than the absence of such patterns. The
occurrence of REM sleep is especially significant in this regard
(108). This finding may be supported by reports of the reap-
pearance of slow-wave sleep with continued absence of the
REM phase in patients who never regained consciousness after
brainstem vascular accidents (113,114). In rare cases, positive
sharp waves in the occipital regions are seen during sleep (21).
In contrast to reports that suggest that sleep or sleep-like
activities point to a favorable outcome, there is no evidence
that sleep or sleep-like activity found at the full stage of the
apallic syndrome months after the brain injury gives any
information about further outcome (115). Patients with rela-
tively well-developed sleep stages persisted clinically at the
stage of an apallic syndrome or remained severely disabled
(Fig. 22.15), whereas patients with severely altered sleep
showed further recovery. However, the value of long-term
EEG recordings in these patients was hampered by the fact
that given stages of sleep or wakefulness lasted no longer than
a few seconds or minutes. This phenomenon appears to be
rather characteristic for apallic patients (109,115). REM sleep
was hardly separated from sleep stage 1; rapid eye movements
appeared at any stage, even during synchronized sleep.
Atypical and asymmetric spindle activity continued to be of
prognostic significance by indicating a persistent structural
hemispheric lesion. Generally, the prognostic value of long-
lasting nocturnal EEG recordings is very limited in apallic or
severely disabled patients if done months after the craniocere-
bral trauma.

Figure 22.14. The EEG patterns of the same 13 patients as in Figure 22.12
at the full stage of the apallic syndrome (dark columns) compared with the
EEG patterns at the acute stage of midbrain syndrome (white columns).
Note the general acceleration and the reappearance of spindles and
altered sleep activities. Abbreviations as in Figure 22.1. (From Rumpl E.
Elektro-neurologische Korrelationen in den frühen Phasen des posttrau-
matischen Komas. II. Das EEG im Übergang zum, und im Vollbild des
traumatischen apallischen Syndroms. Z. EEG-EMG. 1980;11:43–50.)
Chapter 22 ■ EEG and Craniocerebral Trauma 423
Figure 22.13. A: EEG from a 24-year-old woman in acute
midbrain syndrome stage 3 without clinical signs of lateral-
ization. After acoustic stimulus (arrow), appearance of spin-
dles followed by delta and subdelta activity of higher
amplitude. Note the symmetry of reaction. B: EEG from the
same patient in the transition stage to the apallic syndrome
4 days after first recording. Uniform delta and subdelta
activity; loss of spindles. After acoustic stimulus (arrow),
long run of delta. Note the symmetry of reaction.
Figure 22.15. Nocturnal EEG from a 29-year-old woman at the full stage
of an apallic (vegetative) syndrome. Later she has learned to move the
left lower extremity on command but remains severely disabled. Typical
spindles, K complexes during sleep, stage 2. Continuous myogenic spik-
ing in left temporal region due to palatal myoclonus. (From Rumpl E,
Prugger M, Bauer G. Zum prognostischen Wert elektroenzephalo-
graphischer Schlafbeobachtungen im traumatisch bedingten apallischen
Syndrom. Neuropsychiatr Clin. 1984;3:219–232.)
424 Part III ■ Clinical EEG: General
The EEG patterns of patients in the state of stupor and
confusion show a decrease in the variety of EEG patterns sim-
ilar to the one seen in prolonged coma. However, sleep or
sleep-like potentials are more frequently seen in these
patients in diurnal examinations. Generalized theta with less
delta activity is the predominating pattern at this stage; it
is usually seen in waking periods and without further change
after stimulation. The value of most records at this stage
is hampered by the use of the sedative drugs needed to obtain
technically satisfactory recordings. Drug-induced sleep
activities might also indicate a favorable outcome in most of
these patients, and asymmetries in sleep potentials might be
of diagnostic significance in cases of chronic subdural
hematoma.
The reaction to external stimulation is remarkably reduced
in the transitional stage of the apallic syndrome. In most cases,
no response can be observed (48,76). Less frequently, a delayed
reactivity characterized by long runs of delta is noted. No reac-
tivity is seen in patients in the full stage of the apallic syndrome
during the waking state, whereas stimulation during sleep fre-
quently induces long-lasting waking activity (36,109).
Metabolic disturbances decrease in patients in prolonged
comatose states (10). Therefore, the EEG abnormalities are less
influenced by metabolic factors than in the stages of MBS.
Nevertheless, the influence of metabolic disturbances and drugs
must be taken into consideration at these stages.
Epileptic discharges are not seen in patients in the transi-
tional stage or full stage of the apallic syndrome (18,76) but
may occur in the state of stupor and confusion and during the
remission stage from the apallic syndrome (18).
Lateralized signs in the EEG are slightly reduced in pro-
longed comatose states, mainly due to the loss of asymmetrical
typical or atypical sleep potentials. Asymmetry of reactivity is
diagnostically helpful (48).
Prognostic criteria can be derived from EEG findings only
with great care and caution. A gradual increase in frequency
from delta to faster activity points to a favorable index only
for survival, not for recovery (25). Improvement of EEG must
be accompanied by clinical improvement to have any prog-
nostic significance (18,25). An isoelectric EEG indicates loss
of bioelectrical activity, to be discussed in Chapter 23, on
coma and brain death. Patients who survive with an isoelec-
tric EEG are classified as with complete apallic syndromes by
some authors (116,117). In cases of very low-voltage EEG in
repeated recordings, one can expect survival without hope of
recovery as a social human being (117). In patients who
develop the symptomatology of the apallic syndrome from
MBS 4 accompanied by rather rarified EEG patterns, the
prognosis for social recovery is also very poor (76,106). Most
of these patients remain at the full stage of the apallic syn-
drome or at a minimally conscious state (118). The favorable
clinical outcome of patients with organized sleep has been
mentioned above. The persistence of sleep patterns or sleep-
like activities at the transitional stage of the apallic syndrome
might also indicate a favorable outcome although the value of
this observation is hampered by the very small number of
cases studied (76,92).
EEG CHANGES AT EARLY STAGES AFTER
CRANIOCEREBRAL TRAUMA WITHOUT
EVIDENCE OF BRAINSTEM DYSFUNCTION
In mild or moderate head or brain injuries, there may be no or
only mild disturbances of consciousness with no clinical signs
of primary or secondary brainstem dysfunction. In general, the
EEG recovery is fast in these patients, who fortunately form the
vast majority of cases after craniocerebral trauma.
RELATION OF EEG ABNORMALITIES TO
MILD OR MODERATE STAGES OF CEREBRAL
CONCUSSION
In cases of commotio cerebri (cerebral concussion with loss of
consciousness) but with no impairment of consciousness at the
time the records were taken, normal or borderline EEGs were
found in the vast majority of cases within the first 24 hours
(27,28,32). No slowing of the alpha rhythm could be detected in
these recordings (19). The slowing of alpha rhythm is usually
considered to be the slightest degree of generalized disturbance
after contusio cerebri and may only be found with repeated
EEG, because the significant slowing occurs in the alpha fre-
quency range (19,23).
The mildest disturbance of consciousness seen after cranio-
cerebral trauma is drowsiness and hypersomnia, which are
accompanied by the EEG findings of normal sleep (48). There
are usually generalized slowing of all frequencies and altered
sleep patterns in patients with barely existent periods of wake-
fulness. Sensory stimulation will block the slowing and sleep.
There may be gliding clinical and electroencephalographic
transitions to the first stage of MBS. However, the change from
abnormal EEG activity to a more normal pattern is fast in these
cases (Fig. 22.16).
In contrast to the aforementioned fairly good correlation of
EEG and clinical findings, there are many patients with head
injury who show marked discrepancies between EEG patterns
and clinical findings. A normal neurologic status in a patient
with no or little complaint may be accompanied by impressively
abnormal EEGs (Figs. 22.17 and 22.18). Similar EEG findings
may also be seen in patients with cerebral concussion without
loss of consciousness but with a short or longer lasting alter-
ation in mental status (119). The EEG would most likely detect
cortical disease in acute cases with mild injury, even if the CT
scan or MRI is normal. Very little is said in the literature about
these cases, and only a few authors regard them as mild forms
of concussion (120–122). In contrast, a normal EEG can be seen
in patients with significant impairment in neuropsychological
performance after mild traumatic brain injury within 24 hours
and up to 6 weeks after injury (123).
Particularly pronounced discrepancies between electrical
cortical activity and the clinical state may be found in children
(127–129). Children may show high-amplitude generalized delta
activity for weeks or months after a mild craniocerebral trauma
and in the presence of normal consciousness (48). Immaturity
might enhance functional traumatic (postconcussion)
 Chapter 22 ■ EEG and Craniocerebral Trauma 425

Figure 22.16. A: EEG from a 20-year-old woman with cerebral contu-

sion, drowsiness, and clinical signs of left frontoparietal lesion. Delta
and subdelta activity, unilateral slowing, and suppression of altered
spindles in the left hemisphere. B: EEG from the same patient, 1 day
after first recording. Reappearance of reactive alpha activity, local slow-
ing in left fronto-temporo-parietal region. C: EEG from the same patient,
3 days after first recording. The EEG is nearly normal; intermittent slow
waves in left temporal region. Full clinical recovery with the exception
of headache, lack of concentration, and dizziness.

disturbances and produce focal or generalized EEG abnormali-
ties (21). On the other hand, focal changes have been thought
to be a significant sign of contusional brain damage (130).
Elderly persons may have pretraumatic EEG abnormalities,
especially in the temporal regions, due to cerebrovascular insuf-
ficiency. The combination of traumatic and vascular factors
makes it difficult to distinguish the EEG abnormalities caused
by the cerebral contusion. There is particularly high mortality
in elderly patients with early epileptic seizures (131). The loss of
reactivity may appear in light stages of coma, indicating a bad
prognosis (21).
Epileptic Discharges
The term early epilepsy should be confined to fits in the first
week after injury (132). Localized focal motor epilepsy is most
common in the first week, whereas acute temporal lobe
426 Part III ■ Clinical EEG: General

Figure 22.17. EEG from a 35-year-old man complaining of

vertigo after mild cerebral contusion. At the time of EEG
recording, normal neurologic status and CT scan. Intermittent
slow activity (theta and delta) in left frontotemporal region.

Figure 22.18. EEG from a 45-year-old woman after mild cerebral

contusion without signs of brainstem impairment. Some com-
plaints about vertigo. At the time of EEG recording, normal neu-
rologic and CT scan findings. Intermittent left temporal theta
focus. The recovery of EEG abnormalities follows the same pat-
tern as in adults (23). In children with transient cortical blind-
ness following mild head trauma, posterior slowing with
subsequent fast normalization can be observed (124).
Lateralized signs consist mainly of focal slow waves in the theta
and delta range (17,19,23,32), whereas the other cerebral
regions show only minor changes (125). A close correlation
with neurologic deficits is seen within the first months
(17,19,20,126).

epilepsy with psychomotor or complex partial seizures is
never found during this period. Generalized seizures were
found in 50% of the cases, while the other 50% had focal
attacks (21). A higher percentage of focal seizures is assumed,
because of secondary generalization and doubtful clinical
observation (21). After 2 weeks, adults with posttraumatic
epilepsy may have any type of epileptic manifestation, with
the exception of petit mal attacks with generalized rhythmic
synchronous 3/sec spike-and-wave activity in the EEG (133).
Although epilepsy during the first week recurred or persisted
in less than one third of patients (the recurrence rate in the
second week is over 70%), the risk for late epilepsy is signifi-
cantly greater for patients with early fits than for those with-
out (132). A threefold increase in the incidence of both early
and late seizures is seen in patients with depressed skull frac-
tures when compared with injuries accompanied by nonde-
pressed fractures (48). Occipital status epilepticus is a rare
event in posttraumatic blindness (134). While cortical blind-
ness in traumatic brain-injured patients with no or minimal
occipital pathology on CT or MRI scanning is generally tran-
sient (Fig. 22.19), occipital status epilepticus accompanied by
occipital edema may lead to persistent visual deficit if not
detected for early treatment.
Long runs of continuous activity of focal spikes or unilateral
spike-wave discharges are the most common EEG pattern in
early epilepsy (21). In general, focal epileptiform patterns asso-
ciated with diffuse EEG abnormalities suggest a greater proba-
bility of subsequent chronic epilepsy than does focal
epileptiform activity in an otherwise normal EEG (48). Using
standard 24-hour recordings, seizure activity can be seen after
trauma in up to 35% of patients with mild or severe head-brain
injury in the first weeks after trauma, but accompanying epilep-
tic fits are rare (135). In cases of severe brain injury early elec-
troencephalographic seizure activity might be seen in patients
who will have clinical seizures shortly before or after the dis-
charges are found in the EEG (Fig. 22.2).
Epileptiform activities are more frequently seen in children
than in adults. These are isolated spikes or spike waves, sharp
 Chapter 22 ■ EEG and Craniocerebral Trauma 427

Figure 22.19 A: EEG from a 6-year-old girl who had a fall on the back of her head after an unexpected push while playing.
She had a period of blurred vision, delayed answers to questions, disorientation, drowsiness, and stereotyped finger move-
ments, which lasted about 1 hour. The CT and magnetic resonance imaging scans were normal. At the time of the EEG
recording 2 days after the accident, the patient was alert with no neurologic or mental deficit. The EEG showed high-voltage
occipital delta activity intermingled with sharp transients. B: EEG from the same patient 3 days later within the range of
normal.
428 Part III ■ Clinical EEG: General
waves, and spike-wave complexes, localized or generalized. A
single epileptic focus in the EEG after severe head trauma has
the same favorable outcome as in benign focal epilepsy (136).
Generalized irregular spike-wave activity is occasionally seen
(48). The 14 and 6/sec positive spikes can be an abnormal find-
ing after mild or severe head injury in children. This pattern
may appear as a delayed reaction to mild injury, but may also be
seen during recovery from severe injury (137). However, the
significance of 14 and 6/sec positive spikes is of less value,
because it may also be seen in healthy children (138). All of
these discharges can be observed within hours of the trauma
(more commonly within days or weeks), and all are associated
with a good prognosis.
EEG CHANGES AT LATE STAGES AFTER
CRANIOCEREBRAL TRAUMA
In late stages after craniocerebral trauma, clinical and EEG
examination demonstrate diminished correlation, which fur-
ther increases after a 3-month interval (17,19,21,130). This lack
of correlation may also occur in patients during recovery from
an apallic syndrome (18).
CLINICAL CLASSIFICATION
Recovery from the Apallic Syndrome
Although the symptomatology of the apallic syndrome is
remarkably uniform, there are different degrees of telencephalic
lesions, so that recovery is possible (107). During the remission
stage (139), there is evidence of increasing reintegration of higher
brain function. The onset is indicated by primitive emotional
reactions in response to painful stimuli followed by optical fixa-
tion and displeasure reactions. The first voluntary movements
are usually seen in the fingers. The patient starts to respond to
simple orders. In patients who make a reasonable recovery, fur-
ther stages of remission appear, marked by the symptomatology
of the Klüver–Bucy syndrome, the Korsakoff syndrome, and,
finally, the organic brain syndrome of varying severity.
Symptoms of local brain lesions can now be clearly determined.
These lesions may be permanent, but in some cases the apallic
syndrome is transient with complete recovery of the patient.
In mild and moderate brain injuries, the onset of the late
stage can be clinically defined by the disappearance of distur-
bances of consciousness and confusion and by the stabilization
of neurologic deficits and psychic syndromes (21). Usually, a
gradual recovery is seen, but in some cases neurologic and
mental deficits may persist.
Posttraumatic Syndrome
This syndrome comprises headache, vertigo, dizziness, nervous-
ness, irritability, impaired memory, disturbances of concentra-
tion, excessive fatigue, insomnia, and intolerance to alcohol
(140). Practically, in 50% of the patients, the symptoms disap-
pear within 2 weeks after mild injury, while in 25% the symp-
toms continue for months and sometimes years (141).
However, it should be noted that most of the posttraumatic
complaints do not appreciably differ from nontraumatic neu-
rotic syndromes.
RELATION OF EEG ABNORMALITIES
TO THE LATE STAGES OF
CRANIOCEREBRAL TRAUMA
EEG recovery from the full stage of the apallic syndrome is gen-
erally associated with an acceleration of EEG activity. Theta and
(fewer) delta waves characterize the apallic syndrome months
after the brain injury (18,142). Recovery from this stage is char-
acterized by the decline of theta and delta waves and reappear-
ance of reactive alpha activity. Favorable outcome is associated
with faster regression of theta waves and more progressive rein-
stitution of alpha activity (142,143). In a summary of their
findings, Lücking et al. (18) have been able to draw the follow-
ing conclusions on the prognostic value of the EEG in the apal-
lic syndrome: (i) Predominant theta and delta activity
unchanged for several months points to a significant lesion of
the hemispheres. (ii) An early and progressive appearance of
alpha activity in the absence of clinical improvement suggests
slight cortical, but extensive subcortical, lesions. (iii) Alpha fre-
quency may show a gradual acceleration within years without
further clinical improvement. (iv) Focal symptoms become
more marked after the decrease of general changes and may
shift from side to side in cases of temporal foci. (v) Epileptic
discharges may occur in patients with no evidence of epileptic
fits, although patients with focal epileptic attacks may have no
epileptic discharges in repeated EEGs.
EEG recovery of generalized abnormalities is characterized
by a steady increase in frequency. After comatose states, gener-
alized theta activity is one of the most constant features of the
EEG recovery (48) and may show further improvement with
gradually evolving reactive alpha activity. After mild head
injuries with cerebral contusion and slowing of the basic
rhythm, follow-up EEGs are often characterized by a gradual
increase of alpha frequency returning to normal within several
weeks or months (17,19,23,32). Six months after trauma, alpha
slowing persists in only 4% of the cases (23).
EEG recovery of lateralizing signs shows a characteristic
evolution in uncomplicated cases. Unilateral slowing or a delta
focus will gradually change into a theta focus or a focus of irreg-
ular waves of changing frequency. This focus may turn into the
focal slowing of alpha rhythm in the parietal, posterior, and
occipital regions with the slightest degree of local disturbance
(17,19). This focal slowing usually disappears quickly and may
sometimes be followed by focal reduction in amplitude (21).
The diagnostic value of such alpha asymmetries is doubtful,
however, when one considers the frequent occurrence of such
asymmetries in healthy persons (20). Another rare and tran-
sient local sign is focal activation of alpha rhythm (31) charac-
terized by a focal slowing and increase of alpha amplitude by
eye closure; eye opening fails to block the activity.
In cases with traumatic psychosis (delirium, Korsakoff
syndrome, and changes of affect), EEG foci are demonstrated in

95% of cases, bilaterally in 70%. Normalization occurs within
3 months in 48%, with foci persisting more than 2 years in
22% mostly in patients with traumatic epilepsy. Focal signs
initially consist of delta waves (85%) and finally of focal
dysrhythmia (73%) with temporal localization increasing from
58% to 82% (24).
Even in the late posttraumatic stage, focal EEG abnormali-
ties may show a close correlation with the clinical focal deficit
(23,30). After closed head injuries, focal EEG abnormalities
tend to disappear within 6 months (23) to 2 years (19) in the
vast majority of cases. In contrast, focal changes could be
found in 50% of cases 20 years after an open craniocerebral
trauma (144), and focal changes in 40% of patients with pro-
longed confusional states were seen (30). Focal abnormalities
decrease more slowly than generalized ones and may even be
found some years after injury (21). The area of the focal
abnormality usually shrinks in size and may narrow down to
a temporal or parieto-occipital focus (21). In penetrating head
injury, a close correlation between focal slowing and focal
neurologic deficit may be found even 12 to 16 years after
injury (22). In open craniocerebral trauma, foci often persist
over the area of penetration, and migration to posterior
regions is unusual (127). In rare cases, the location of the focal
abnormality may shift from one area to another (23,29). More
frequently, the foci shift to the frontal rather than to the occip-
ital region, but they may return to the original locus in both
cases (29).
Epileptic Discharges
Apart from early attacks, late epilepsy may develop within a
month or two of the injury or may be delayed for many years.
Early attacks will predispose to late ones (132). A detailed dis-
cussion of posttraumatic epilepsy is presented in Chapter 27.
The EEG in the posttraumatic syndrome can neither prove
nor rule out the existence of traumatic symptoms (145,146).
Even pronounced EEG abnormalities are not necessarily associ-
ated with posttraumatic syndromes (21). In contrast, a normal
EEG cannot exclude severe cerebral damage.
EEG Changes in Whiplash Injuries, Cervical
Fractures, and Dislocations
Whiplash injuries may be accomplished by EEG abnormali-
ties; similarly, patients with cervical fractures and dislocations
may demonstrate EEG changes even in the absence of loss of
consciousness at the moment of the accident (21). So the
incidence of generalized and focal EEG changes (focal spikes
and focal slow waves) found in a selected group of whiplash
injury cases was the same as in patients with closed head injury
(147). An additional injury to the head was carefully excluded
in these patients. On the average, the EEGs were done 6 months
after the accident. Temporal lobe foci were more frequently seen
in whiplash injury than in closed head trauma, but this did not
prove to be statistically significant. Slow waves appeared to be
accentuated over the posterior regions of the brain. In contrast
to the EEG changes after closed head injuries, the EEG became
increasingly abnormal in many of the whiplash injury cases
when follow-up records were studied. Moreover, many cases
Chapter 22 ■ EEG and Craniocerebral Trauma 429
showed marked fluctuations in the degree of abnormality sug-
gesting a vascular mechanism for these changes (147).
The EEG changes might indeed be due to fluctuations in the
blood supply from the vertebral arteries, which can be impaired
by hyperextension of the neck (148,149). Vertebral fractures with
displacement of fragments may directly compress the vertebral
arteries. Similar patterns in the EEG may appear in this case.
There are also alternative explanations for the genesis of EEG
abnormalities in the wake of neck injuries. The acceleration–
deceleration as such may directly affect the cerebral parenchyma
(50), or the rapid sequence of backward–forward movements of
head and neck may cause traumatic lesions as the cerebrum
strikes the inner table of the skull (147). Mostly normal EEG (and
also 24-hour cassette recording) were found in 68 cases with
whiplash injuries without accompanying head injury (150).
COMPRESSED SPECTRAL ARRAY
AND BRAIN MAPPING
Continuous EEG monitoring of patients in the first days fol-
lowing injury requires the use of computer-assisted (Fourier
transform) analysis using compressed spectral analysis (CSA)
or similar techniques (see Chapters 54 and 56). With this
method the power spectra or specific EEG frequency bands may
be monitored for the long term. The dominant frequencies,
their distribution, and their amplitude may be better assessed
with CSA than with conventional EEG (15). Two or four chan-
nels appear to be adequate to regain significant data regarding
the outcome of patients (75,97,151,152). Unfavorable outcome
was seen in patients with slow and monotonous CSA and in
patients with significant interhemispheric amplitude asymme-
try. Favorable outcome was frequently found in patients with
changeable CSA, especially with sleep-like CSA (97,152).
Persistence or return of activity within the alpha or theta range
pointed to good outcome in most cases (151). Repeated studies
of CSA showed an increase in the absolute and relative ampli-
tudes of the alpha and theta ranges in patients who survived
(153). Difficulties in interpretation may appear in patients in
alpha or theta pattern coma, as well as in patients undergoing
barbiturate therapy. Further difficulties are that CSA may fail to
show paroxysmal patterns, such as spikes, brief seizures, burst
suppression, PLEDs, triphasic waves, and intermittent theta and
delta activity (84).
Calculation of the theta/beta power ratios in patients with
deep coma revealed long-lasting EEG reactions after standard-
ized repetitive stimulation (154). With this method EEG reac-
tion may be elicited even in cases where other methods fail to
evoke an EEG response—acute or delayed—at all. The
theta/beta ratio decreased with recovery of cerebral functions. A
bolus injection of 100 to 500 mg of thiopental evokes short-
lasting beta stimulation in CSA. This response was of high
prognostic significance since 84% of patients with this reaction
survived (155).
In mild head trauma discriminating EEG power spectral
analysis indicated three classes of variables that are attributable
to mechanical head injury: (i) increased coherence and
430 Part III ■ Clinical EEG: General
decreased phase in frontal and frontotemporal regions; (ii)
decreased power differences between anterior and posterior cor-
tical regions; (iii) reduced alpha power in posterior cortical
regions (156). These authors speculate that these electrophysio-
logic features are consistent with the mechanics of cerebral
trauma and may explain persistent symptoms after mild trauma
such as headache, dizziness, memory loss, short attention span,
and reduced ability to process complex information.
Subsequently, EEG measures of cerebral asymmetry such as EEG
coherence and phase and amplitude symmetry were found to be
best predictors of outcome also in patients with severe brain
injury. EEG coherence and phase have been shown to reflect to
some extent the magnitude of diffuse axonal injury (157).
Topographic brain mapping (TBM) can demonstrate sub-
tle asymmetries and lateralization and localization effects
more efficiently than standard EEG, when the recording is
done properly from a technical standpoint (158,159). In 135
patients after mild or moderate head injury, abnormal TBM
was found in 56% of cases years after injury. The tempo-
rofrontal regions were involved in 65%, the temporo-occipital
regions in 25%, and the parieto-occipital regions in 9% of
patients. The most common type of abnormality was absolute
voltage asymmetry. In contrast the standard EEG showed
abnormalities in 30% of cases consisting mainly of mild, non-
specific generalized slowing (158). TBM appears to provide
better detection of low- amplitude slow activity not easily
discernible by routine EEG (159).
HEAD INJURY IN SPORTS
A generalized reduction in amplitude and slow irregular theta
activity has been observed in boxers within 15 to 30 minutes
after fighting, with a significant increase of these findings in
boxers who had sustained a knockout blow (160–162).
Although the EEG might detect even slight degrees of cere-
brovascular autoregulation disturbance, the method is not rou-
tinely used to avoid second impact syndrome in contact-sports
head trauma (119,163). However, the method should be more
sensitive than CT or MRI scanning.
In long-term observations a correlation between the fre-
quency of abnormal EEGs and the number of fights and, fur-
thermore, a correlation between EEG abnormalities and short
time intervals between fights was found (162). Repeated fights
within a short period have particularly ill effects on the brain,
and the extent and degree of EEG abnormalities increased after
each fight (21,162). Usually the EEG of young boxers shows
more serious changes than that of older ones. No severe abnor-
malities were found in amateur boxers either with many or few
matches. There was a somewhat higher incidence of slight or
moderate EEG deviations among boxers than among soccer
players and track and field athletes. No neurophysiologic vari-
able was correlated with the number of bouts, number of lost
fights, or length of boxing career. Also an abnormal EEG did
not predict the degree of clinical impairment. EEG abnormali-
ties showed no correlation with neurologic findings or the
number of bouts the fighter has waged (164). In general, no
signs of serious brain damage were found among amateur ath-
letes (165). On the other hand, a significantly increased
incidence of focal theta activity was found in soccer players at
the end of their careers, probably as the result of a cumulative
effect due to repeated heading (166). Slightly abnormal to
abnormal EEGs were found in 35% of active and in 32% of for-
mer soccer players. Interestingly enough, there were fewer
abnormal EEG changes among the typical “headers” than
among the “nonheaders” (167).
There is a chronic progressive traumatic encephalopathy in
boxers, especially in those with a history of repeated knockout
defeats. A significant increase of EEG abnormalities in boxers
who suffered from this encephalopathy, which is known as the
punch drunk state (168) or dementia pugilistica (169), in com-
parison to the EEG of more successful boxers was found (170).
However, no correlation between the degree of boxer’s
encephalopathy and the EEG was eventually seen (170,171).
Therefore, the EEG may be useful in determining the severity of
trauma after a fight, but it is a less sensitive indicator of boxer’s
encephalopathy during and beyond the active career. No corre-
lation with clinical findings was established even with the use of
psychometric tests (172). As in EEG recovery after craniocere-
bral trauma of other origins, a normal EEG cannot exclude the
existence or the development of encephalopathy.
Other sports, such as horseback riding, American football,
and rugby, may carry a similar risk of repeated concussion. The
number of injuries might be quite different from that of boxers,
but some of them may be more severe than those usually occur-
ring in the boxing ring. Temporal lobe epilepsy and persisting
mental impairment was found in professional jockeys after
repeated riding accidents (173).
CONCLUSION
There is a close correlation between clinical and electroen-
cephalographic findings in acute posttraumatic comatose
states. A declining number of observable EEG patterns may
suggest increasing rostrocaudal deterioration and poor progno-
sis, as well as the development of a prolonged comatose state.
An assessment of the EEG abnormalities without knowledge of
the clinical symptoms is difficult in these stages. Metabolic dis-
turbances and drugs may also influence the EEG patterns. Early
epileptic discharges in the EEG are of prognostic significance if
they appear prior to clinical seizures. In cases of primary brain-
stem damage, the EEG can help to differentiate these lesions
from secondary brainstem involvement. In comatose states, fol-
low-up EEG is necessary to determine the state of cortical activ-
ity and its evolution. Early EEG recordings are stressed in all
cases of craniocerebral trauma because of increasing lack of
correlation between clinical and EEG data in later stages. No
conclusion can be drawn from a single EEG at a late stage as to
the severity of the original trauma. A single record may only be
useful in the case of epileptic manifestations. Serial tracings
during acute and late stages accompanied by close clinical
examinations will increase the value of the EEG in any case.
Secondary deterioration of the EEG after a period of recovery

suggests late intracranial complications, such as epilepsy,
hematoma, hydrocephalus, abscess, and meningitis, or the
development of a concomitant cerebral disease.
REFERENCES
1. Jellinger K. Häufigkeit und Pathogenese zentraler Hirnläsionen
nach stumpfer Gewalteinwirkung auf den Schädel. Wien Z
Nervenheilkd. 1967;25:223–249.
2. Maciver JN, Lassman LP, Thomas CW, et al. Treatment of severe
head injuries. Lancet. 1958;2:544–550.
3. Jefferson M. Altered consciousness associated with brain stem
lesions. Brain. 1952;75:55–67.
4. Mitchell DE, Adams JH. Primary focal impact damage to the brain
stem in blunt injuries. Does it exist? Lancet. 1973;2:215–218.
5. McNealy DE, Plum F. Brain stem dysfunction with supratentorial
mass lesions. Arch Neurol. 1962;7:26–48.
6. Plum F, Posner JB. Diagnosis of Stupor and Coma. Philadelphia:
F.A. Davis; 1966.
7. Gerstenbrand F, Lücking CH. Die akuten traumatischen
Hirnstammschäden. Arch Psychiatr Nervenkr. 1970;213:264–281.
8. Osborn AG. Diagnosis of descending herniation by cranial com-
puted tomography. Radiology 1977;123:93–96.
9. Reich JB, Sierra J, Camp W, et al. Magnetic resonance imaging meas-
urements and clinical changes accompanying transtentorial and
foramen magnum brain herniation. Ann Neurol. 1993;34:748–749.
10. Arfel G. Introduction to clinical and EEG studies in coma. In:
Remond A, Harner R, Naquet R, eds. Handbook of
Electroencephalography and Neurophysiology, vol. 12. 1975;5–23.
11. Bricolo A, Turella G. Electroencephalographic patterns of acute
traumatic coma: diagnostic and prognostic value. J Neurosurg Sci.
1973;17:278–285.
12. Chatrian GE, White JR, Daly P. Electroencephalographic patterns
resembling those of sleep in certain comatose states after injuries to
the head. Electroencephalogr Clin Neurophysiol. 1963;15:272–280.
13. Rumpl E, Lorenzi E, Hackl JM, et al. The EEG at different stages of
acute secondary traumatic midbrain and bulbar brain syndromes.
Electroencephalogr Clin Neurophysiol. 1979;46:487–497.
14. Silverman D. Retrospective study of EEG in coma.
Electroencephalogr Clin Neurophysiol. 1963;15:486–503.
15. Stone JL, Ghaly RF, Hughes JR. Electroencephalography in acute
head injury. J Clin Neurophysiol. 1988;5:125–133.
16. Synek VM. Value of a revised EEG coma scale for prognosis after
cerebral anoxia and diffuse head injury. Clin Electroencephalogr.
1990;21:25–30.
17. Jung R. Neurophysiologische untersuchungsmethoden. In:
Bergmann G, Frey W, Schwiegk H. eds. Handbuch der Inneren
Medizin V, vol. 1. Berlin: Springer; 1953:1286–1293.
18. Lücking CH, Müllner E, Pateisky K, et al. Electroencephalographic
findings in the apallic syndrome. In: Dalle Ore G, Gerstenbrand F,
Lücking CH, et al. eds. The Apallic Syndrome. Berlin/Heidelberg/
New York: Springer; 1977:144–154.
19. Meyer-Mickeleit RW. Das Elektroenzephalogramm nach gedeck-
ten Kopfver-letzungen. Dtsch Med Wschr. 1953;1:480–484.
20. Radermecker J. Das EEG bei gedeckten Hirnschäden und seine
Beziehung zu den subjektiven Beschwerden. Med Wschr. 1964;
106:1315–1322.
21. Courjon J, Scherzer E. Traumatic disorders. In: Remond A, Magnus
O, Courjon J. eds. Handbook of Electroencephalography and Clinical
Neurophysiology, vol. 14. Amsterdam: Elsevier; 1972:8–95.
Chapter 22 ■ EEG and Craniocerebral Trauma 431
22. Jabbari B, Vengrow MI, Salazar AM, et al. Clinical and radiological
correlates of EEG in the late phase of head injury: a study of 515
Vietnam veterans. Electroencephalogr Clin Neurophysiol. 1986;64:
285–293.
23. Koufen H, Dichgans J. Häufigkeit und Ablauf von traumatischen
EEG-Veränderungen und ihre klinischen Korrelationen. Fortschr
Neurol Psychiatr. 1978;46:165–177.
24. Koufen H, Hagel KH. Systematic EEG follow-up study of trau-
matic psychosis. Eur Arch Psychiatr Neurol Sci. 1987;237:2–7.
25. Strnad P, Strnadová V. Long-term follow-up EEG studies in patients
with traumatic apallic syndrome. Eur Neurol. 1987;26:84–89.
26. Walter W, Hill D, Williams D. Discussion of the electroencephalogram
in organic cerebral disease. Proc Soc Med Lond. 1948;41:237–250.
27. Williams D. The electro-encephalogram in acute head injuries.
J Neurol Psychiatry 1941;4:107–130.
28. Dow R, Ulett G, Raaf J. Electroencephalographic studies immedi-
ately following head injury. Am J Psychiatr. 1944;101:174–183.
29. Dawson EE, Webster JE, Gurdjian ES. Serial electroencephalogra-
phy in acute head injuries. J Neurosurg. 1951;8:613–630.
30. Schneider E, Hubach H. Das EEG der traumatischen Psychosen.
Dtsch Z Nervenheilkd. 1962;183:600–627.
31. Duensing F. Die Alphawellenaktivierung als Herdsymptom in
Elektroencephalogramm. Nervenarzt. 1948;19:544–552.
32. Scherzer E. Wert der Elektroenzephalographie beim Schädel-
trauma. Wien Klin Wschr. 1965;77:543–547.
33. Davis PA, Davis H. The electrical activity of the brain. Its relation
to physiological states and to states of impaired consciousness. Res
Publ Assoc Res Nerv Ment Dis. 1939;19:50–80.
34. Hill D. Psychiatry. In: Hill D, Parr G, eds. Electroencephalography.
London: MacDonald; 1950:319–363.
35. Schwab RS. Electroencephalography in Clinical Practice.
Philadelphia, PA: Saunders; 1951:195.
36. Fischgold H, Mathis P. Obnubilations, comas et stupeurs, etudes
electro-encephalographiques. Electroencephalogr Clin Neurophysiol.
1959;11(suppl):27–68.
37. Loeb C. Electroencephalographic changes during the state of
coma. Electroencephalogr Clin Neurophysiol. 1958;10:589–606.
38. Mathis P, Turrubia H, Fischgold H. Réactivité, périodicité et cor-
rélation cortico-cardio-respiratoire dans le coma. Electro-
encephalogr Clin Neurophysiol. 1957;6(suppl):27–68.
39. Naquet R, Vigouroux RP, Choux Bourand C, et al. Étude élec-
troencéphalographique des traumatismes craniens récents dans
un service de ré-animation. Rev Neurol. 1967;117:512–513(abst).
40. Bergamasco B, Bergamini L, Doriguzzi T, et al. EEG sleep patterns
as a prognostic criterion in post-traumatic coma. Electro-
encephalogr Clin Neurophysiol. 1968;24:374–377.
41. Bricolo A, Gentilomo A, Rosadini G, et al. Akinetic mutism follow-
ing cranio-cerebral trauma. Physiopathological considerations
based on sleep studies. Acta Neurochir (Wien). 1968;18:68–77.
42. Valente M, Placidi F, Oliveira AJ, et al. Sleep organization pattern
as a prognostic marker at the subacute stage of post-traumatic
coma. 2002. Clin Neurophysiol. 2002;113:1798–1805.
43. Fischgold H, Torrubia H, Mathis P, et al. Réaction EEG d’éveil
(arousal) dans le coma. Corrélations cortico-cárdio-respiratoires.
Presse Méd. 1955;61:1231–1233.
44. Fischer C, Mutschler V. Traumatic brain injuries in adults: from
coma to wakefulness. Neurophysiological data. Ann Readapt Med
Phys. 2002;45:448–455.
45. Courjon J, Naquet R, Baurand C, et al. Valeur diagnostique de
l’E.E.G. dans les suites immédiates des traumatismes crâniens. Rev.
EEG Neurophysiol. 1971;1:133–150.
432 Part III ■ Clinical EEG: General
46. Chatrian GE. Electrographic and behavioral signs of sleep in com-
atose states. In: Remond A, Harner R, Naquet R, eds. Handbook of
Electroencephalography and Clinical Neurophysiology, vol. 12.
Amsterdam: Elsevier; 1975:63–77.
47. Vigouroux RP. Between life and death. In: Remond A, Harner R,
Naquet R, eds. Handbook of Electroencephalography and Clinical
Neurophysiology, vol. 12. Amsterdam: Elsevier; 1975:95–98.
48. Stockard JJ, Bickford RG, Aung MH. The electroencephalogram in
traumatic brain injury. In: Vinken, PJ, Bruyn, GW, eds. Handbook
of Clinical Neurology, vol. 23. Amsterdam/Oxford: North-Holland;
1975;217–367.
49. Williams D, Denny-Brown D. Cerebral electrical changes in exper-
imental concussion. Brain. 1941;64:223–238.
50. Denny-Brown D, Russel WR. Experimental cerebral concussion.
Brain. 1941;64:93–164.
51. Ommaya AK, Gennarelli A. Cerebral concussion and traumatic
unconsciousness. Brain. 1974;97:633–654.
52. Ommaya AK, Grupp RL, Jr, Naumann RA. Coup and contre-coup:
observations on the mechanics of visible brain injuries in the rhe-
sus monkey. J Neurosurg. 1971;35:503–517.
53. Ommaya AK, Corrao PG, Letcher F. Head injury in the chim-
panzee: part I. Biodynamics of traumatic unconsciousness. J
Neurosurg. 1973;39:152–166.
54. Shaw NA. The neurophysiology of concussion. Prog Neurobiol.
2002;67:281–344.
55. Gloor P, Ball G, Schaul N. Brain lesions that produce delta waves in
the EEG. Neurology. 1977;27:326–333.
56. Akert K, Koella WP, Hess R, Jr. Sleep produced by electrical stimu-
lation of the thalamus. Am J Physiol. 1952;168:260–267.
57. Hunter J, Jasper HH. Effects of thalamic stimulation in unanaes-
thetized animals. Electroencephalogr Clin Neurophysiol. 1949;1:
305–324.
58. Jouvet M. Telencephalic and rhombencephalic sleep in the cat. In:
Wolstenholme GEW, O’Connor M. eds. The Nature of Sleep.
Boston, MA: Little, Brown; 1961:188–208.
59. Fischer CM. The neurological examination of the comatose
patient. Acta Neurol Scand. 1969;36(suppl):45–56.
60. Hockaday JM, Potts F, Epstein E, et al. EEG changes in acute cere-
bral anoxia from cardiac or respiratory arrest. Electroencephalogr
Clin Neurophysiol. 1965;18:575–586.
61. Kubicki S, Rieger H, Busse G. EEG in fatal and near fatal poisoning
with soporific drugs. I. Typical EEG patterns. Clin
Electroencephalogr. 1970;1:5–13.
62. Prior PF. The EEG in Acute Cerebral Anoxia. Amsterdam: Excerpta
Medica; 1973:43–46.
63. Gerstenbrand F, Rumpl E. Das prolongierte Mittelhirnsyndrom
traumatischer genese. In: Neumärker KJ. ed. Hirnstammläsionen.
Leipzig: Hirzel; 1983:236–248.
64. Bricolo, A. Neurosurgical exploration and neurological pathology
as means for investigating human sleep semiology and mechanism.
In: Lairy GC, Salzarulo P. eds. Experimental Study of Human Sleep:
Methodological Problems. Amsterdam: Elsevier; 1975:55–82.
65. Hess R. Sleep and sleep disturbances in the electroencephalogram.
Prog Brain Res. 1965;18:127–139.
66. Lücking GH. Sleep-like patterns and abnormal arousal reactions in
brain stem lesions. Electroencephalogr Clin Neurophysiol.
1970;28:214 (abst).
67. Rumpl E. Elektro-neurologische Korrelationen in den frühen Phasen
des posttraumatischen Komas. I. Das EEG in den verschiedenen
Phasen des akuten traumatischen sekundären. Mittelhirn und
Bulbärhirnsyndroms. Z. EEG-EMG. 1979;10:148–157.
68. Hutchinson DO, Frith RW, Shaw NA, et al. A comparison between
electroencephalography and somatosensory evoked potentials for
outcome prediction following severe head injury. Electroence-
phalograph Clin Neurophysiol. 1991;78:228–233.
69. Rae-Grant AD, Barbour PJ, Reed J. Development of a novel EEG
rating scale for head injury using dichotomous variables.
Electroencephalogr Clin Neurophysiol. 1991;79:349–357.
70. Chokroverty S. “Alpha-like” rhythms in electroencephalograms in
coma after cardiac arrest. Neurology (Minneapolis) 1975;25:655–663.
71. Synek VM, Synek BJL. Theta pattern coma, a variant of alpha pat-
tern coma. Clin Electroencephalogr. 1984;15:116–121.
72. Horton EJ, Goldie WD, Baram TZ. Rhythmic coma in children.
J Child Neurol. 1990;5:242–247.
73. Chatrian GE, Lowell MD, White E, Jr., et al. EEG pattern resem-
bling wakefulness in unresponsive decerebrate state following
traumatic brain stem infarct. Electroencephalogr Clin Neurophysiol.
1964;16:285–289.
74. Jasper HH, Van Buren J. Interrelationships between cortex and
subcortical structures. Clinical electroencephalographic studies.
Electroencephalogr Clin Neurophysiol. 1953;4(suppl):168–202.
75. Steudel WI, Krüger J, Grau H. Zur Alpha- und Spindel-Aktivitt bei
komatösen Patienten nach einer Schädel-Hirn-Verletzung unter
besonderer Berücksichtigung der Computertomographie. Z. EEG-
EMG. 1979;10:143–147.
76. Rumpl E. Elektro-neurologische Korrelationen in den frühen
Phasen desposttraumatischen Komas. II. Das EEG im Übergang
zum, und im Vollbild des traumatischen apallischen Syndroms. Z.
EEG-EMG. 1980;11:43–50.
77. Goulon M, Pocidalo JJ, Christophe M, et al. Clinical, electroen-
cephalographic and biological correlations in 34 cases of bron-
chopneumonia with asphyxia. In: Gastaut H, Meyer JS. eds.
Cerebral Anoxia and the Electroencephalogram. Springfield, IL:
Charles C Thomas; 1959:565–577.
78. Merill JP, Hampers CL. Uremia (part 1). Engl J Med. 1970;289:
953–961.
79. Wilson WP, Sieker HL. A study of the factors responsible for
changes in the electroencephalogram in chronic pulmonary insuf-
ficiency. Electroencephalogr Clin Neurophysiol. 1958;10:89–96.
80. Vespa PM, Miller C, McArthur D, et al. Nonconvulsive electro-
graphic seizures after traumatic brain injury result in a delayed,
prolonged increase in intracranial pressure and metabolic crisis.
Crit Care Med. 2007;35:2830–2836.
81. Grand W. The significance of post-traumatic status epilepticus in
childhood. J Neurol Neurosurg Psychiatry. 1974;37:178–180.
82. Kiloh LG, Osselton JW. Clinical Electroencephalography. London:
Butterworths; 1966;125.
83. Bauer G, Niedermeyer E. Acute convulsions. Clin Electroen
Cephalogr. 1979;10:127–144.
84. Jordan KG. Continuous EEG and evoked potential monitoring in
the neuroscience intensive care unit. J Clin Neurophysiol. 1993;
10:445–475.
85. Toyonaga K, Schlagenhauff RE, Smith BH. Periodic lateralized
epileptiform discharges in subdural hematoma. Case-reports and
review in literature. Clin Electroencephalogr. 1974;5:113–118.
86. Chatrian GE, Shaw CM, Leffman H. The significance of periodic
lateralized epileptiform discharges in the EEG: an electroen-
cephalographic, clinical and pathologic study. Electroencephalogr.
Clin Neurophysiol. 1964;17:177–193.
87. Markand ON, Daly DD. Pseudoperiodic lateralized paroxysmal
discharges in electroencephalogram. Neurology (Minneapolis).
1971;21:975–981.

88. Bickford RG, Butt HR. Hepatic coma: the electroencephalic pat-
tern. J Clin Invest. 1955;34:790–799.
89. Hall RC, Joffe JR. Hypomagnesemia: physical and psychiatric
symptoms. JAMA. 1973;244:1749–1751.
90. Jennet B, Bond M. Assessment of outcome after severe brain
damage. Lancet. 1975;1:480–484.
91. Rumpl E, Prugger M, Bauer G, et al. Incidence and prognostic
value of spindles in post-traumatic coma. Electroencephalogr Clin
Neurophysiol. 1983;56:420–429.
92. Bergamasco B, Bergamini L, Bricolo A, et al. Studies on sleep dur-
ing the apallic syndrome. In: Dalle Ore G, Gerstenbrand F,
Lücking CH, et al., eds., The Apallic Syndrome. Berlin/Heidel-
berg/New York: Springer; 1977;155–159.
93. Hansotia P, Gottschalk P, Green P, et al. Spindle coma: incidence,
clinicopathologic correlates, and prognostic value. Neurology
(Minneapolis) 1981;31:83–87.
94. Hughes JR, Boshes B, Leestma J. Electro-clinical and pathologic
correlations in comatose patients. Clin Electroencephalogr.
1976;7:13–30.
95. Lorenzoni E. Das EEG im posttraumatischen Koma. Fortschr
Neurol Psychiatr. 1975;43:155–191.
96. Vigouroux R, Naquet R, Baurand C, et al. Évolution électroradio-
clinique de comas graves prolongés post-traumatiques. Rev
Neurol. 1964;110:72–81.
97. Bricolo A, Turazzi S, Faccioli F, et al. Clinical application of com-
pressed spectral array in long-term EEG monitoring of comatose
patients. Electroencephalogr Clin Neurophysiol. 1978;45:211–225.
98. Dusser A, Navelet Y, Devictor D, et al. Short- and long-term prog-
nostic value of the electroencephalogram in children with severe
head injury. Electroencephalogr Clin Neurophysiol. 1989;73:85–93.
99. Britt RH, Herrick MK, Hamilton RD. Traumatic locked-in syn-
drome. Ann Neurol. 1977;1:590–592.
100. Miller JD. Barbiturates and raised intracranial pressure (editor-
ial). Ann Neurol. 1979;6:189–193.
101. Myers RR, Stockard JJ, Saidman LJ. Monitoring of cerebral per-
fusion during anesthesia by time-compressed Fourier analysis of
the electroencephalogram. Stroke. 1977;8:331–337.
102. Newlon PG, Greenberg RP, Enas GG, et al. Effects of therapeutic
pentobarbital coma on multimodality evoked potentials
recorded from severely head-injured patients. Neurosurgery.
1983;12:613–619.
103. Overgaard J, Hvid-Hansen O, Land AM, et al. Prognosis after
head injury based on early clinical examination. Lancet.
1973;2:631–635.
104. Avenarius HJ, Gerstenbrand F. The transition stage from mid-
brain syndrome to the traumatic apallic syndrome. In: Dalle Ore
G, Gerstenbrand F, Lücking CH, et al., eds. The Apallic Syndrome.
Berlin/Heidelberg/New York: Springer; 1977;22–25.
105. Hörtnagl H, Hammerle AF, Hackl JM, et al. The activity of the
sympathetic nervous system in the course of severe head injury.
Intensive Care Med. 1980;6:169–177.
106. Gerstenbrand F, Rumpl E, Prugger M. Die apallische
Symptomatik. In: Wiek HH., ed. Neurotraumatologie, Derzeitige
Schwerpunkte, 8. Internationales Symposium Erlangen 1979.
Stuttgart: Thieme; 1980:235–239.
107. Gerstenbrand F. Das Traumatische Apallische Syndrome.
Wien/New York: Springer; 1967.
108. Dolce G, Kaemmerer E. Contribute anatomico-clinico ed elet-
troencephalografico all conoscenza della sindromi apallica.
Studio dell evoluzione dell EEG da sonne in 5 casi. Sist Nerv
1967;1:12–23.
Chapter 22 ■ EEG and Craniocerebral Trauma 433
109. Butenuth J, Kubicki S. Klinisch-elektroenzephalographische
Schlafbeobachtungen im apallischen Syndrome. Z EEG-EMG.
1975;6:185–188.
110. Passouant P, Cadilhac J, Delange M, et al. Différentes stades élec-
triques et organization en cycle des comas posttraumatiques.
Enrégistrement polygraphique de long durée. Rev Neurol.
1964;111:391(abst).
111. Cadilhac J, El Kassabgui M, Passouant P. La réorganisation du
sommeil nocturne après les comas posttraumatiques. Rev Neurol.
1966;115:529(abst).
112. Bergamasco B, Bergamini L, Doriguzzi T, et al. EEG sleep pat-
terns as a prognostic criterion in post-traumatic coma. Electro-
encephalogr Clin Neurophysiol. 1968:24:374–377.
113. Chase TN, Moretti L, Prensky AL. Clinical and electroencephalo-
graphic manifestations of vascular lesions of the pons. Neurology
(Minneapolis) 1968;18:357–368.
114. Ferguson JM, Benett DR. Sleep in a patient with pontine infarc-
tion. Electroencephalogr Clin Neurophysiol. 1974;36:210–211(abst).
115. Rumpl E, Prugger M, Bauer G. Zum prognostischen Wert elek-
troenzephalographischer Schlafbeobachtungen im traumatisch
bedingten apallischen syndrom. Neuropsychiatr Clin. 1984;3:
219–232.
116. Biniek R, Ferbert A, Rimpel J, et al. The complete apallic syn-
drome—a case report. Intensive Care Med. 1989;15:212–215.
117. Ingvar DH, Brun A. Das komplette apallische syndrom. Arch
Psychiatr Nervenheilkd. 1972;215:219–239.
118. Giacino JT, Ashwal S, Childs N, et al. The minimally conscious state:
definition and diagnostic criteria. Neurology. 2002;58:349–353.
119. Kelly JP, Rosenberg JH. Diagnosis and management of concus-
sion in sports. Neurology (Minneapolis) 1997;48:575–580.
120. Denny-Brown D. Brain trauma and concussion. Arch Neurol.
1961;5:1–3.
121. Tönnis A. Beobachtungen an frischen gedeckten Hirnschädi-
gungen. In: Rehwald E., ed. Das Hirntrauma. Stuttgart: Thieme;
1956:772.
122. Verjaal A. Trauma capitis. In: van den Bergh R, Folkerts JF, eds.
Neurologie vor de Algemene Praktijk. Brussels: Agon-Elsevier;
1972:170–187.
123. Voller B, Benke T, Benedetto K, et al. Neuropsychological, MRT
and EEG findings after very mild traumatic brain injury. Brain
Inj. 1999;13:821–827.
124. Yamamoto LG, Bart RT. Transient blindness following mild head
trauma. Clin Pediatr. 1989;27:479–483.
125. Hess R. Significance of EEG signs for localization of cerebral
tumours. Electroencephalogr. Clin Neurophysiol. 1961;19(suppl):
75–110.
126. Mifka P, Scherzer E. Über die Wertigkeit des EEG im Spätstadium
der Gehirnverletzung. Wien Klin Wschr. 1962;74:573–576.
127. Dumermuth G. Elektroencephalographie im Kindesalter. Stutt -
gart: Thieme; 1965:262–266.
128. Lenard HG. EEG-Veränderungen bei frischen Schädeltraumen
im Kindesalter. Med Wschr. 1965;107:1820–1827.
129. Melin KA. Electroencephalography following head injuries in
children. Acta Pediatr (Uppsala). 1949;75:152–174.
130. Christian W. Klinische Elektroenzephalographie. Lehrbuch und
Atlas. Stuttgart: Thieme; 1968:264–274.
131. Vigouroux RP, Naquet R, Gastaut H. Aspects particuliers de cer-
taines manifestations épileptiques précoces lors de traumatismes
crânio-cérébraux survénuschez des sujects âgés. Rev Neurol.
1966;115:976–980.
132. Jennett WB. Early traumatic epilepsy. Lancet. 1969;1:1023–1025.
434 Part III ■ Clinical EEG: General
133. Courjon J, Revol M. Traumatismes crâniens et épilepsie. Cah Méd
Lyon. 1966;42:1343–1350.
134. Black K, Obayan A, Zafonte RD, et al. Occipital status epilepticus:
an unusual case of posttraumatic blindness. Neuro Rehabili-
tation. 1996;7:219–221.
135. Kazibutowska Z, Stelmach-Wawrzyczek M, Majchrzak R. Results
of prospective 24-hour EEG studies of patients after craniocere-
bral injuries. Neurol Neurochir Pol. 1992;26:304–310.
136. Wohlrab G, Schmitt B, Botshauser E. Benign focal epileptiform
discharges in children after severe head trauma: prognostic value
and clinical course. Epilepsia. 1997;38:275–278.
137. Gibbs FA, Gibbs, EL. Electroencephalographic study of head
injury in childhood. Clin Electroencephalogr. 1987;18:10–11.
138. Niedermeyer E, Knott J. Über die Bedeutung der 14 and 6/sec
positiven Spitzen in EEG. Arch Psychiatr Nervenkr. 1961;202:
266–280.
139. Gerstenbrand F. The symptomatology of the apallic syndrome.
In: Dalle Ore G, Gerstenbrand F, Lücking CH, et al., eds. The
Apallic Syndrome. Berlin/Heidelberg/New York: Springer;
1977:14–21.
140. Caveness WF. Posttraumatic sequelae. In: Caveness WF, Walker
AE, eds. Head Injury, Conference Proceedings. Philadelphia/
Toronto: Lippincott; 1966:209–219.
141. Verjaal A, Van’t Hooft F. Commotio and contusio cerebri (cere-
bral concussion). In: Vinken PJ, Bruyn GW, eds. Handbook of
Clinical Neurology, vol. 23. Amsterdam/Oxford: North-Holland;
1975:417–444.
142. Jellinger K, Gerstenbrand F, Pateisky K. Die protrahierte form der
posttraumatischen Enzephalopathie. Nervenarzt. 1963;34:
145–163.
143. Pateisky K, Gerstenbrand F, Jellinger K. The EEG in the posttrau-
matic apallic syndrome. Electroencephalogr Clin Neurophysiol.
1963;15:713(abst).
144. Goetze W. Das Hirnstrombild bei offenen Hirnverletzungen.
Mschr Unfallheilk. 1953;56:297–305.
145. Courjon J. La place de l’électroencéphalographie en traumatolo-
gie crânienne. Cah Méd Lyon. 1962;38:315–317.
146. Kugler J. Elektroenzephalographie in Klinik und Praxis 2. Aufl.,
Stuttgart: Thieme; 1966:203.
147. Torres F, Shapiro SK. Electroencephalograms in whiplash injury.
Arch Neurol. 1961;5:28–35.
148. Kuhlendahl K. Die neurologischen syndrome bei der
Überstreckungsverletzung der Halswirbelsäule und dem soge-
nannten Schleudertrauma. M Med Wschr. 1964;106:1025–1030.
149. Mifka P, Scherzer E. Zur Pathogenese zerebraler Symptomatik bei
Verletzungen der Halswirbelsäule. Münch Med Wschr. 1962;10:
1686–1690.
150. Jacome DE. EEG in whiplash: a reappraisal. Clin Electroence-
phalogr. 1987;18:41–45.
151. Cant BR, Shaw NA. Monitoring by compressed spectral array in
prolonged coma. Neurology. 1984;34:35–39.
152. Sironi VA, Ravagnati L, Signorini G, et al. Diagnostic and prog-
nostic value of EEG compressed spectral analysis in posttraumatic
coma. In: Vilani R, Papo J, Giovanelli M, et al. eds. Advances in
Neurotraumatology. Amsterdam: Excerpta Medica; 1983:328–330.
153. Steudel WI, Krüger J. Using the spectral analysis of the EEG for
prognosis of severe brain injuries in the first post-traumatic
week. Acta Neurochir. 1979;28(suppl):40–42.
154. Pfurtscheller G, Schwartz G, List W. Long-lasting EEG reactions
in comatose patients after repetitive stimulation. Electro-
encephalogr Clin Neurophysiol. 1986;64:402–410.
155. Klein HJ, Rath SA, Göppel F. The use of EEG spectral analysis
after thiopental bolus in the prognostic evaluation of comatose
patients with brain injuries. Acta Neurochirurg. 1988;
42(suppl):31–34.
156. Thatcher RW, Walker RA, Gerson I, et al. EEG discriminant
analysis of mild head trauma. Electroencephalogr Clin
Neurophysiol. 1989;73:94–106.
157. Thatcher RW, Cantor DS, McAlaster R, et al. Comprehensive pre-
dictions of outcome in closed head-injured patients. Ann NY
Acad Sci. 1991;620:82–101.
158. Hooshmand H, Beckner E, Radfor F. Technical and clinical
aspects of topographic brain mapping. Clin Electroencephalogr.
1989;20:235–247.
159. Jerrett SA, Corsak J. Clinical utility of topographic EEG brain
mapping. Clin Electroencephalogr. 1988;19:134–143.
160. Busse EW, Silverman AJ. Electroencephalographic changes in
professional boxers. JAMA. 1952;149:1522–1525.
161. Larsson LE, Melin KA, Nordström-Öhrberg BO, et al. Acute head
injuries in boxers. Acta Psychiatr Scand. 1954;95(suppl):1–42.
162. Pampus E, Grote W. Elektroencephalographische und klinische
Befunde bei Boxern und ihre Bedeutung fr die Pathophysiologie
der traumatischen Hirnschädigung. Arch Psychiatr Nervenkr.
1956;194:152–178.
163. Saunders RL, Harbaugh RE. The second impact in catastrophic
contact sports head trauma. JAMA. 1984;252:538–539.
164. Roberts AH. Brain Damage in Boxers. London: Pitman Medical
Scientific Publishing; 1969.
165. Haglund Y, Persson HE. Does Swedish amateur boxing lead to
chronic brain damage? A retrospective clinical neurophysiologi-
cal study. Acta Neurol Scand. 1990;82:353–360.
166. Tysvaer AT, Storli OV, Bachen NI. Soccer injuries to the brain. A
neurologic and electroencephalographic study of former players.
Acta Neurol Scand. 1989;80:151–156.
167. Tysvaer AT. Head and neck injuries in soccer. Impact of minor
trauma. Sports Med. 1992;14:200–213.
168. Martland HS. Punch drunk. JAMA. 1928;91:1103–1107.
169. Millspaugh JA. Dementia puglistica (punch drunk). US Nav Med
Bull. 1937;35:297–303.
170. Critchley M. Medical aspects of boxing, particularly from a neu-
rological standpoint. Br Med J. 1957;1:357–362.
171. Mawdsley C, Ferguson FR. Neurological disease in boxers. Lancet.
1963;2:795–801.
172. Johnson J. The EEG in the traumatic encephalopathy of boxers.
Psychiatr Clin. 1969;2:204–211.
173. Forster JB, Leiguarda R, Tilley PJB. Brain damage in national
hunt jockeys. Lancet. 1976;1:981–987.

Anoxia, Coma, and Brain Death
PETER W. KAPLAN AND GERHARD BAUER
CLINICAL DEFINITIONS OF IMPAIRED
RESPONSIVENESS: FROM LOCKED-IN
STATES TO COMA
Consciousness
Consciousness can be described as an awareness of self and the
environment. The two primary components of consciousness
are awareness and arousal with the determination of conscious-
ness being made by the examination of the patient’s response to
stimuli. The examination of brainstem reflexes and responses to
command is helpful in localizing the causes of coma to higher
or lower CNS structures.
A further component of awareness depends on cortical
integrity and subcortical connections. When these are dis-
rupted, a patient may have a decreased consciousness to the
point of being in a minimally conscious state or a vegetative state,
but less frequently is in actual coma (1).
Coma
Coma is a state of unconsciousness without wakefulness, char-
acterized by the lack of arousal from internal or external stim-
uli. The eyes are usually closed, and there is no awareness of self
or surroundings. Brainstem reflexes may be retained. Coma is
usually a transient condition that evolves to either wakefulness
or death over days to weeks.
Minimally Conscious States (MCS)
Relatively recently, a minimally conscious state (MCS) has
been declared a separate clinical condition (2). A MCS is that
of a severely altered consciousness. Behavioral manifestations
of awareness are minimal, but present. There may be fluctuat-
ing attention, as well as purposeful visual tracking and com-
munication, but all these elements should be present. fMRI has
demonstrated some patterns of brain activity in minimally
conscious patients that are also seen in normal people (3).
These patterns are thought to represent conscious cognitive
processing.
Vegetative States (VS)
In vegetative states (VS), the patient is awake. However, there
is no evidence of awareness of self or of the environment,
although sleep–wake cycling occurs. There is also no evidence
(during wakefulness) of cognitive processing. A VS is persistent
if it has lasted at least a month, but “persistent” does not neces-
sarily imply irreversibility. A permanent VS is irreversible, and
CHAPTER
23
is often diagnosed after about 3 months in nontraumatic brain
injury, or a year after brain trauma. Different reports of this
syndrome have used names such as apallic syndrome (4), com-
plete and incomplete apallic syndrome, prolonged coma (5),
and persistent vegetative state (PVS) (6). Absence of any clini-
cal sign of cortical function combined with alternating periods
of sleep and wakefulness is the essential of this condition. The
eyes open periodically without evidence of appreciation of the
environment. More recently, multimodal imaging has shed
light on brain activity in the VS (1,3,7). Pathologically, no com-
mon pattern of lesion can be defined. Prognosis, EEG signs, and
appropriate management rest on the strict or broadened defini-
tion of this condition (8–11). Terminological and prognostic
controversies have been cleared by the Multi-Society Task Force
on PVS (12). Questions remain about the boundaries to coma
and VS (12).
Locked-in Syndrome (LIS)
An LIS consists of almost total paralysis of limb and head
movements so that the patient cannot move or speak, but
remains fully conscious (13,14). It is due to the interruption of
corticospinal and corticobulbar pathways, typically in the basis
pontis or midbrain (15–20) or from bilateral lesions in the
internal capsules (21), or less frequently from total peripheral
paralysis (ALS, polyradiculopathy). This pathology disconnects
the brain from the ability to converse (interruption of efferent
pathways). The only retained movements are of the eyelids and
of vertical eye movements. The patient can communicate with
a pattern of eye movements that the patient and observer have
established to allow discourse. In this way, the patient demon-
strates awareness and hence consciousness. EEG has shown dif-
ferent patterns, but the most typical is a posterior, often reactive
alpha rhythm (22,23).
Akinetic Mutism
The term akinetic mutism was coined by Cairns and colleagues
(1941) to describe a state of mutism, or minimal speech, and
hence communication (24). There are a number of anatomical
lesions that can cause this syndrome, and it is a term less used
at present.
Coma Vigile
A term also in declining use, it was used to refer to several of the
syndromes mentioned above including LIS, akinetic mutism
(6), or the apallic syndrome (25).
435
436 Part III ■ Clinical EEG: General
Sleep
Although sleep represents a transient absence of consciousness,
a patient can be aroused to conscious wakefulness. In light
coma, it can be difficult to differentiate pathologic hypersom-
nolence (e.g., from drugs, alcohol, or sleep deprivation) from
physiologic sleepiness.
THE ANATOMICAL AND PHYSIOLOGIC
SUBSTRATES OF COMA
Consciousness is maintained by the integration of ascending
arousal inputs via the ascending reticular activating system
through the thalami to both cerebral hemispheres. In animals
and humans, these ascending pathways pass through the pons
and midbrain tegmentum (26). Structural or functional
impairment or interruption of these pathways in the brainstem
or in both cerebral hemispheres leads to decreased conscious-
ness and coma (27,28).
Structural Causes of Coma
The signal process needed to induce coma is compromise of
brainstem function. This typically happens from compression
or ischemia. Compression can arise from unilateral cerebral
mass lesions exerting traction and distortion of the brainstem
structures or compromise of circulation. Although bilateral
cerebral hemisphere dysfunction without brainstem involve-
ment can suppress cortical activity, coma per se rarely occurs.
Further, a markedly raised ICP in itself does not induce coma,
except by brainstem compromise. Pathologic examination and
radiologic studies have demonstrated the effects of uncal and
central herniation on diencephalic structures that mediate
arousal (29), but more current imaging techniques (CT and
MRI) give greatest credence to the transverse or horizontal type
of herniation compromise (30,31), while evidence of down-
ward, transtentorial herniation can be obtained with MRI
(32–34). Brainstem compromise in the latter injury arises more
from anoxia than direct pressure.
More direct destruction by compression, trauma, or
ischemia of the paramedian pontomesencephalic reticular for-
mation can occur at the level of the pons and midbrain (35,36),
although a direct pontine effect by trauma has been questioned
(37). Other authors note rare cases of traumatic brainstem
coma supported by MRI (38). Diffuse axonal injuries from tor-
sional or shearing stresses can markedly adversely affect prog-
nosis and be more difficult to evidence in imaging studies (39).
Toxic/ Metabolic Causes of Coma
Despite the prevalence and longstanding recognition of these
conditions as causes of coma, there is a relative paucity of infor-
mation on their mechanism. It is believed that synaptic transmis-
sion is variably impaired, with neuronal energy failure occurring
late in the process. There is early impairment of complex interac-
tions among the cortical regions and polysynaptic pathways pro-
ducing an initial impairment of higher cortical functions. In this
way, judgment, memory, attention, and concentration are early
heralds of impending encephalopathy and coma (40). Coma is
believed to be produced by the induction of brainstem or brain
edema (41–43). The progressive impairment of consciousness
mirrors, to some extent, that of anesthesia induction.
A number of mineral and electrolyte disturbances can all
contribute to encephalopathy, and in the extreme, to coma.
These include hyper- and hyponatremia, hyper- and hypocal-
cemia, hyper- and hypoglycemia, and uremia. Contributing to
these electrolyte disturbances, or constituting disorders in
themselves, are metabolic and endocrine disorders, and organ
failures such as renal failure and dialysis effects, hypothy-
roidism, hypocortisolism, and porphyria. There are rarely if any
pathognomonic EEG patterns with these encephalopathies and
coma (44,45). The patterns reflect largely the degree of suppres-
sion (and level of coma), the presence of associated epilepti-
form disturbances, and occasionally focal effects. One major
exception is hepatic failure that may frequently be reflected on
EEG by triphasic waves (TWs) and progressive disorganization
of the EEG as coma deepens even though such waves can be
seen in other conditions (see later on in this chapter).
Drug intoxications in coma can produce patterns that in part
reflect their cortical effects. Barbiturates progressively suppress
the EEG; benzodiazepines suppress voltage, increase the power of
frequencies in the beta range, and eventually slow the back-
ground. Lithium may produce epileptiform discharges, slowing,
and TWs (46). Some stimulant drugs increase, voltage and fre-
quencies, but often do not produce coma. For more detailed
pathophysiology and EEG findings on toxic/metabolic and organ
failure impairment of CNS function, please turn to Chapter 42.
Infection and Coma
Infection of the CNS, of individual body organs, as well as sep-
ticemia can all induce confusion and encephalopathy, but only
direct CNS infection readily causes coma (47–49). A large num-
ber of microbes may directly affect the brain and/or the
meninges, along with the increasing emergence of patients with
the acquired immunodeficiency syndrome (AIDS). In addition,
there are conditions that may simulate infections. Conditions
such as neoplasms, autoimmune conditions, and slow viruses
are dealt elsewhere in this text. The pathophysiology of infection
and coma has not been clearly elucidated. Mechanisms impli-
cated include a direct release of toxins and their effect on CNS
structures, the involvement of microvasculature and larger ves-
sels, cerebral edema, a focal or generalized increase in intracra-
nial pressure, and changes in cerebral blood flow (CBF), which
may all play a part. Once again, EEG changes in these various
conditions reflect the focality, epileptogenicity, level of con-
sciousness, and the mechanism of CNS compromise produced
by the respective infectious process as it induces coma.
Seizures, Epileptiform Activity, Status
Epilepticus, and Coma
Epileptic activity when generalized usually causes coma.
Seizures of themselves and in the form of ongoing activity
(status epilepticus) generally may cause either transient or a
more lasting loss of consciousness hours. Some conditions with
partial brain seizure activity may impair vigilance, but do not in
all cases cause coma. Coma from many causes may additionally
reveal epileptiform activity and seizure activity as an “epiphe-
nomenon.” In this sense, the underlying coma etiology is not

that of seizure activity, but rather the reflection of cortical irri-
tability or insult. Please see also Chapters 25, 26, and 28 to 30.
Anoxia and Coma
Because anoxia is such a common cause of coma, as well as being
one of its most morbid etiologies, more attention will be paid to
this entity in this chapter. The brain at rest consumes 20% of the
total body oxygen consumption, even though it represents only
2% by weight (50). Consciousness and normal background EEG
patterns are impaired if the brain oxygen tension falls below 15 to
20 mm Hg (51). An interruption in oxygen supply results in cell
dysfunction, and ultimately death, and can be induced by a
decrease in CBF, hypoxia or anoxia, a low oxygen concentration
in respired air (hypoxic hypoxia). When ischemia reduces CBF
below 15 mL/min/100 g of tissue for several minutes, there may
be subsequent cell death and degeneration over the ensuing days,
worsened by injury produced by resumed blood flow or reperfu-
sion (52–54). The postischemic period before cell death may rep-
resent a therapeutic window of opportunity. Causes of global
ischemia include cardiorespiratory arrest (CRA), circulatory col-
lapse, strangulation, drowning, suffocation, muscle paralysis, and
toxins such as carbon monoxide. The less frequent hypoxia in
pure form is rare as cardiac dysrrythmias usually also occur, as
may systemic hypotension—both producing superadded
ischemic damage (55). Clinical signs of anoxic–ischemic brain
injury largely follow the rate and severity of the deficiency.
Unconsciousness with cardiac arrest occurs in about 8 seconds in
the upright position, and about 15 seconds supine, followed by
tonic posturing (56). When anoxic coma is prolonged, there may
be persisting deficits, with permanent cognitive decline, rigidity,
ataxia and spasticity, amnesia, visual dysfunction, movement dis-
orders, and in some cases persistent coma or a VS (57). Epileptic
phenomena include myoclonus, myoclonic status, or stimulus-
sensitive myoclonus (15,58–63). It is important to distinguish
between the morbid postanoxic myoclonus in coma from the
entity in which the patient returns to consciousness, but with
stimulus-induced myoclonus following briefer anoxia/hypoxia
(64,65). When the hypoxia is less marked, survivors are able to
speak, but may have some degree of cognitive deficit. CT of the
head may be normal in the early hours of the day, but will even-
tually reveal edema, sulcal effacement, basal ganglia lucency, and
watershed changes (66,67). With MRI, there is better resolution
of abnormalities in the globus pallidus, putamen, caudate nuclei,
hippocampus, cerebellum, brainstem, and subcortical white mat-
ter (68–73). Magnetic resonance spectroscopy (MRS), diffusion-
weighted MRI, and positron emission tomography (PET) are all
more sensitive for hypoxic insults, and at an earlier point. With
PET, De Volder et al. (1990) showed hypometabolism that corre-
lated with both location and severity of hypoxia. PET studies of
patients in a VS have uncovered evidence of irreversible damage
to supratentorial cortical zones (74,75).
EEG IN COMA
EEG patterns in coma may correlate with the depth of coma
(76). As consciousness decreases, there is a slowing of the preva-
lent EEG frequencies from alpha toward delta. When there are
excess theta frequencies in the posterior alpha rhythm, it may be
Chapter 23 ■ Anoxia, Coma, and Brain Death 437
difficult in adults to differentiate light coma from drowsiness.
Occasionally, higher frequencies can be seen, for example, in the
beta range, particularly with benzodiazepines and barbiturates.
High doses of CNS-suppressant drugs can decrease EEG voltage.
Drug effects aside, low-voltage or flatline tracings after anoxia
predict an abysmal prognosis for return to consciousness.
Intermittent Delta Rhythms
Frontal intermittent rhythmic delta activity (FIRDA) (77) can be
seen in mild obtundation, but rarely occurs in coma. It is believed
to arise most frequently from diffuse cerebral dysfunction (78),
but was initially described in deep midline dysfunctions in sub-
cortical regions (79), deep frontal (80), and with early herniation.
When FIRDA exists with slow background activity, a more dif-
fuse cerebral dysfunction, including toxic, metabolic, or infec-
tious etiologies, may be suspected (81,82). (see Fig. 23.1).
Prolonged Bursts of Delta Waves and the
Reactivity of EEG
Some patients with head injury have paroxysms of potentiated
slow activity, lasting up to minutes (83–86). These bursts may
occur spontaneously or after stimulation, particularly as the
patient tries to communicate.
In several of the patterns described above, spontaneous EEG
variability or change in EEG patterns after stimulation indicates
a more favorable prognosis. Conversely, the loss of variability or
reactivity to stimuli can forebode deepening coma.
Epileptiform Morphologies
Epileptiform morphologies can be seen in many causes of
coma, even without apparent clinical or electrographic seizure
activity. Continuous, unilateral epileptiform discharges with
little variability suggest pseudoperiodic lateralized epilepti-
form discharges (PLEDs), an indication of cortical irritability,
believed by most authors to fall short of actual seizures.
Frequent causes are structural abnormalities (e.g., strokes,
trauma) with a recent seizure. Such patterns are frequently seen
after seizure resolution in these settings, (87–90). When the dis-
charges are more generalized and synchronous, usually with a
relative paucity of intervening background activity, the term
generalized periodic epileptiform discharges (GPEDs) has been
used (91). This pattern usually reflects a more severe, diffuse
cerebral insult such as is seen after anoxia (92), CNS infections,
or less frequently in the end-stages of status, and correspond-
ingly indicates a worse prognosis. Generalized discharges can be
accompanied by myoclonic movements, or conversely with lit-
tle evident motor manifestations (15,92,93). Bilateral inde-
pendent discharges (BIPLEDs) probably lie in between with
regards to both severity of etiology and prognostic significance
(88). There is ongoing controversy regarding where on the
ictal–interictal continuum these periodic discharge patterns lie,
and hence the need (or not) of more intensive seizure suppres-
sion with anesthetic agents or anesthetic coma. Evidence that
PLEDs represent a form of nonconvulsive status epilepticus
(94) arises from PET studies revealing increased metabolism or
blood flow (95). In most cases, parenteral AEDs fail to lastingly
suppress PLEDs or improve the patient’s level of consciousness.
See Figures 23.2 and 23.3.
438 Part III ■ Clinical EEG: General

Figure 23.1 This EEG shows

medium- to high-voltage theta and
a predominance of arrhythmic delta
activity in an unresponsive patient
with multiple organ failure includ-
ing renal and respiratory failure.

Triphasic Waves (TWs)
Although this pattern was initially attributed to patients with
hepatic failure and subsequently with raised serum ammonia,
later case series found this pattern with several different toxic or
metabolic problems, including renal failure (96). Some authors
have sought to distinguish a “typical” from an “atypical” form
(Figs. 23.4 and 23.5) (97,98). Typical TW pattern consist of
medium- to high-voltage three-phased bilateral, symmetric
waveforms in runs, with a discharge frequency of 1.5 to 2.5 Hz.
Typical TWs possess an anterior–posterior phase lag, which often
is not seen when using transverse or referential electrode arrays.
TWs are seen with hepatic coma, after hypoxia (99–101), and fol-
lowing toxic exposure (see Fig. 23.6) (2,6,96,102,103). A number
of other metabolic problems have been reported with TWs
(78,104). Structural lesions have occasionally been reported with
TWs, including brainstem strokes and subdural hematomas
(105). Infection (106) and malignancy (cerebral carcinomatosis)
may also induce this pattern (107). Degenerative and spongi-
form encephalopathies may also produce TWs in the conscious
patient (108). An important differential diagnostic consideration
is the continuously occurring sharp- and slow-wave runs seen
with atypical absence status in Lennox–Gastaut syndrome (109).

Figure 23.2. This EEG sample

shows pseudoperiodic lateralized
epileptiform discharges (PLEDS) at
about 1 Hz over the right hemi-
sphere in an unresponsive patient.

Burst-Suppression Patterns
This pattern is largely associated with deeper levels of coma,
and prominently occurs with anoxic insults after cardiac arrest.
In this setting it carries an abysmal prognosis. Other more
benign causes are in the setting of anesthetic use. In the thera-
peutic suppression of status epilepticus, propofol and other
agents typically produce a burst-suppression pattern on the way
to producing more extensive, global cerebral suppression.
Overdoses or therapeutic use of barbiturates can also induce
reversible patterns of burst-suppression. The pattern typically
consists of high-voltage bursts of sharply contoured morpholo-
gies, and polyspikes and spikes superimposed or associated
with underlying slow-wave activity. These paroxysms may last
Figure 23.4 This EEG shows a pattern along the continuum of interictal to ictal, generalized periodic discharges (GPEDs)
in a patient who has had a prolonged cardiorespiratory arrest, but before therapeutic hypothermia.
Chapter 23 ■ Anoxia, Coma, and Brain Death 439
Figure 23.3 This EEG shows gene -
ralized pseudoperiodic epileptiform
discharges (GPEDs) in an anoxic
patient.
from less than a second to upwards of 10 seconds, but are char-
acterized as burst-suppression patterns by the appearance of
marked diffuse bilateral suppression of brain EEG activity
between these bursts. The periods of suppression again may
vary from about 1 second to periods exceeding 10 seconds.
There may be a variable amount of background activity during
these suppression periods (110,111). The bursts typically repeat
quasi-periodically, appear diffusely and bilaterally, and exist
generally throughout the recording (112). They may be associ-
ated with single or salvos of myoclonic face, arm, chest, or leg
jerks (15,93,113,114). In many patients, the jerks can be
induced by stimulating the patient by touch, suctioning, or loud
sounds. See Figure 23.7.
440 Part III ■ Clinical EEG: General
Figure 23.5 This EEG shows the effects of lithium toxicity. There are runs of triphasic waves and some sharper, epilepti-
form morphologies seen diffusely, bilaterally. The background is slow. The patient was obtunded during the recording, but
recovered over several weeks.
Alpha and Theta Patterns in Coma
Alpha frequency patterns may be recorded in some deep comas.
Originally described with brainstem dysfunction (strokes,
tumors, trauma) (115–119), they were also recorded after
marked anoxia (119–124). Differentiating alpha coma from nor-
mal, physiologic waking alpha are its diffuse and indeed frontal
predilection, general lack of reactivity to noxious stimulation (or
at least a failure to induce a return to discernible clinical con-
sciousness or wakefulness), and its steady presence. Alpha coma
must be distinguished from the spindle-like patterns that can
also appear in unresponsive states (125,126). Drug toxicity may
cause faster rhythms in the 10- to 18-Hz range (123). In the
unresponsiveness of the LIS due to de-efferentation, the awake,
but paralyzed patient will evince waking alpha frequencies, typ-
ically reactive to eye-opening and closure. Alpha coma may fol-
low burst-suppression pattern (111,125), heralding death. Some
studies including unresponsive patients with alpha frequencies
have found that prognosis may vary according to the etiology of
coma, and the presence of a degree of reactivity. Unreactive pat-
terns and anoxia, or brainstem infarction carried the worst prog-
nosis (see Fig. 23.8) (115–130).
Sleep-Like Patterns in Coma
Some coma EEG patterns contain spindles, slow activity, and
vertex waves, resembling stage II sleep (81,131). They are usu-
ally seen after head trauma or structural lesions of the brain-
stem (126,132,133), but may be seen with viral encephalitis
(134), and benzodiazepines (135). Such patterns may evolve
to waking patterns, or diffuse slowing, and are thought to
represent functional impairment of arousal mechanisms on
thalamocortical circuits responsible for sleep patterns. With
deepening coma, there is less variability and cycling, and
the appearance of more monomorphic slow activity (see Fig.
23.9) (136).
Low -Voltage or Diffusely Suppressed Patterns
Marked voltage suppression can precede electrocerebral inac-
tivity (typically after anoxia), or occur (reversibly) with anes-
thetic suppression of cortical activity. It is usually defined as the
totality of EEG-recorded brain activity that remains below
20 V. Low-voltage records can occur in conscious patients
after head trauma, alcoholism, and hydrocephalus, and are not
to be confused with the morbid pattern seen in coma.

Figure 23.6 The EEG shows disorganized activity, with no dominant alpha background activity. There are high-voltage
triphasic waves, some exhibiting an anteroposterior lag. This patient had baclofen toxicity.
EEG AND THE ETIOLOGY OF COMA
The EEG is not specific for etiology of coma, but can reflect
some types of underlying disease.
Infectious and Inflammatory Processes
With massive cerebral infection, there may be prominent, diffuse,
often rhythmic slowing, with few faster frequencies in the alpha
and theta ranges. With the late, immune disorder of measles-
associated subacute sclerosing panencephalitis (SSPE), there may
be extremely high-voltage complexes reaching 300 V, lasting 1
to 2 seconds, with a periodicity exceeding 5 to 10 seconds (137)
or Jakob–Creutzfeldt disease (138,139). These patterns can also
exist in other diffuse encephalitides. Periodic complexes are typ-
ically lateralized in the encephalitis of herpes simplex (140,141).
EEG in Metabolic Coma
The enormous complexity of pathophysiologic mechanisms in
metabolic coma stands in contrast to the lack of specificity of
accompanying EEG signs. Diffuse slowing is always seen in
metabolic coma (see Fig. 23.10) (44,45). In cases of dispropor-
tionate degrees of coma and slowing, a metabolic cause is
unlikely (78). Focal slowing does not exclude the diagnosis. TWs
Chapter 23 ■ Anoxia, Coma, and Brain Death 441
are suggestive of a metabolic disorder (see Fig. 23.11), especially
of hepatic origin (46,96,102,103,142). The diagnostic signifi-
cance of TWs may have been overstressed, but it is still impor-
tant when this pattern is encountered in an undiagnosed,
insidiously developing coma. With hepatic coma, spikes may
also be encountered. Uremia is frequently complicated by gener-
alized tonic–clonic seizures and generalized polyspikes associ-
ated with diffuse slow activity. Photomyoclonic responses are
not infrequent. Diffuse slowing may become exaggerated during
or after dialysis with or without clinical signs of a disequilibrium
syndrome. Paroxysmal bursts of slow waves and slowing of the
basic rhythm were seen in all cases of the dialysis dementia syn-
drome (143,144). Modern advances in dialysis techniques made
this complication an exceptional event.
Epileptic Conditions
Epileptic conditions lead to prolonged coma in convulsive
status epilepticus, postictal states with lingering subclinical
paroxysmal activities, typical and atypical absence status
(145,147,148), and other types of nonconclusive status epilep-
ticus (145–148). In most instances, the EEG shows prominent
seizure activity and is instrumental in the establishment of
the correct diagnosis, although an EEG without spikes does not
442 Part III ■ Clinical EEG: General

Figure 23.7 This EEG shows a burst-suppression pattern in a patient who had anoxia and cardiac arrest. There are bilate -
ral, generalized polyspike–slow-wave complexes. The patient died without regaining consciousness.

categorically exclude the epileptic nature of the altered state of
consciousness. On the other hand, many comatose states are
complicated by interspersed epileptic seizures. Coma with acute
convulsions is accompanied by a markedly abnormal EEG,
which in most instances is characterized by seizure discharges
(see Fig. 23.12) (15). Occasionally, seizures on EEG take the
form of rhythmic theta and delta activity (see Fig. 23.13).
EEG in Relation to the Depth of Coma
The EEG as a functional test should be closely related to the
degree of impaired cerebral function. There are several reasons
why this relation is far from being perfect. The main reason is
found in local brainstem coma, which leaves the cortex and the
EEG comparatively unaffected.

Figure 23.8 This composite of four 10-second segments of EEG shows a diffuse alpha frequency pattern, present also
anteriorly. After two arousal efforts, the EEG continues to show alpha frequencies. This pattern constitutes alpha coma.
 Chapter 23 ■ Anoxia, Coma, and Brain Death 443

Figure 23.9 This EEG shows sleep

spindle patterns in an unresponsive
patient, consistent with a spindle
coma pattern. This pattern carries a
good prognosis for return to con-
sciousness, if not due to anoxia.

Attempts to classify coma into clinical stages are numerous. rostral–caudal deterioration can be supported by EEG evi-
The most logical coma grading is based on an orderly sequence dence in the following manner:
of neurologic symptoms corresponding to the levels of the
1. The degree and generalization of slowing is related to the
neuraxis in cases of transtentorial herniations. Regardless of
level of unresponsiveness (78,149,150). The most important
the etiology, the following information about the grade of

Figure 23.10 This EEG shows sinusoidal diffuse theta activity in a comatose patient, from hypoglycemia.
444 Part III ■ Clinical EEG: General

Figure 23.11 There are runs of

triphasic waves seen synchronously,
bilaterally in this patient with hyper-
ammonemia.

exception to this rule is prolonged bursts of delta waves They might be mediated by the lower brainstem or even by
secondary to exogenous stimuli. Furthermore, in children the spinal cord and can be present with cortical death and
the degree of slowing is frequently disproportionate to the electrocerebral silence (ECS).
clinical state.
EEG activity seen during sleep becomes progressively scarcer
2. The effect of stimulation yields good information about the
and finally disappears with deepening coma (132,152–154).
depth of coma. As one proceeds down the scale of coma, the
Patterns highly suggestive of a late midbrain or initial bulbar
blocking type of response is replaced by the alerting type.
brain syndrome are (i) progressive voltage depression
Finally, the EEG becomes unreactive even to repeated stim-
(83,84,151); (ii) extreme slowing with extinction of superim-
ulation (83,84,151). Motor responses without changes of
posed fast activities (84); (iii) intermittent isoelectric periods;
EEG should not be regarded as a sign of cortical functions.

Figure 23.12 The high-voltage

bursts of generalized polyspike–slow-
wave activity seen synchronously,
bilaterally in this lethargic patient
with an idiopathic generalized
epilepsy indicate generalized non-
convulsive status epilepticus. Note
that in contrast to some burst-
suppression patterns after anoxia,
there is background activity in the
form of spindles, alpha, and theta
activity.

Figure 23.13 This EEG shows rhythmic theta and delta waves, diffusely, but with slight frequency shift over time from
side-to-side. The patient was in coma with limbic encephalitis.
(iv) periodic spiking or burst-suppression activity (15,84,101);
and (v) monorhythmic unreactive alpha and theta frequencies
(119,155) and spindle patterns (132).
EEG ABNORMALITIES IN ANOXIA
AND ANOXIC COMA
Basic Mechanisms of EEG Changes
The EEG reflects decreases in cerebral metabolism due to
anoxia. As pO2 falls, there is a fall in membrane potential pro-
gressing to a cessation of electrical activity. EEG voltage declines
to flatline (51). An increase in glutamate is associated with cell
death, initially affecting the forebrain, followed by brainstem
structures (156).
EEG Changes during Arrest of Cerebral
Circulation and w ith Syncope
A progression of abnormalities follows the cerebral ischemia
that arises from hypotension and cardiac arrest. In the first sev-
eral seconds, little change is seen, but after about 10 seconds,
there is higher voltage slow activity, followed by flattening as the
period of ischemia increases. Recovery may occur in the reverse
order.
Chapter 23 ■ Anoxia, Coma, and Brain Death 445
EEG in Anoxic Coma
Patients in postanoxic coma may exhibit a number of EEG pat-
terns including TWs, continuous epileptiform discharges in
“electrographic status in coma,” burst-suppression patterns,
and monorhythmic alpha, theta, or mixed alpha–theta activity.
The patient typically is deeply comatose, and may have
myoclonic jerks and an EEG reflective of bursts of epileptiform
activity (status myoclonicus) (59,61,157,158).
In prolonged anoxia, the EEG patterns are similar to those
seen in other causes of coma. There may be diffuse slowing (see
Fig. 23.14) (100,101). FIRDA, a pattern of frontal delta waves,
can also be seen, but is nonspecific, occurring in deep midline
dysfunction as well as in other diffuse cerebral dysfunctions,
particularly in the elderly (82) (Fig. 23.15). Another pattern in
postcardiac arrest coma is the appearance of sharp activity
with some morphologies assuming a triphasic appearance
(99,101), PLEDs (Fig. 23.2), or bilaterally independent peri-
odic discharges (BIPLEDs) (88). With prolonged, severe
hypoxia, there may be intermittent suppression with bursts of
spikes. A frank burst-suppression EEG pattern (Figs. 23.16
and 23.17) indicates severe prolonged diffuse cortical and
subcortical insults (15). Similarly, there is variability and evo-
lution among patterns of burst-suppression, and alpha–theta
coma patterns overtime. As some have noted, there is often an
446 Part III ■ Clinical EEG: General

Figure 23.14 This EEG is from a patient several days after cardiorespiratory arrest. It shows generalized, sinusoidal 4- to
5-Hz activity, and carries a moderate prognosis for good outcome. The patient awoke with significant cognitive deficits.

inconstant association among clinical signs and paroxysmal
EEG activity. With spike or burst activity, there may be sponta-
neous or stimulus-induced eye-opening, and facial and limb
myoclonus, but similar activities can occur without an EEG
correlate (15,159). As coma progresses, posturing can regress,
suggesting further involvement of brainstem structures. Some
anoxic comas induce alpha and theta frequency patterns
(100,119,127,155,109–161). Both can occur in the same record
(130) or appear in sequence. Reactivity signals a better progno-
sis, although overall the mortality of these patterns is 80% to
90% (127). An irreversible depolarization of membrane poten-
tial is reflected as a flatline EEG pattern.

Figure 23.15 This EEG panel shows a

burst of frontal intermittent, rhythmic
delta activity.
 Chapter 23 ■ Anoxia, Coma, and Brain Death 447

Figure 23.16 This EEG panel

shows polyspike bursts on a sup-
pressed background in a patient
after cardiac arrest and anoxia. The
patient was comatose with
myoclonic face and limb move-
ments. The patient died without
regaining consciousness.

PROGNOSIS IN COMA USING EEG
Electrophysiology is optimally used in guiding early diagnostic
investigation, and in excluding other differential considera-
tions including severe encephalopathy, NCSE, or psychogenic
issues.
The EEG is certainly of great prognostic value in comatose
states (162). Its significance is limited by the fact that the outcome
is often not determined by the brain disturbance itself but by the
underlying metabolic or cardiovascular problems. Furthermore,
prognostic criteria do not apply if anesthetics or other CNS-
depressant drugs have been used. This fact has become especially
important with the widespread use of sedative drugs in comatose
states. In these patients, neurologic and EEG examinations may be
misleading, and valuable monitoring seems to be confined to
evoked potentials (163–165). The barbiturate treatment of coma-
tose states, however, has never been generally accepted (166,167),
and its use has declined progressively (168,169).
In general, the prognostic significance of EEG patterns par-
allels their correlations to the grade of rostral–caudal deteriora-
tion. Prognostically favorable signs are reactions to exogenous
stimuli (83,170–172) and normal-looking sleep potentials or sleep
cycles (83,131,152,173–175). Spindle coma indicates relative
integrity of the cerebral hemispheres (126). Its prognostic signif-
icance depends on the underlying etiology (176). Best outcomes
occurred when spindle coma was due to drugs, encephalopathy,

Figure 23.17 This EEG shows pro-

longed suppression with variable
polymorphic bursts occurring syn-
chronously, bilaterally in a patient
who had anoxia and cardiac arrest.
This is also a burst-suppression pat-
tern. Care was withdrawn for this
patient.
448 Part III ■ Clinical EEG: General
Figure 23.18 This patient was
obtunded with an excess of benzodi-
azepines. The EEG shows diffuse 20-
to 25-Hz beta activity, reflecting the
medication effect, in a pattern
resembling sleep spindles, but of
higher frequency.
or seizures (Fig. 23.18). With posttraumatic coma, normal-
looking sleep spindles carry a good prognosis (Fig. 23.9). On the
other hand, several posttraumatic cases exhibiting spindle coma
have a poor outcome probably due to prominent primary or sec-
ondary brainstem lesions (174,176). The absence of spindles in
anoxic coma was associated with poor outcome, whereas its pres-
ence did not indicate a favorable course (177). Improvement of
EEG signs at serial recordings may represent the first indication
of a favorable course.
On the other hand, absence of reactivity has been judged an
ominous sign (83). With rostrocaudal deterioration, the number
of patterns decreases (153). A very poor prognosis is indicated by
monotonous, high-voltage 0.5- to 3-Hz activity without changes
occurring spontaneously or secondary to exogenous stimuli
(178), the burst-suppression pattern (15,100,101,159,179–181),
generalized periodic phenomena (100,120,180,181), intermittent
flattening, by monorhythmic alpha frequencies (83,100,119,120),
and a low-voltage-output EEG (15,100,101,120,159,179). With
intoxications, the aforementioned patterns are of less serious sig-
nificance. TWs indicate a poor prognosis in liver diseases (182; see
Fig. 23.11). Paroxysmal abnormalities in coma represent an unfa-
vorable sign (120), except in posttraumatic coma (15).
Patients typically evolve over hours to days from comatose
states. A typical progression would be from coma to alertness,
coma to death, or, in a minority of patients, coma to VS. In
some reports, progression from coma to VS is not accompanied
by a change in EEG. In some cases, EEG change will not be
reflected in a clinical change.
PROGNOSTIC CRITERIA FROM EEG
For prognosis, EEG can be obtained shortly after an irreversible
insult such as ischemia/anoxia from CRA. It can then be used to
help predict outcome, and the probabilities of making a good,
poor, or no recovery. After anoxic coma, EPs also provide reliable
prognostication of a poor or negative outcome. Combined with
EPs, an Index of Global Cortical Function (IGCF) can also be
generated to provide a graded prognostication, but frequently
the EP testing resources are not available in many hospitals,
excepting some academic centers. Imaging techniques are still
largely experimental, but may be useful in individual cases.
When patients reach the chronic stages with LIS, VS, PVS, or
MCS, electrophysiology does not accrue further prognostic value.
Vegetative States
Regarding VS, there is a wide variety of possible EEG patterns,
including normal reactive alpha, diffuse slowing and ranging to
continuous arrhythmic delta slowing, seen in most patients.
Other types of pattern include an unreactive alpha, theta, or
spindle coma pattern, reactive theta background, or diffuse
voltage suppression consistent with electrocerebral inactivity.
Hence, EEG is of limited value in VS. There may be EEG evi-
dence of wakefulness or sleep activity. Such testing does not
evaluate consciousness, and EEG and SSEP may not reliably
separate VS from LIS in some patients.
PROGNOSIS IN ANOXIC COMA;
CLINICAL AND ELECTROPHYSIOLOGIC
CONSIDERATIONS
Methodologic approaches can vary widely in studies of coma.
The duration of anoxia is a major determinant, as is the duration
of CPR. The duration of hypoxia may be difficult to quantify, but
data suggest that if the coma lasts 2 to 3 days, there is rarely a
complete recovery (183). Most comas lasting a few hours recover
(184). About 80% of CA survivors regained consciousness in 1
hour after CPR (185), while others have found that about 20% of
patients who are unconscious for 10 minutes after CPR are nor-
mal or nearly normal at 1 year (186)—the remainder had persist-
ent severe deficits. Edgren et al. (1994) and Levy et al. (1981)
found that absent motor responses on day 3 with unreactive
pupils predicted PVS in all patients. Unreactive pupils at initial
examination foretold that 0/52 patients had a moderate or good
outcome (187,188). Wiilloughby and Leach (1974) also found
that no response or stereotyped reflexes at 1 hour after CA bode

a poor prognosis (189). Poor outcome occurred if at 3 days, no
motor response or flexor and extensor posturing was present. A
minority of 5% to 7% of patients with no pupillary light reaction
and motor responses had favorable outcomes (190). The pres-
ence of myoclonic status epilepticus after CA is also indicative of
a poor outcome (15,59–61). Briefer anoxia or hypoxia may result
in stimulus-sensitive myoclonus in an awake patient
(Lance–Adams syndrome), with good prognosis (64).
Overall 50% of out-of-hospital CA patients survive after
resuscitation (191–193), and 20% at 1 year (168); about one
sixth to one half will recover to normal wakefulness and cogni-
tion. Other workers report a 20% good recovery; 80% died or
vegetative (194).
After in-hospital CPR, more than 90% of patients had a
favorable recovery (195). The use of EEG in predicting outcome
has been the subject of much study over the years
(110,125,127,137,196–203). Quite aside from the challenges
faced in determining the value of EEG in accurately predicting
prognosis for the range of possibilities from normal to death or
PVS are the problems of interference in producing accurate EEG
data because of administered medications that modify the EEG
or the clinical examination. Furthermore, more recent use of
hypothermia also modifies the interpretation of EEG and clini-
cal findings. Electrophysiology (EEG and evoked potentials) has
been most successful in accurately prognosticating patients at
the extremes of outcome—that is to say those who will do well
or very poorly. Patients who lie in between these extremes have
been far more problematic with regards to accurate prognostica-
tion (100,101,159,200,204). Furthermore, the time of electro-
physiologic assessment has also been found to be important,
particularly for EEG. Early EEG ( 24 hours) after coma onset is
less reliable for prognostic purposes as a record obtained days
Figure 23.19 This EEG shows a pattern that is intermediate between that of generalized periodic discharges (GPEDs) and
generalized electrographic status epilepticus in coma. This patient has had prolonged cardiorespiratory arrest.
Chapter 23 ■ Anoxia, Coma, and Brain Death 449
later, and there may be a progression through several EEG pat-
terns of coma, seizure, or prognostic significance in the early
period (200,205). A number of investigators have promulgated
different EEG scales of pattern severity over the years, with sub-
division varying from 5 to 13 grades, and with greater or lesser
homogeneity of the study populations. Early work by Fischgold
and Mathis, and by Prior (1974) provided excellent qualitative
guidance to the types of abnormalities, and some insight into
the evolution over time (84). Scales by Hockaday as early as
1965, later modified by Synek (1988, 1990), Scollo-Lavizzari and
Bassetti (1987), and Young and colleagues (2004, 2005, 2006),
have led to increasing use of grading scales to predict outcome
after CRA (159,206). For the high and low grades, an accuracy of
above 80% has been touted by a number of investigators. Data
analysis of the early hours of EEG after CRA may yield greater
opportunities for therapeutic intervention. The availability of
electrophysiology has a proven role in the decision-making of
ICU specialists in the withdrawal of care.
All scales agree that the EEG patterns of electrographic status
in coma, status myoclonicus with a burst-suppression pattern, a
low-voltage EEG, or ECS are reliable indicators of death or a VS
(Figs. 23.19 to 23.23). More recent studies have in part removed
patterns such as spindle or alpha coma from the extremely poor
prognosis, particularly if there is some reactivity (160,176).
Others have more systematically excluded EEG responsiveness to
external stimuli in the diagnosis of these coma patterns (com-
plete or incomplete alpha/theta coma), and hence found the
worse prognosis (119,125). A similar distinction can be found
with types of burst-suppression patterns. A summary of our
present understanding of the reliability of electrophysiology in
CRA outcome has been provided by the excellent meta-analysis
of 33 studies, totaling more than 4500 patients. Reliable predictors
450 Part III ■ Clinical EEG: General
Figure 23.20 In this obtunded patient after cardiac arrest,
there is diffuse theta activity with some arrhythmic delta
activity, representing an intermediate grade 3 on the Synek
scale of recovery after anoxic insult.
of poor outcome (death or PVS) were found to be the absence of
early cortical somatosensory evoked responses (0% to 2% for
95% confidence interval [CI] of pooled false-positive tests), elec-
trographic seizures, grade 4 to 5 EEG ( 5 V or flatline tracing),
and burst-suppression (196). From this analysis and for subse-
quent prognostication, the authors suggest a stepwise approach
beginning 3 days after coma onset. Those patients without pupil-
lary reflexes or motor response to pain (better than flexion)
should be tested with SSEPs. Patients lacking cortical responses
can then be considered as having no prognosis for meaningful
recovery and guided toward the option of palliative care. The pre-
dictive value of neuroimaging, and of particular laboratory
parameters in the serum and cerebrospinal fluid fall outside the
scope of this chapter and will not be addressed here.
BRAIN DEATH
Definition
The entire problem of cerebral death is a by-product of improved
medical care; it has become urgent because of the need for organs
for transplantation surgery. Numerous terms are used for the
same clinical entity: (i) aperceptive, areactive, apathic, and atonic
syndrome (183); (ii) brain death; (iii) stage IV coma (84); (iv)
coma depasse; (v) irreversible coma (5); (vi) cerebral death;
Figure 23.21 This EEG sample is from a patient after
cardiac arrest, with a Glasgow Coma Scale of 3. A flatline
tracing like this is classified as a 5 on the 5-point Synek
scale, and represents electrocerebral inactivity. If after car-
diac arrest and in the absence of hypothermia, or CNS-
suppressant drugs, it carries a 100% bad outcome.
(vii) cerebral death syndrome; and (viii) irreversible breakdown
of cerebral functions.
From a neurologic point of view, there are two problems with
non-heart-beating donors. No data exist about how long the heart
beat has to be arrested until the brain is irreversibly out of any
function. Category 3 patients, according to Koostra et al. (1995),
are ICU patients dying from severe brain trauma with no chance
for recovery but who do not meet the criterion of brain death. To
harvest organs, medical support is withdrawn until cardiac arrest
occurs. Ethically, this procedure can hardly be accepted. With the
other categories of non-heart-beating donors, neurologists and
electroencephalographers might not be involved in the decision-
making process.
Neurologic Signs
In the diagnosis of cerebral death, the first step is the demon-
stration of a brain death syndrome based on neurologic exam-
ination. Neurologic signs of cortical functions and brainstem
activity must be absent. Pupils must be fixed even with a strong
light stimulus. Peripheral third nerve injuries should be
excluded. Oculovestibular responses must be absent (185,186).
No motor activity must be observed in the muscles innervated
by cranial nerves. Spontaneous respiration must be absent; no
respiration movements should be detectable after removal from

the respirator. Testing for apnea is necessary to confirm the
neurologic brain death syndrome.
The practical procedure differs markedly within different
centers (186). A standardized procedure has been suggested for
apnea testing. The varying criteria used across the United States
have been reviewed, and a more standardized approach was
suggested.
Electrocerebral Silence (ECS)
An isoelectric EEG is confirmatory for the establishment of
cerebral death. It must be emphasized, however, that ECS per
se does not mean cerebral death. Isoelectric EEG recordings
have been found with intoxication and full recovery (187),
hypothermia (100), and transient decorticate states followed
by varying degrees of recovery (84). Therefore, ECS can be
taken as a sign of brain death only if neurologic signs of corti-
cal and brainstem functions are lacking, and intoxication and
marked hypothermia can be excluded. A brain death syndrome
at the neurologic examination is based on brainstem death and
does not exclude ongoing cortical activities reflected by EEG
activity (188,189). To consider EEG activity with brainstem
death as a pitfall of EEG in determining brain death, because
brainstem death equals brain death, turns the logic upside
down. Brainstem death does not mean brain death, defined as
irreversible loss of functions of all parts of the brain. Ongoing
Chapter 23 ■ Anoxia, Coma, and Brain Death 451
Figure 23.22 After anoxia this patient showed prolonged
bursts and brief suppressions (burst-suppression pattern),
and died.
EEG activity with isolated brainstem death proves this state-
ment. In cases of primary brainstem lesions, the EEG is not
only confirmatory, but also essential for correct diagnosis. The
same is true for fulminant demyelinating polyradiculopathy
resembling brainstem death.
The relationship between brain death diagnosed on the basis
of neurologic and EEG examinations and respirator brain at
autopsy has been reassessed. Overly rigid criteria for the differ-
entiation of ECS from the very low-voltage-output EEG and
excess noise have been criticized. In spite of certain shortcom-
ings, EEG has proved to be of paramount importance in the
evaluation of brain death. Shortcomings are of a technical
nature, as are conceptual misunderstandings, such as with pri-
mary brainstem death and the overinterpretation of artifacts.
However, real cerebral EEG waves exclude brain death by defi-
nition. To bypass these difficulties, several attempts have been
made to redefine brain death and to confine it to brainstem
death. The concept of whole-brain death delivers high diagnos-
tic practicability and certainty and has achieved a high level of
acceptance in the Western society. There is no urgent need to
change it to brainstem death or neocortical death.
The definition and reliability of ECS is based on the technical
standards applied for its registration. Many records referred to as
showing ECS show low-voltage-output EEG or use insufficient
technical standards. Several national EEG societies have published
Figure 23.23 This EEG tracing was obtained after CRA,
but during hypothermia. Although SE after CRA usually is
fatal, hypothermia may improve prognosis in some, but
complicates EEG interpretation and prognostication. Is this
GPEDs or status epilepticus?
452 Part III ■ Clinical EEG: General
recommendations for EEG recordings in suspected brain death.
These suggestions do not differ essentially among the national
recommendations. They include the following: a minimum of
eight scalp electrodes and reference electrodes to cover the major
brain areas; interelectrode impedances under 10,000 but over
100 ; testing the integrity of the entire recording system;
interelectrode distances of at least 10 cm to magnify the ampli-
tudes and to pick up electrical fields originating in deep struc-
tures; sensitivity increased up to 2 V/mm during most of the
recording to distinguish ECS from low-voltage-output EEG;
the use of time constants of 0.3 to 0.4 second; the use of mon-
itoring techniques, with simultaneous ECG recording to be
mandatory; testing of EEG reactivity to exogenous stimuli; a
recording time of at least 30 minutes; and a recording to be
made only by a qualified technologist. The most important
recommendation is the use of high instrument sensitivity
along with the need for continuous recognition and elimina-
tion of artifacts. This is indeed a painstaking task in a busy
ICU. If the EEG is chosen as a confirmatory laboratory test, the
length of observation time until a second and final EEG should
be recorded has been under discussion for a long time, with
intervals between recordings and clinical observation varying
from 6 to 24 hours. In cases with diagnostic or technical uncer-
tainties, repeated recordings are advised. The usefulness of
EEG in the evaluation of brain death in children has been dis-
cussed and remains under contention. With optimal technical
standards, similar criteria to those used in adults can be applied,
particularly for older children.
REFERENCES
1. Schiff ND, Fins JJ. Disorders of consciousness. Mayo Clin Proc.
2007;82(2):250–251.
2. Giacino JT, Ashwal S, Childs N, et al. The minimally conscious
state. Neurology. 2002;58:349–353.
3. Schiff ND. Bringing neuroimaging tools closer to diagnostic use in
the severely injured brain. Brain. 2007;30(pt 10):2482–2483.
4. Gerstenbrand F. Das Traumatische Apallische Syndrome. Wien-
New York: Springer-Verlag; 1967.
5. Vigouroux RP. Between life and death. In: Harner R, Naquet R,
eds. Altered States of Consciousness, Coma, Cerebral Death.
Handbook of Electroencephalography and Clinical Neurophysiology.
Vol 11. Amsterdam: Elsevier; 1975:95–99.
6. Jennet B, Plum F. Persistent vegetative state after brain damage. A
syndrome in search of a name. Lancet. 1972;1:734–737.
7. Schiff ND, Rodriguez-Moreno D, Kamal A, et al. fMRI reveals
large-scale network activation in minimally conscious patients.
Neurology. 2005;64(3):514–523.
8. Levin HS, Saydjari Ch, Eisenburg HM, et al. Vegetative state after
closed-head injury. Arch Neurol. 1991;48:580–585.
9. Molinari GF. Persistent vegetative state, do not resuscitate … and
still more words doctors use. J Neurol Sci. 1991;102:125–127.
10. Munsat ThL, Stuart WH, Cranford RE. Guidelines on the vegeta-
tive state. Neurology. 1989;39:123–124.
11. Spudis EV. The persistent vegetative state—1990. J Neurol Sci.
1991;102:128–136.
12. Fins JJ, Schiff ND. Shades of gray: new insights into the vegetative
state. Hastings Cent Rep. 2006;36(6):8.
13. Plum F, Posner JB. The Diagnosis of Stupor and Coma.
Philadelphia, PA: F.A. Davis; 1966.
14. Laureys S, Owen AM, Schiff ND. Brain function in coma, vegeta-
tive state, and related disorders. Lancet Neurol. 2004;3(9):537–546.
15. Bauer G, Niedermeyer E. Acute convulsions. Clin Electro-
encephalogr. 1979;10:127–144.
16. Chia LG. Locked-in syndrome with bilateral ventral midbrain
infarcts. Neurology. 1991;41:445–446.
17. Dehaene J, Dom R. A mesencephalic locked-in syndrome. J Neurol.
1982;227:255–259.
18. Hochmann MS, Sowers JJ, Bruce-Gregorius J. Syndrome of the
mesencephalic artery: report of a case with CT and necropsy find-
ings. J Neurol Neurosurg Psychiatry. 1985;48:1179–1181.
19. Karp JS, Hurtig HI. “Locked-in” state with bilateral midbrain
infarcts. Arch Neurol. 1974;30:176–178.
20. Uematsu D, Suematsu M, Fukunchi Y, et al. Midbrain locked-in
state with oculomotor subnucleus lesion. J Neurol Neurosurg
Psychiatry. 1985;48:952–956.
21. Chia LG. Locked-in state with bilateral internal capsule infarcts.
Neurology. 1984;34:1365–1367.
22. Gütling E, Isenmann S, Wichmann W. Electrophysiology in the
locked-in-syndrome. Neurology. 1996;46(4):1092–1101.
23. Jacome DE, Morilla-Pastor D. Unreactive EEG: pattern in locked-
in syndrome. Clin Electroencephalogr. 1990;21(1):31–36.
24. Wijdicks EF, Cranford RE. Clinical diagnosis of prolonged states of
impaired consciousness in adults. Mayo Clin Proc. 2005;80(8):
1037–1046.
25. Matsuo F. EEG features of the apallic syndrome resulting from
cerebral anoxia. Electroencephalogr Clin Neurophysiol. 1985;61(2):
113–122.
26. French JD, Magoun HW. Effects of chronic lesions in central cephalic
brain stem of monkeys. Arch Neurol Psychiatry. 1952;68:591–604.
27. Plum F. Organic disturbances of consciousness. In: Critchley M,
O’Leary JLO, Jennet B, eds. Scientific Foundations of Neurology.
Philadelphia, PA: F.A. Davis; 1972:193–201.
28. Parvizi J, Damasio R. Neuroanatomical correlates of brainstem
coma. Brain. 2003;126:1524–1536.
29. McNealy DE, Plum F. Brainstem dysfunction with supratentorial
mass lesions. Arch Neurol. 1962;7:10–32.
30. Ropper AH. Lateral displacement of the brain and level of con-
sciousness in patients with an acute hemispheral mass. N Engl
J Med. 1986;314:953–958.
31. Ropper AH. Syndrome of transtentorial herniation: is vertical dis-
placement necessary? J Neurol Neurosurg Psychiatry. 1993;56(8):
932–935.
32. Feldmann E, Gandy SE, Becker R, et al. MRI demonstrates
descending transtentorial herniation. Neurology. 1988;38:697–701.
33. Reich JB, Sierra J, Camp W, et al. Magnetic resonance imaging and
clinical changes accompanying transtentorial and foramen mag-
num brain herniation. Ann Neurol. 1993;33:159–170.
34. Wijdicks EFM, Miller GM. MR imaging of progressive downward
herniation of the diencephalon. Neurology. 1997;48:1456–1459.
35. Nyberg-Hansen R, Loken AC, Tenstad O. Brainstem lesion with
coma for five years following manipulation of the cervical spine.
J Neurol. 1978;218:97–105.
36. Dinsdale HB. Spontaneous hemorrhage in the posterior fossa.
Arch Neurol. 1964;10:200–217.
37. Mitchell DE, Adams JM. Primary focal impact damage to the
brainstem in blunt head injuries. Does it exist? Lancet. 1973;
II:215–218.
38. Gentry LR, Godersky JC, Thompson BH. Traumatic brain stem
injury: MR imaging. Radiology. 1989;171:177–187.
39. Adams JH, Graham DI, Genarelli TA, et al. Diffuse axonal injury in
non-missile head injury. J Neurol Neurosurg Psychiatry. 1991;54:
481–483.

40. Gastaut H, Fischer-Williams M. The physiopathology of epileptic
seizures. In: Field Y, et al., eds. Handbook of Physiology. Vol 1.
Washington: American Physiological Society; 1959:329–363.
41. New PFJ, Scott WR. Computerized Tomography of the Brain and
Orbit. Baltimore, MD: Williams & Wilkins; 1975.
42. Ruggiero G, Sabattini L, Nuzzo G. Computerized tomography and
encephalography. Neuroradiology. 1977;13:45–48.
43. Torack RM, Alcala H, Gado M, et al. Correlative assay of comput-
erized cranial tomography (CCT), water content and specific
gravity in normal and pathological post mortem brain.
J Neuropathol Exp Neurol. 1976;34:385–392.
44. Kaplan PW. Stupor and coma: metabolic encephalopathies. Suppl
Clin Neurophysiol. 2004;57:667–680.
45. Kaplan PW. The EEG in metabolic encephalopathy and coma.
J Clin Neurophysiol. 2004;21(5):307–318.
46. Kaplan PW, Birbeck G. Lithium-induced confusional states: non-
convulsive status epilepticus or triphasic encephalopathy?
Epilepsia. 2006;47(12):2071–2074.
47. Kalita J, Misra UK. EEG in dengue virus infection with neurologi-
cal manifestations: a clinical and CT/MRI correlation. Clin
Neurophysiol. 2006;117(10):2252–2256.
48. Clarke M, Newton RW, Klapper PE, et al. Childhood encephalopa-
thy: viruses, immune response, and outcome. Dev Med Child
Neurol. 2006;48(4):294–300.
49. Sejvar JJ, Haddad MB, Tierney BC, et al. Neurologic manifestations
and outcome of West Nile virus infection. JAMA. 2003;290(4):
511–515.
50. Clarke DD, Sokoloff L. Circulation and energy metabolism of the
brain. In: Siegel GJ, et al., eds. Basic Neurochemistry: Molecular,
Cellular, and Medical Aspects. 5th ed. New York, NY: Raven Press;
1994:645–680.
51. Martin RL, Lloyd HGE, Cowan AL. The early events of oxygen and
glucose deprivation: setting the scene for neuronal death. Trends
Neurosci. 1994;17:251–257.
52. Hori A, Hirose G, Kataoka S, et al. Delayed postanoxic
encephalopathy after strangulation. Arch Neurol. 1991;48:871–874.
53. Petito CH, Feldmann E, Pulsinelli WA, et al. Delayed hippocampal
damage in humans following cardiorespiratory arrest. Neurology.
1987;37:1281–1286.
54. Hallenbeck JM, Dutka AJ. Background review and current con-
cepts of reperfusion injury. Arch Neurol. 1990;47:1245–1254.
55. Miyamoto O, Auer RN. Hypoxia, hyperoxia, ischemia, and brain
necrosis. Neurology. 2000;54:362–371.
56. Heiss WD, Rosner G. Functional recovery of cortical necrosis as
related to degree and duration of ischemia. Ann Neurol. 1983;14:
294–301.
57. Snyder BD, Loewenson RB, Gumnit RJ, et al. Neurological progno-
sis after cardiac arrest: III. Seizure activity. Neurology. 1980;30:
1292–1297.
58. Jumao-as A, Brenner RP. Myoclonic status epilepticus: a clinical
and electroencephalographic study. Neurology. 1990;40:1199–1202.
59. Krumholz A, Stern BJ, Weiss HD. Outcome from coma after car-
diopulmonary resuscitation: relation to seizures and myoclonus.
Neurology. 1988;38:401–405.
60. Wijdicks EFM, Parisi JE, Sharbrough FW. Prognostic value of
myoclonic status in comatose survivors of cardiac arrest. Ann Neurol.
1994;35:239–243.
61. Young GB, Gilbert JJ, Zochodne DW. The significance of
myoclonic status epilepticus in postanoxic coma. Neurology. 1990;
40:1843–1848.
62. Gatzonis SD, Zournas Ch, Michalopoulos A, et al. Area-selective
stimulus-provoked seizures in post-anoxic coma. Seizure. 2001;10:
294–297.
Chapter 23 ■ Anoxia, Coma, and Brain Death 453
63. Lance JW, Adams RD. The syndrome of intention or action
myoclonus as a sequel to hypoxic encephalopathy. Brain. 1963;86:
111–133.
64. Morris HR, Howard RS, Brown P. Early myoclonic status and out-
come after cardiorespiratory arrest. J Neurol Neurosurg Psychiatry.
1998;64:267–268.
65. Werhan KJ, Brown P, Thompson PD, et al. The clinical features and
prognosis of chronic post hypoxic myoclonus. Mov Disord.
1997;12:216–220.
66. Kjos BO, Brant-Zawadzki M, Young RG. Early CT findings of
global central nervous system hypoperfusion. AJNR Am
J Neuroradiol. 1983;4:1043–1048.
67. Morimoto Y, Kemmotsu O, Kitami K, et al. Acute brain swelling
after out-of-hospital cardiac arrest: pathogenesis and outcome.
Crit Care Med. 1993;21:104–110.
68. Birbamer G, Aichner F, Felber S, et al. MRI of cerebral hypoxia.
Neuroradiology. 1991;33:53–55.
69. Christophe C, Fonteyne C, Ziereisen F, et al. Value of MR imaging
of the brain in children with hypoxic coma. AJNR Am
J Neuroradiol. 2002;23:716–723.
70. Sawada H, Udaka F, Seriu N, et al. MRI demonstration of cortical
laminar necrosis and delayed white matter injury in anoxic
encephalopathy. Neuroradiology. 1990;32:319–321.
71. Singhal AB, Topcuoglu MA, Koroshetz WJ. Diffusion MRI in three
types of anoxic encephalopathy. J Neurol Sci. 2002;196:37–40.
72. Takahashi S, Higano S, Ishii K, et al. Hypoxic brain damage: corti-
cal laminar necrosis and delayed changes in white matter at
sequential MR imaging. Radiology. 1993;189:449–456.
73. Chalela JA, Wolf RL, Maldjian JA, et al. MRI identification of early
white matter injury in anoxic–ischemic encephalopathy.
Neurology. 2001;56:481–485.
74. Rudolf J. Positron emission tomography in diagnosis and progno-
sis of postanoxic cerebral dysfunctions. Fortschr Neurol Psychiatr.
2000;68:344–351.
75. Schaafsma A, de Jong BM, Bams JL, et al. Cerebral perfusion and
metabolism in resuscitated patients with severe post-hypoxic
encephalopathy. J Neurol Sci. 2003;210:23–30.
76. Gibbs FA, Gibbs EL, Lennox WG. Effect on the electroencephalo-
gram of certain drugs which influence nervous activity. Arch Intern
Med. 1937;60:154–166.
77. Daly DD. Use of EEG for diagnosis and evaluation of epileptic
seizures and nonepileptic episodic disorders. In: Klass DD, Daly
DD, eds. Current Practice of Clinical Electroencephalography. New
York, NY: Raven Press; 1979:221–268.
78. Harner RN, Katz RI. Electroencephalography in metabolic coma.
In: Harner R, Naquet R, eds. Altered States of Consciousness,
Coma, Cerebral Death. Handbook of Electroencephalography and
Clinical Neurophysiology. Vol 12. Amsterdam: Elsevier; 1975:
47–62.
79. Gloor P. Generalized and widespread bilateral paroxysmal activi-
ties. In: Cobb W, ed. EEG Interpretation in Clinical Medicine.
Handbook of Electroencephalography and Clinical Neurology. Vol
11B. Amsterdam: Elsevier; 1976:52–87.
80. Arfel G, Fischgold H. EEG signs in tumors of the brain. In:
Electroencephalography and Cerebral Tumors. Electroencephalogr
Clin Neurophysiol Suppl. 1961;19:36–50.
81. Hess R. Die bioelektrischen Zeichen der cerebralen Massenver -
schiebung bei Hirntumoren. Schweiz Med Wochenschr. 1962;92:
1537–1542.
82. Schaul N, Gloor P, Gotman J. The EEG in deep midline lesions.
Neurology (New York). 1981;31:157–167.
83. Arfel G. Introduction to clinical and EEG studies in coma. In:
Harner R, Naquet R, eds. Altered States of Consciousness, Coma,
454 Part III ■ Clinical EEG: General
Cerebral Death. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 12. Amsterdam: Elsevier; 1975:5–23.
84. Fischgold H, Mathis P. Obnubilations, comas et stupeurs: ƒtudes
Electroencephalographique. Electroencephalogr Clin Neurophysiol.
(suppl 11). Paris: Masson; 1959.
85. Kubicki S, Rieger H, Busse G, et al. Elektroenzephalographische
Befunde bei schweren Schlafmittelvergiftungen. Z EEG-EMG.
1970;1:80–93.
86. Schwartz MS, Scott DF. Pathological stimulus-related slow
wave arousal response in the EEG. Acta Neurol Scand. 1978;57:
300–304.
87. Chatrian GE, Shaw CM, Leffman H. The significance of periodic
lateralized epileptiform discharges in EEG: an electrographic,
clinical and pathological study. Electroencephalogr Clin
Neurophysiol. 1964;17:177–193.
88. de la Paz D, Brenner RP. Bilateral independent periodic lateral-
ized epileptiform discharges. Clinical significance. Arch Neurol.
1981;38(11):713–715.
89. Kuroiwa Y, Celesia GG. Clinical significance of periodic EEG pat-
terns. Arch Neurol. 1980;37:15–20.
90. Schwartz MS, Prior PF, Scott DF. The occurrence and evolution
in the EEG of a lateralized periodic phenomenon. Brain. 1973;
96:613–622.
91. Husain AM, Mebust KA, Radtke RA. Generalized periodic
epileptiform discharges: etiologies, relationship to status epilep-
ticus, and prognosis. J Neurophysiol Clin. 1999;16(1):51–58.
92. Lowenstein DH, Aminoff MJ. Clinical and EEG features of status
epilepticus in comatose patients. Neurology. 1992;42:100–104.
93. Suter C. Clinical advances in the evaluation of deep coma. MCV
Q. 1974;10:152–162.
94. Terzano MG, Parrino L, Mazzuchi A, et al. Confusional states with
periodic lateralized epileptiform discharges (PLEDs): a peculiar
epileptic syndrome in the elderly. Epilepsia. 1986;27:446–457.
95. Handforth A, Cheng JT, Mandelkern MA, et al. Markedly
increased mesiotemporal lobe metabolism in a case with PLEDs:
further evidence that PLEDs are a manifestation of partial status
epilepticus. Epilepsia. 1994;35:876–881.
96. Fisch BJ, Klass DW. The diagnostic specificity of triphasic wave
patterns. Electroencephalogr Clin Neurophysiol. 1988;70(1):1–8.
97. Bickford RG, Butt HR. Hepatic coma: the electroencephalo-
graphic pattern. J Clin Invest. 1955;34:790–799.
98. Reiher J. The electroencephalogram in the investigation of meta-
bolic comas. Electroencephalogr Clin Neurophysiol. 1970;28:104.
99. Calham CL, Ettinger MG. Unusual EEG in coma after cardiac
arrest. Electroencephalogr Clin Neurophysiol. 1966;21:385–388.
100. Prior PF. The EEG in Acute Cerebral Anoxia. Amsterdam:
Excerpta Medica; 1973.
101. Silverman D. The electroencephalogram in anoxic coma. In:
Harner R, Naquet R, eds. Altered States of Consciousness. Coma,
Cerebral Death/Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 12. Amsterdam: Elsevier; 1975: 81–94.
102. Vulliemoz S, Iwanowski P, Landis T, et al. Levetiracetam accumu-
lation in renal failure causing myoclonic encephalopathy with
triphasic waves. Seizure. 2009;18:376–378.
103. Neufeld MY. Periodic triphasic waves in levodopa-induced
encephalopathy. Neurology. 1992;42(2):444–446.
104. Madison D, Niedermeyer E. Epileptic seizures resulting from acute
cerebral anoxia. J Neurol Neurosurg Psychiatry. 1970;33:381–386.
105. Towsend JB, Drury J. Triphasic waves in coma from brainstem
infarction. Eur Neurol. 1991;31:47–49.
106. Young GB, Bolton CF, Archibald YM, et al. The electroencephalo-
gram in sepsis-associated encephalopathy. J Clin Neurophysiol.
1992;9:145–152.
107. Miller JW, Klass DW, Mokri B, et al. Triphasic waves in cerebral
carcinomatosis. Arch Neurol. 1986;43:1191–1193.
108. Blatt I, Brenner RP. Triphasic waves in a psychiatric population: a
retrospective study. J Clin Neurophysiol. 1996;13:324–329.
109. Bauer G, Aichner F, Mayr U. Status atypischer Absencen in
Jugend- und Erwachsenenalter. Nervenarzt. 1983;54:100–105.
110. Bauer G, Prugger M, Bohr K, et al. Relationship between the sup-
pression burst pattern and rhythmical activities in the alpha and
theta bands of EEG of comatose patients. Electroencephalogr Clin
Neurophysiol. 1989;76:72P.
111. Zaret BS. Prognostic and neurophysiological implications of
concurrent burst suppression and alpha patterns in the EEG of
postanoxic coma. Electroencephalogr Clin Neurophysiol. 1985;61:
199–209.
112. Niedermeyer E, Sherman DL, Geocadin RJ, et al. The burst-
suppression electroencephalogram. Clin Electroencephalogr.
1999;30(3):99–105.
113. Butenuth J, Kubicki S. Über die prognostische Bedeutung bes-
timmter Formen der Myoklonien und korrespondierender EEG-
Muster nach Hypoxien. Z EEG-EMG. 1971;2:78–83.
114. Pampiglione G, Harden A. Resuscitation after cardio-vascular
arrest. Prognostic evaluation and early electroencephalographic
findings. Lancet. 1968;1:1261–1264.
115. Chatrian GE, White LE, Shaw C-M. EEG pattern resembling
wakefulness in unresponsive decerebrate state following trau-
matic brain stem infarct. Electroencephalogr Clin Neurophysiol.
1964;16:285–289.
116. Loeb C, Poggio G. Electroencephalograms in a case with ponto-
mesencephalic haemorrhage. Electroencephalogr Clin Neuro-
physiol. 1953;5:295–296.
117. Lundervold A, Hange T, Lšken AC. Unusual EEG in unconscious
patient with brain stem atrophy. Electroencephalogr Clin
Neurophysiol. 1956;8:665–670.
118. Rohmer R, Kurtz D, Kiffer A. étude critique de l’activité E.E.G.
dan les syndromes vascularies du tronc cérébral. Rev Neurol. 1965;
113:278–284.
119. Westmoreland BF, Klass DW, Sharbrough FM, et al. Alpha-coma.
Electroencephalographic, pathologic and etiologic correlations.
Arch Neurol. 1975;32:713–718.
120. Binnie CD, Prior PF, Lloyd DSL, et al. Electroencephalographic
prediction of fatal anoxic brain damage after resuscitation from
cardiac arrest. Br Med J. 1970;4:265–268.
121. Chokroverty S. “Alpha-like” rhythms in electroencephalograms
in coma after cardiac arrest. Neurology (Minneapolis). 1975;25:
655–663.
122. Grindal AB, Suter C. “Alpha-pattern coma” in high voltage electri-
cal injury. Electroencephalogr Clin Neurophysiol. 1975;38:521–526.
123. Sharbrough FW, Westmoreland BF, Reagan TJ, et al. The signifi-
cance of a transitional monorhythmic EEG pattern in patients
after cardiopulmonary arrest. Neurology (Minneapolis). 1975;27:
384–385.
124. Vignaendra VMB, Wilkus RJ, Copass MK, et al. Electro -
encephalographic rhythms of alpha frequency in comatose
patients after cardiopulmonary arrest. Neurology (Minneapolis).
1974;24:582–588.
125. Berkhoff M, Donati F, Bassetti C. Postanoxic alpha (theta) coma:
a reappraisal of its prognostic significance. Clin Neurophysiol.
2000;111:297–304.
126. Nogueira de Melo A, Krauss GL, Niedermeyer E. Spindle coma:
observations and thoughts. Clin Electroencephalogr. 1990;21:
151–161.
127. Iragui VJ, McCutchen ChB. Physiologic and prognostic significance
of “alpha coma”. J Neurol Neurosurg Psychiatry. 1983;46:632–638.

128. Kaplan PW, Genoud D, Ho TW, et al. Etiology, neurologic corre-
lations, and prognosis in alpha coma. Clin Neurophysiol. 1999;
110(2):205–213.
129. Tomassen W, Kamphuisen HAC. Alpha coma. J Neurol Sci.
1986;76:1–11.
130. Young GB, Blume WT, Campbell VM, et al. Alpha, theta and
alpha–theta coma: a clinical outcome study utilizing serial
records. Electroencephalogr Clin Neurophysiol. 1994;91:93–99.
131. Chatrian GE, White LE, Daly D. Electroencephalographic pattern
resembling those of sleep in certain comatose states after injuries to
the head. Electroencephalogr Clin Neurophysiol. 1963;15:272–280.
132. Seet RC, Lim EC, Wilder-Smith EP. Spindle coma from acute
midbrain infarction. Neurology. 2005;64(12):2159–2160.
133. Ganji SS, Henry R, Furlow J. Diffuse theta activity and spindle-like
bursts during coma after cardiac arrest. Clin Electroencephalogr.
1996;27(2):89–94.
134. Dadmehr N, Pakalnis A, Drake ME Jr. Spindle coma in viral
encephalitis. Clin Electroencephalogr. 1987;18(1):34–37.
135. Mouradian MD, Penovich PE. Spindle coma in benzodiazepine
toxicity: case report. Clin Electroencephalogr. 1985;16(4): 213–218.
136. Hess R. Sleep and sleep disturbances in the electroencephalo-
gram. Prog Brain Res. 1965;18:127–139.
137. Radermecker J, Macken J. Aspects électroencéphalographiques et
cliniques de la leucoencéphalite sclérosante subaigue. Rev Neurol.
1951;83:341.
138. Gloor P, Kalabay O, Giard N. The electroencephalogram in dif-
fuse encephalopathies: electroencephalographic correlates of
grey and white matter lesions. Brain. 1968;91:779–802.
139. Lee RG, Blair RDG. Evolution of EEG and visual evoked response
changes in Jakob–Creutzfeldt disease. Electroencephalogr Clin
Neurophysiol. 1975;35:133–142.
140. Illis LS, Taylor FM. The electroencephalogram in herpes-simplex
encephalitis. Lancet. 1972;1:718–721.
141. Kugler J, Martin JJ, Radermecker FJ, et al. Periodische Komplexe
im EEG bei nekrotisierender Herpes-Enzephalitis. Z EEG-EMG.
1976;7:63–71.
142. Kaplan PW. Electrophysiological prognostication and brain
injury from cardiac arrest. Semin Neurol. 2006;26:403–412.
143. Chang Chui H, Damasio AR. Progressive dialysis encephalopathy
(“dialysis dementia”). J Neurol. 1980;222:145–157.
144. Garret PJ, Mulcahy D, Carmody M, et al. Aluminium encephalopa-
thy: clinical and immunological features. Q J Med. 1988;69:775–783.
145. Bauer G. Der Wert von EEG-Kontrollen mšglichst bald nach
einem epileptischen Anfall. Z EEG-EMG. 1975;6:125–130.
146. Bauer G. Psychische Veränderungen bei kontinuierlichen epilep-
tischen Entladungen. Schweiz Arch Neurol Neurochir Psychiatr.
1975;116:241–255.
147. Brett EM. Minor epileptic status. J Neurol Sci. 1966;3:53–75.
148. Niedermeyer E, Khalifeh R. Petit mal status (“spike wave
stupor”). Epilepsia. 1965;6:250–262.
149. Loeb C. Correlative EEG and clinico-pathological studies of patients
in coma. In: Harner R, Naquet R, eds. Altered States of Consciousness,
Coma, Cerebral Death. Vol 11. Amsterdam: Elsevier; 1975:24–36.
150. Romano J, Engel GL. Delirium. I: electroencephalographic data.
Arch Neurol Psychiatry. 1944;52:290–295.
151. Vigouroux R, Naquet R, Baurand C, et al. Evolution électroradio-
clinique de comas graves prolongés post traumatiques. Rev
Neurol. 1964;110:72–80.
152. Bergamini L, Bergamasco B, Mombelli A, et al. Autocorrelazione
in sogetti in coma. Rev Neurol. 1966;36:233–236.
153. Rumpl E, Lorenzi E, Hackl JM, et al. The EEG at different stages
of acute secondary traumatic midbrain and bulbar brain syn-
dromes. Electroencephalogr Clin Neurophysiol. 1979;46:487–497.
Chapter 23 ■ Anoxia, Coma, and Brain Death 455
154. Silverman D. Retrospective study of the EEG in coma.
Electroencephalogr Clin Neurophysiol. 1963;15:486–503.
155. Synek VM, Synek BJL. Theta pattern coma, a variant of alpha
pattern coma. Clin Electroencephalogr. 1984;15:116–121.
156. Naquet R, Fernandez-Guardiola A. Effects of various types of
anoxia on spontaneous and evoked cerebral activity in the cat. In:
Gastaut H, Meyer JS, eds. Cerebral Anoxia and the Electroencephalo-
gram. Springfield, IL: Charles C. Thomas; 1959:72–88.
157. Kuroiwa Y, Celesia GG, Chung HD. Periodic EEG discharges of
the cerebral cortex in anoxic encephalopathy: a necropsy case
report. J Neurol Neurosurg Psychiatry. 1982;45:740–742.
158. Simon RP, Aminoff MJ. Electroencephalographic status epilepti-
cus in fatal anoxic coma. Ann Neurol. 1986;20:351–355.
159. Van Cott AC, Blatt I, Brenner RP. Stimulus-sensitive seizures in
postanoxic coma. Epilepsia. 1996;37:868–874.
160. Hockaday JM, Potts F, Epstein E, et al. EEG changes in acute cere-
bral anoxia from cardiac or respiratory arrest. Electroencephalogr
Clin Neurophysiol. 1965;18:575.
161. McKeown MJ, Young GB. Comparison between the alpha pattern
in normal subjects and in alpha pattern coma. J Clin Neurophysiol.
1997;14:414–418.
162. Celesia GG. EEG and coma: is there a prognostic role for EEG?
Editorial. Clin Neurophysiol. 1999;110:203–204.
163. Hume AI, Cant BR. Central somatosensory conduction after
head injury. Ann Neurol. 1981;10:411–419.
164. Hutchinson DO, Frith RW, Shaw NA, et al. A comparison
between electroencephalography and somatosensory evoked
potentials for outcome predictions following severe head injury.
Electroencephalogr Clin Neurophysiol. 1991;78:228–233.
165. Rumpl E, Prugger M, Gerstenbrand F, et al. Central somatosen-
sory evoked potentials in posttraumatic coma. Electroencephalogr
Clin Neurophysiol. 1983;56:583–596.
166. Cremer OL, Moons KGM, Bouman EAC, et al. Long-term propo-
fol infusion and cardiac failure in adult head-injured patients.
Lancet. 2001;357:117–118.
167. Schwab S, Spranger M, Schwarz S, et al. Barbiturate coma in
severe hemispheric stroke: useful or obsolete? Neurology. 1997;48:
1608–1613.
168. Brain Resuscitation Clinical Trial I Study Group. Randomized
clinical study of thiopental loading in comatose survivors of car-
diac arrest. N Engl J Med. 1986;314:397–403.
169. Dearden NM. Ischaemic brain. Lancet. 1985;2:255–259.
170. Evans BM, Bartlett JR. Prediction of outcome in severe head
injury based on recognition of sleep related activity in the poly-
graphic electroencephalogram. J Neurol Neurosurg Psychiatry. 1995;
59:17–25.
171. Gutling E, Gonser A, Imhof H-G, et al. EEG reactivity in the
prognosis of severe head injury. Neurology. 1995;45:915–918.
172. Luecking CH. Sleep-like patterns and abnormal arousal reactions
in brain stem lesions. Electroencephalogr Clin Neurophysiol. 1970;
28:214 [abstract].
173. Bergamasco B, Bergamini L, Doriguzzi T, et al. EEG sleep pat-
terns as a prognostic criterion in post-traumatic coma. Electroen-
cephalogr Clin Neurophysiol. 1968;24:374–377.
174. Rumpl E, Prugger M, Bauer G, et al. Incidence and prognostic
value of spindles in posttraumatic coma. Electroencephalogr Clin
Neurophysiol. 1983;56:420–429.
175. Valente M, Placidi F, Oliveira AJ, et al. Sleep organization pattern
as a prognostic marker at the subacute stage of post-traumatic
coma. Clin Neurophysiol. 2002;113:1798–1805.
176. Kaplan PW, Genoud D, Ho TW, et al. Clinical correlates and
prognosis in early spindle coma. Clin Neurophysiol. 2000;111:
584–590.
456 Part III ■ Clinical EEG: General
177. Hulihan JF, Syna DR. Electroencephalographic sleep patterns in
post-anoxic stupor and coma. Neurology. 1994;44:758–760.
178. Dusser A, Nevelet Y, Devictor D, et al. Short- and long-term
prognostic value of electroencephalogram in children with
severe head injury. Electroencephalogr Clin Neurophysiol. 1989;
73:85–93.
179. Bickford RG, Dawson B, Takeshita H. EEG evidence of neuro-
logic death. Electroencephalogr Clin Neurophysiol. 1965;18:513.
180. Pampiglione G. EEG studies after cardio-pulmonary resuscita-
tion. Proc R Soc Med. 1962;55:653.
181. Pampiglione G. Resuscitation after cardiac arrest. Electroen-
cephalogr Clin Neurophysiol. 1962;14:294.
182. MacGillivray BB. The EEG in liver disease. In: Daly DD, ed.
Metabolic, Endocrine and Toxic Diseases. Handbook of Electroen-
cephalography and Clinical Neurophysiology. Vol 15. Amsterdam:
Elsevier; 1976:26–50.
183. Bell JA, Hodgson HJF. Coma after cardiac arrest. Brain. 1974;
97:361–372.
184. Thomassen A, Wernberg M. Prevalence and prognostic signifi-
cance of coma after cardiac arrest outside intensive care and
coronary units. Acta Anaesth. 1979;23:143–148.
185. Bates D, Caronna JJ, Cartlidge NEF, et al. A prospective study of
non-traumatic coma: methods and results in 310 patients. Ann
Neurol. 1977;2:211–220.
186. Abramson NS, Meisel A, Safar P, Deferred consent. A new
approach for resuscitation research on comatose patients. JAMA.
1986;255:2466–2471.
187. Edgren E, Hedstrand U, Kelsey S, et al., and BRCT 1 Study Group.
Assessment of neurological prognosis in comatose survivors of
cardiac arrest. Lancet. 1994;343:1055–1059.
188. Levy DE, Bates D, Caronna JJ, et al. Prognosis in nontraumatic
coma. Ann Intern Med. 1981;94:293–301.
189. Willoughby JO, Leach BG. Relation of neurological findings after
cardiac arrest to outcome. Br Med J. 1974;3:437–439.
190. Levy DE, Caronna JJ, Singer BH, et al. Predicting outcome from
hypoxic ischemic coma. JAMA. 1985;253:1420–1426.
191. Earnest MP, Breckinridge JC, Yarnell PR, et al. Quality of survival
after out-of-hospital cardiac arrest; predictive value of early neu-
rologic evaluation. Neurology. 1979;29:56–60.
192. Longstreth WT Jr, Diehr P, Inui TS. Prediction of awakening
after out-of-hospital cardiac arrest. N Engl J Med. 1983;308:
1378–1382.
193. Snyder BD, Loewenson RB, Gumnit RJ, et al. Neurologic progno-
sis after cardiac arrest: II. Level of consciousness. Neurology. 1980;
30:52–58.
194. Bassetti C, Bomio F, Mathis H, et al. Early prognosis after cardiac
arrest: a prospective clinical, electrophysiological, and biochemical
study of 60 patients. J Neurol Neurosurg Psychiatry. 1996;61:
610–615.
195. Murphy DJ, Murray AM. Outcome of cardiopulmonary resusci-
tation in the elderly. Ann Intern Med. 1989;111:199–203.
196. Zandbergen EGJ, de Haan RJ, Stoutenbeek CP, et al. Systematic
review of early prediction of poor outcome in anoxic-ischaemic
coma. Lancet. 1998;352:1808–1812.
197. Guérit JM, Amantini A, Amodio P, et. al. Consensus on the use of
neurophysiological tests in the intensive care unit (ICU): elec-
troencephalogram (EEG), evoked potentials (EP), and elec-
troneuromyography (ENMG). Neurophysiol Clin. 2009;39:71–83.
198. Guérit JM. Electroencephalography: the worst traditionally rec-
ommended tool for brain death confirmation. Intensive Care
Med. 2007;33:9–10.
199. Young GB. Clinical practice. Neurologic prognosis after cardiac
arrest. N Engl J Med. 2009;361:605–611.
200. Chen R, Bolton CF, Young B. Prediction of outcome in patients
with anoxic coma: a clinical and electrophysiological study. Crit
Care Med. 1996;24:672–678.
201. Zandbergen EG, Hijdra A, Koelman JH, et. al. PROPAC Study
Group. Prediction of poor outcome within the first 3 days of
postanoxic coma. Neurology. 2006;66:62–68.
202. Jia X, Koenig MA, Shin HC, et. al. Quantitative EEG and neuro-
logical recovery with therapeutic hypothermia after asphyxial
cardiac arrest in rats. Brain Res. 2006;1111:166–175.
203. Rothstein TL, Thomas EM, Sumi SM. Predicting outcome in
hypoxic-ischemic coma. A prospective clinical and electrophysio-
logic study. Electroencephalogr Clin Neurophysiol. 1991;79:101–107.
204. Pressler RM, Boylan GB, Morton M, et al. Early serial EEG in
hypoxic ischaemic encephalopathy. Clin Neurophysiol. 2001;112:
31–37.
205. Synek VM. EEG abnormality grades and subdivisions of prog-
nostic importance in traumatic and anoxic coma in adults. Clin
Electroencephalogr. 1988;19:160–166.
206. Synek VM. Value of a revised EEG coma scale for prognosis after
cerebral anoxia and diffuse head injury. Clin Electroencephalogr.
1990;21:25–30.

The EEG in Patients with Migraine and
Other Forms of Headache
ERNST NIEDERMEYER AND DONALD L. SCHOMER
P
atients with headaches are usually referred to the elec-
troencephalography (EEG) laboratory to rule out under-
lying cerebral pathology rather than for a clarification of
the type of headache. This type of referral has become less fre-
quent with the greater availability of modern neuroimaging.
Headache is one of the most common complaints. As a
symptom, it may herald a wide variety of infectious, neoplastic,
and vascular intracranial lesions, but it also may be a sign of var-
ious dysfunctions impinging on neural, vascular, and muscular
structures. It may arise from the vicinity of the cranial cavity or
even from distant structures. Metabolic, toxic, and hormonal
disturbances are further causes. In other words, headache is a
challenge for the diagnostic acumen of neurologists and other
specialists. It has been stated that about 20% of the U.S. popula-
tion complains of headache, and about half of them receive
some form of symptomatic medical treatment (1). Table 24.1
shows a classification of the types of headache. New criteria of
classification have been proposed by Silberstein et al. (3).
MIGRAINE (CLASSICAL AND
COMPLICATED FORMS)
General Considerations and Clinical Features
Migraine has been known to humanity for ages; a Sumerian
poem written 5000 years ago gives an account of this disorder.
In spite of a remarkable upsurge of research interest in this
CHAPTER
24
field, migraine has remained a poorly understood disorder. A
plethora of clinical data is found in the work of Sacks (4).
The clinical symptomatology of the migrainous attack is
well known. In the classical form, visual symptoms herald the
attack; there are scintillating scotoma with teichoscopy and
other forms of visual field cuts. Within a short time (about 5 to
20 minutes), this stage is supplanted by headache, which is
mostly unilateral, with shifting lateralization from attack to
attack in most cases. This is accompanied by severe nausea,
vomiting, irritability, and photophobia. This stage lasts for
hours or a full day. Wolff (5) has ascribed the initiating visual
symptoms to intracerebral local vasoconstriction and the ensu-
ing phase of headache and nausea to an abnormal degree of
extracranial vasodilation, which can easily be palpated along
the temporal artery on the painful side.
A genetic predisposition is present or even pronounced, but
the mode of genetic transmission is not fully understood. The
attacks tend to start in adolescence; in childhood, attacks of
abdominal pain may be precursors of migraine. It has been
thought that a fall of the plasma serotonin level (6) plays a cru-
cial role in the physiopathogenesis of the migrainous attack, but
this concept has not been generally accepted. Allergic (anti-
gen–antibody) reactions, free fatty acids, and prostaglandin E
are also believed to be involved in migraine (7). Activation of 5-
hydroxytryptamine receptors has been stressed by Fozard (8).
According to Moskowitz (9), migraine is caused by a distur-
bance of the “trigeminovascular system” (connections between

Tabl e 2 4 . 1

Classification of Headache

Vascular Headache Muscle Contraction Traction and Inflammatory

(Psychogenic) Headache Headache
Migraine Cervical osteoarthritis Mass lesions (tumors, edema, hematomas, cerebral hemorrhage)
1. Classic
2. Common
3. Hemiplegic Complicated migraine
4. Ophthalmoplegic
Cluster (histamine) Chronic myositis Diseases of the eye, ear, nose, throat, teeth
Toxic vascular Depressive equivalents and Infection
conversion reactions
Hypertensive Arteritis, phlebitis; cranial neuralgias; occlusive vascular disease

Modified from Dalessio, DJ. Classification and mechanism of migraine. Headache. 1979;19:114–120.

457
458 Part III ■ Clinical EEG: General
trigeminal ganglia and cerebral blood vessels) involving the
neurotransmitter peptide, substance P. Experimentally, “neu-
roinflammation” was produced in animals by electrical stimu-
lation of the trigeminal ganglion causing the release of sensory
neuropeptides from nerve terminals (the model of “neurogenic
inflammation” (10)). Biggs and Johnson (11) have placed spe-
cial emphasis on the adrenergic system and its role in migraine
pathogenesis. A unified neurogenic concept of migraine has
been proposed by Diamond and Dalessio (12).
Not all cases of migraine are due to an inherited dysfunction;
neuropathologic processes such as arteriovenous malformations
of neurosyphilis are known as cases of “symptomatic migraine.”
This differential diagnosis may occasionally become difficult,
because migraine attacks are capable of proceeding to a state of
ischemic infarction (13,14) and of producing regional computed
tomography (CT) scan changes (15). In regional cerebral blood
flow studies using xenon-133 intra-arterially, reduced blood flow
was demonstrable during migraine attacks starting posteriorly
and very slowly spreading to the rolandic region (16). According
to Hansen et al. (17) and Olesen (18), spreading depression is
considered a useful model of migraine aura and presumably also
for the subsequent headache. This concept was reemphasized by
Olesen (19), especially on the basis of the positron emission
tomography (PET) scan observation (oxygen-15-labeled water)
of Woods et al. (20). There is good evidence of hypoperfusion
within the occipital lobe (20), which can just as well (if not bet-
ter) be used as supportive of the vascular concept.
Special involvement of the central visual system is a very
common feature of classical migraine with visual initiation.
Huang et al. (21) have shown with the use of functional mag-
netic resonance imaging (fMRI) the excessive responses to
visual stimuli and a particular sensitivity to a pattern of regu-
larly spaced parallel lines of stripes (see also Lashley (22)).
Whoever reads the original experimental techniques used in
the production of spreading depression (23,24) will have nag-
ging doubts concerning the appropriateness of the spreading-
depression model for a neurogenic migraine concept. The mode
of elicitation implies all sorts of mechanical and chemical
trauma to the brain tissue; electrocorticographic recording (24)
shows flattening of the record followed by several recurrent
prominent spikes (against a flat background) and another phase
of flattening before the baseline character of the record is
restored. It is difficult to imagine that similar electrical processes
would occur in migrainous human beings. The primordial
nature of vasomotor changes according to H. G. Wolff’s (5) orig-
inal theory seems to be a lot more plausible.
On the basis of data derived from animal experiments
(Wistar rats), Ebersberger et al. (25) doubt that spreading
depression initiates migraine.
Electroencephalographic Findings
The literature in this field is very confusing because almost
equal numbers of reports stress the predominance of normal
and abnormal tracings. Relatively few records have been
obtained during the attacks; these data are discussed later. Daly
and Markand (26) have pointed out that previous studies of
migraine and EEG were frequently flawed by sampling prob-
lems and heterogeneous populations of migrainous persons.
The contrast between various reports on the EEG in migraine
in the interval between attacks is due to (i) composition of mate-
rial (adults vs. children, inclusion or exclusion of hemiplegic
cases), (ii) different criteria for normality and abnormality in
the investigators’ EEG interpretation, and (iii) difficulties in the
delineation of migraine as a nosologic entity (inclusion or strict
exclusion of cluster headaches or symptomatic forms of
migraine with cerebral pathology). Keeping all this in mind, it is
still difficult to understand the disparity of the reports.
The predominance of normal-interval EEG records was
stressed by Ulett et al. (27), Jung (28), Becher (29), Krischek
(30), Wissfeld and Neu (31), Bille (32), and Gibbs and Gibbs
(33). The work of other authors places the emphasis on a variety
of abnormalities. Heyck (34,35) found mainly “hypersynchro-
nous bursts” and occasional focal slowing. Weil (36,37) noted
pronounced delta responses to hyperventilation. Various types
of abnormalities were noted by Dow and Whitty (38) and Selby
and Lance (39). A high incidence of abnormal EEG records was
also emphasized by Barolin (40), Gschwend (41), and Pithova
(42). Almost equal numbers of normal and abnormal records
(with about 45% abnormal tracings) were reported in the exten-
sive work of Smyth and Winter (43). The reported abnormali-
ties, however, were predominantly mild to moderate, with some
bursts, slowing, or sharp transients. With the use of computer
frequency analysis, Jonkman and Lelieveld (44) demonstrated
abnormal interval EEG findings in 55% of migrainous patients.
According to Drake et al. (45), the EEG of patients with migraine
does not differ significantly from the EEG of normal individu-
als. This is essentially congruent with my personal views.
Intermittent photic stimulation often shows an occipital
driving response extending into the range above 20 flashes/sec
(“H response” after Golla and Winter (46)); according to Smyth
and Winter (43), this is almost specific for migraine. This has
been substantiated by Slater (47). Personal observations essen-
tially support this view. Further substantiation of these findings
has been provided by Simon et al. (48) with the use of spectral
analysis during photic stimulation (Fig. 24.1).
EEG findings in the migraine attack range from normal to
mildly abnormal (alpha depression) in the initiating ophthalmic
phase; even severely abnormal findings have been reported in
special cases (49–51). Based on his large material, Heyck (34)
found normal tracings in the ophthalmic vasoconstrictive as
well as in the headache-nausea phase. Schoenen et al. (52) found
reduced alpha activity over one occipital region in 19 out of 22
patients recorded during an attack of common migraine. In the
light of these observations, the statement that “EEGs have
almost always been normal in migraineurs during attacks” (53)
might be slightly exaggerated, but EEG abnormalities should be
viewed as exceptions.
A neuronal dysfunction as the cause of migraine was
assumed by Soysal et al. (54) on the basis of significantly pro-
longed P100 latencies of visual evoked potentials in the interval
between attacks. On the other hand, EEG abnormalities were
observed in only 4 of the 13 patients.
 Chapter 24 ■ The EEG in Patients with Migraine and Other Forms of Headache
Figure 24.1 A: A 31-year-old woman with a history of classical migraine
(experiences flashing lights and also some left-sided numbness). EEG
obtained in interval. Note good occipital photic driving response to a flash
rate of 22/ sec (“H response”). The right frontal spiky discharge is artifac-
tual. B: Same patient. Very good occipital lambda activation presentation
of pattern vision test tables.
When the migraine attack is complicated by mild hemiparetic
or dysphasic deficits (migraine accompagnŽe), the EEG may
remain normal (55). In cases with pronounced hemiplegia and
aphasia, there is good evidence of delta and theta activity over the
affected hemisphere (34,56–59). The delta activity over the
affected hemisphere may be very impressive (see cases of Isler
(60)). The neurologic deficit subsides within days (sometimes
within weeks), and the focal or lateralized slowing my also linger
on for some period of time (Fig. 24.2). Cases of familial hemi-
plegic migraine have been reported by Whitty (14), Rosenbaum
(59), Bradshaw and Parsons (56), and Müller and Müller (61). It
459
has been pointed out that in familial hemiplegia the lateralization
of the affected hemisphere remains unchanged in every attack
and is the same in all involved family members. This is not con-
gruent with a personal observation. EEG studies in familial hemi-
plegia show a varying degree of slowing over the brain’s affected
side.
The contingent negative variation (CNV) has been used for
the differential diagnosis of headaches (62). Ahmed (63) clearly
demonstrated an enhanced CNV in patients with classic and
common migraine; their CNV was moderately larger than in
patients with tension headaches and much larger than in nor-
mal controls. These statements pertain to the pain-free interval
between migraine attacks. The CNV changes are believed to
reflect catecholamine hyperactivity. According to CNV
research, this slow potential measures expectancy, attention,
preparation, and motivation. When one considers that, accord-
ing to Ahmed (63), the enlargement of the CNV would persist
after successful antimigraine treatment, then one wonders if
enhanced motivation and a powerful desire to perform tasks
perfectly are personality traits of migrainous persons.
ATYPICAL FORMS OF MIGRAINE
Abdominal manifestations of migraine are relatively common in
children. Headache is usually absent in these attacks, whereas
abdominal pain is in the foreground. Such attacks may last for
hours. Temporary agitation and obnubilation during the attack
have been noted by Lérique-Koechlin and Mises (64).
In childhood migraine, normal EEG records are the rule
(89%), but “benign” focal spikes, mostly rolandic, were found in
9% (65). Kellaway et al. (66) stressed the high incidence of 14
and 6/sec positive spikes in the sleep records of these children.
During the attack of abdominal pain, the EEG is unremarkable
(personal observations). Gibbs and Gibbs (33) separate abdom-
inal pain attack (with frequent 14 and 6/sec positive spike dis-
charges in the interval) from migraine with normal-interval
tracings. Lérique-Koechlin and Mises (64) reported a very high
incidence of paroxysmal EEG changes in children with a history
of abdominal pain attacks. Moore (67) introduced the term
abdominal epilepsy for such cases of acute abdominal pain, but
this has not been widely accepted. This term is rather mislead-
ing, and its use should be discouraged. In migrainous children,
temporal lobe-type seizures have been described (68). These
children experienced strange sensations or misperceptions; the
EEG in most cases showed sharp activity over the temporal
region. “Benign paroxysmal vertigo of childhood” has been
found to be a migrainous equivalent (69).
Basilar migraine represents a syndrome described by
Bickerstaff (70), consisting of a sudden transient blurring of
vision or blindness, vertigo, gait ataxia, dysarthria, acroparesthe-
sias, and pulsatile occipital headache with vomiting. Even syn -
copal states and loss of consciousness may occur (71). Basilar
migraine manifestations are more common in older children,
adolescents, and young adults; females are more often affected. In
adulthood, these attacks may be supplanted by attacks of classical
or common migraine. According to Caplan (72), basilar migraine
460 Part III ■ Clinical EEG: General
Figure 24.2 A: A 13-year-old boy with migraine attacks. EEG recorded during a migraine attack with moderate right hemi-
paresis and global aphasia. Note marked delta activity over left occipital-posterotemporal region. B: EEG taken 5 days later.
Clinically normal, EEG normalized.
is not always benign. It affects both sexes and a wider age range;
it also may be associated with strokes.
The EEG literature is meager in this domain; a case recorded
by Slater (47) during a presumed attack showed diffuse 1.5 to
4/sec activity with subsequent normalization. This patient, how-
ever, was 46 years old; this age could cast some doubt on the diag-
nosis. Lapkin et al. (73) reported two cases (ages 12 and 10) with
diffuse and chiefly posterior slowing in the 1.5 to 2/sec and 3 to
4/sec ranges, respectively, during the attack. This activity van-
ished with serial recordings. Camfield et al. (74) observed four
adolescents (two girls and two boys) with presumed basilar
migraine. These cases were complicated by epileptic seizures
(mostly grand mal, but also focal motor). The EEG showed very
prominent spike activity and slowing accentuated over the poste-
rior regions. All of these patients had a strongly positive history
of classical or common migraine. Similar paroxysmal findings
over posterior regions were reported by Panayatopoulos (75).
Another form with predominant beta activity during the attack
was delineated by Parain and Samson-Dollfus (76). Gastaut (77)
has cast much doubt on the observations of Camfield et al. (74);
he feels that these patients were suffering from benign occipital
lobe epilepsy (see Chapter 27). Simple partial status epilepticus in
the occipital lobe can be misdiagnosed as migraine (78).
According to Ramelli et al. (79), the EEG of children (11
to 13 years) with basilar artery migraine showed diffuse
subdelta–delta activity during the attack and occipital
delta–theta activity hours afterward. The authors warn against
concepts of structural lesions (infarction, inflammation) and
presume a temporary disfunction.
Cernibori and Bouquet (80) observed children ages 2 to 14
years with episodes of loss of consciousness ranging from 1 hour
to several days; these episodes were attributed to basilar artery
migraine of childhood. Diffuse or lateralized slow activity was
noted in the EEG, which improved over the ensuing days.
Slightly abnormal interval EEG tracings were found in 4 of 12
children. Episodic comatose states have also been reported by
Ganji et al. (81).
Status Migrainosus
On the basis of personal observations, there is good reason to
presume that prolonged or constantly recurrent migraine
attacks are the result of severe mental–emotional stress. In this
condition, EEG findings are bland and noncontributory.
Whatever neurologic–organic basis may be demonstrated
in migraine, the involvement of psychogenic factors must not
be counted out. Evidence of a link between migraine and
 Tabl e 2 4 . 2

Differentiation of Migrainous and Certain Epileptic Conditions

Scintillating Visual Hallucination, Nausea

Condition Occurrence Sex Age Scotoma (Elementary, Figurative) Headache (Vomiting)
Common Very common Probably females Mostly after age 0 0
migraine preponderant 15 years
Classic Common Mostly males if without Mostly after age ? 0– 0–
migraine headache-nausea 15 years
Basilar artery Moderately Probably females only 15–35 years 0– 0– 0– 0–
migraine common
Migraine- Very rare ? (? mainly females) ? 15–40 years (immediately ? – (after 0– (after
triggered followed by grand mal) grand mal) grand mal)
epilepsy
Benign Rare No sex 5–40 years 0 – – (after
occipital preponderance (after seizure) seizure)
lobe epilepsy
Benign Very rare ? (? mainly females) 5–20 years for 0 (for epileptic), 0 0– 0–
rolandic epileptic, past 15 (for migrainous attacks) (after seizure), (after seizure),
epilepsy- for migrainous
migraine manifestations (in migraine) (in migraine)
Epileptic EEG (During
Syncope Seizures EEG (Interval) Episodes) Therapy Prognosis
0 reduction 0 Mostly normal Mostly normal Antimigrainous, Mostly very good (albeit
stress reduction no real cure)
0 (? ) 0 Mostly normal Mostly normal Antimigrainous, Mostly very good (albeit
stress reduction no real cure)
0– 0– Normal to slightly Normal to slightly Antimigrainous, Mostly very good
abnormal (nonspecific) abnormal (with stress reduction, (? spontaneous cure
changes due to may require in middle adulthood)
syncope or antiepileptics
convulsion)
0 Normal Typical EEG of a Probably preventive Unclear
(grand mal) tonic–clonic antimigrainous therapy,
convulsion antiepileptics ineffective
0 Abnormal with Abnormal with Antiepileptics Mostly good
(mostly visual, recurrent posterior bilateral posterior
also others) spike-waves or spikes spike-waves
0 (focal Normal or abnormal Probably abnormal Antiepileptics, in later Probably very good
motor) with rolandic spikes, during focal motor phase antimigrainous, (albeit no real cure
in adulthood normal attacks, normal stress reduction for migraine)

461
during migraine
462 Part III ■ Clinical EEG: General
neuroticism has been reported by Breslau et al. (82). Let us
assume that, while many migrainous persons are in full control
of such subthreshold mechanisms, others may lose their control
temporarily (under stress) or perhaps even permanently. Olesen’s
(19) concept of a “clean neurologic” migraine—uncontaminated
with psychogenic features—cannot be upheld in view of the
clinical facts.
RELATIONSHIPS BETWEEN MIGRAINE
AND EPILEPTIC SEIZURE DISORDERS
A combination of migraine and epileptic seizure disorder may
occur, but it is uncommon (40,83,84). A true link between both
disorders is highly debatable (85–87). There is some reason to
support the view of Camfield et al. (74) that, under certain cir-
cumstances, migraine can trigger an epileptic attack. This subject
has been reviewed by Hess (88), who presumes that headache
may be a secondary symptom in patients with epileptic seizure
disorder. There is no cogent need to establish a special form
called “dysrhythmic migraine” (36,37).
The entire subject of migraine and epilepsy was reinvesti-
gated in extensive work edited by Andermann and Lugaresi (89)
and in the overview of Andermann and Andermann (90).
An entity named “migraine-triggered epilepsy” has been
described by Niedermeyer (91–93). Typical grand mal seizures
occur after a very short visual initiation; migrainous headache
and nausea follow the seizure. The EEG is mostly normal in
the interval and the response to antiepileptic medication
leaves much to be desired. Postepileptic headaches are quite
common and there is also a possibility of “epilepsy-triggered
migraine” (94).
On the basis of very extensive data collected (1957 adult
probands with epilepsy), Ottman and Lipton (95) were unable
to support the theory of a shared genetic susceptibility for
migraine and epileptic disorders. Table 24.2 demonstrates the
differentiation of migrainous and certain epileptic conditions.
THE EEG IN OTHER TYPES OF HEADACHE
The EEG in nonmigrainous forms of headaches has been thus
far a barren field; one is unable to correlate any type of
headache with some type of EEG change during the attacks or
in the interval. Cluster headache is a well-defined entity, but has
no special EEG correlate. Hyperventilation-related headache
associated with EEG slowing has been reported by Sbrascini
and Bassi (96). Hypnic headache has been described by Raskin
(97) and related to REM sleep; the EEG is normal (98).
In general, patients with habitual headaches and no organic
disease may show EEG patterns that are believed to reveal some
degree of “neuronal hyperexcitability”; this is a vague term
without any precise scientific foundation, but with some merit
in the domain of medical practice. Rolandic mu rhythm is quite
often found in patients referred for headaches without demon-
strable organic cause. Other patients show mildly paroxysmal
flicker responses; still others show 14 and 6/sec positive spikes
or even categorical EEG abnormalities such as psychomotor
variant pattern and 6/sec spike waves. The possibility of vasomo-
tor headaches as a symptom of larval epilepsy has been discussed
by Heyck and Hess (99). A physiopathogenetic basis of such
dysfunctional states with headache is still obscure.
A rather nonspecific type of headache (“fullness of head,
pressure, heat, pounding”) was found to be an ictal epileptic
manifestation in the limbic portion of the right temporal lobe
recorded with depth electrodes (100). These patients benefited
from surgical resection.
It was found that EEG studies done with spectral analysis in
patients with tension headaches did not differ from normal
persons (and were also not significantly different from the EEG
of migrainous patients) (45).
ACKNOWLEDGMENT
The assistance of Dr. Fowzia Siddiqui is gratefully acknow-
ledged.
REFERENCES
1. Diamond S. Headache: its diagnosis and management (introduc-
tion to symposium). Headache. 1979;19:113.
2. Dalessio DJ. Classification and mechanism of migraine. Headache.
1979;19:114–120.
3. Silberstein SD, Lipton RB, Sliwinski M. Classification of daily and
near-daily headaches: field trial of revised IHS criteria. Neurology.
1996;47:871–875.
4. Sacks O. Migraine. Stuttgart: Kohlhammer. (Original English ver-
sion: Migraine. The Evolution of a Common Disorder. London:
Faber & Faber; 1985.)
5. Wolff HG. Headache and Other Head Pain. New York: Oxford
University Press; 1948.
6. Anthony M, Hinterberger H, Lance JW. Plasma serotonin in
migraine and stress. Arch Neurol (Chicago). 1967;16:544–552.
7. Anthony M. Plasma fatty acids and prostaglandin E in migraine
and stress. Headache. 1970;16:58.
8. Fozard JR. Proceedings of the International Headache Congress,
Florence, 1980. Clin Psychiatry News. 1980;8(9).
9. Moskowitz MA. The neurobiology of vascular head pain. Ann
Neurol. 1984;16:157–168.
10. Moskowitz MA, Nozaki K, Draig RP. Neocortical spreading depres-
sion provokes the expression of c-fos-protein-like immunoreactivity
within trigeminal nucleus caudalis via trigeminovascular mecha-
nisms. J Neurosci. 1993;13:1167–1177.
11. Biggs MJ, Johnson ES. The autonomous nervous system and
migraine pathogenesis. In: Amery WK, Van Nueten JM, Wauquier
A, eds. The Pharmacological Basis of Migraine Therapy. London:
Pitman; 1984:99–107.
12. Diamond S, Dalessio DJ. The Practicing Physician’s Approach to
Headache. Baltimore: Williams & Wilkins; 1982.
13. Schumacher GA, Wolff HG. Experimental studies on headache.
Arch Neurol Psychiatry (Chicago). 1941;45:199–214.
14. Whitty CWM. Familial hemiplegic migraine. J Neurol Neurosurg
Psychiatry. 1953;16:172–177.
15. Skinhoj E, Hoedt-Rasmussen K, Paulson OB, et al. Regional cere-
bral blood flow and its autoregulation in patients with transient
focal ischemic attacks. Neurology (Minneapolis). 1970;20:485–493.
16. Lauritzen M, Skyhoj Olsen T, Lassen NA, et al. Changes in regional
cerebral blood flow during the course of classic migraine attacks.
Ann Neurol. 1983;13:633–641.
17. Hansen AJ, Lauritzen M, Tfeldt-Hansen P. Spreading cortical depres-
sion and antimigrainous days. In: Amery WK, Van Neuten JM,
 Chapter 24 ■ The EEG in Patients with Migraine and Other Forms of Headache
Wauquier A, eds. The Pharmacological Basis of Migraine Therapy.
London: Pitman; 1984:161–170.
18. Olesen J. Conclusions and prospects for the future. In: Olesen J, ed.
Migraine and Other Headaches. New York: Raven Press; 1991:
347–349.
19. Olesen J. Understanding the biologic basis of migraine. N Engl J
Med. 1994;331:1713–1714.
20. Woods RP, Iacoboni M, Mazziota JC. Brief report: bilateral spread-
ing cerebral hypoperfusion during spontaneous migraine
headache. N Engl J Med. 1994;331:1689–1692.
21. Huang J, Cooper TG, Santana B, et al. Visual distortion provoked
by a stimulus in migraine associated with hyperneuronal activity.
Headache. 2003;43:664–671.
22. Lashley KS. Patterns of cerebral integration indicated by scotomas
of migraine. Arch Neurol Psychiatry (Chicago). 1941;46:331–339.
23. Leão AAP. Spreading depression of activity in the cerebral cortex.
J Neurophysiol. 1944;7:359–390.
24. Leão AAP. Spreading depression. In: Purpura DP, Penry JK, Tower
D, et al., eds. Experimental Models of Epilepsy. New York: Raven
Press; 1972:173–196.
25. Ebersberger A, Schaible HG, Averbeck B, et al. Is there a correlation
between spreading depression, neurogenic inflammation, and noci-
ception that might cause migraine headache? Ann Neurol. 2001;49:
7–13.
26. Daly DD, Markand ON. Focal brain lesions. In: Daly DD, Pedley
TA, eds. Current Practice of Clinical Electroencephalography. 2nd
ed. New York: Raven Press; 1990:35–370.
27. Ulett GA, Evans D, O’Leary JL. Survey of EEG findings in 1,000
patients with chief complaint of headache. Electroencephalogr Clin
Neurophysiol. 1952;4:463–470.
28. Jung R. Neurophysiologische Untersuchungsmethoden. In:
Handbuch der Inneren Medizin. 4th ed., Vol. 5/1. Berlin: Springer;
1953:1206–1314.
29. Becher F. Hirnelektrische und elektromyographische. Unter -
suchungen beim Migränesyndrom. Dtsch Z Nervenheilk. 1955;
172:556.
30. Krischek J. Elektroenzephalographische Befunde bei Migräne.
Dtsch Z Nervenheilk. 1956;175:43.
31. Wissfeld E, Neu O. Über die EEG-Verändeungen bei Migräne und
die Bedeutung occipitaler Delta-Wellen im EEG. Nervenarzt. 1960;
31:418.
32. Bille B. Migraine in school children. Acta Paediatr Suppl (Uppsala).
1962;51:136.
33. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol. 3.
Reading: Addison-Wesley; 1964.
34. Heyck H. Neue BeitrŠge zur Klinik und Pathogenese der MigrŠne.
Stuttgart: Thieme; 1956.
35. Heyck H. Der Kopfschmerz. Stuttgart: Thieme; 1958.
36. Weil AA. EEG findings in a certain type of psychosomatic headache:
dysrhythmic migraine. Electroencephalogr Clin Neurophysiol. 1952;
4:181–186.
37. Weil AA. Observation on dysrhythmic migraine. J Neurol Ment
Dis. 1962;134:277–281.
38. Dow DJ, Whitty CW. Electroencephalographic changes in migraine.
Review of 51 cases. Lancet. 1947;2:52–54.
39. Selby G, Lance JW. Observations on 500 cases of migraine and allied
vascular headache. J Neurol Neurosurg Psychiatry. 1960;23:23–32.
40. Barolin GS. Migraines and epilepsies—a relationship? Epilepsia
(Amsterdam). 1966;7:53–66.
41. Gschwend J. EEG-Befunde und ihre Interpretation bei einfacher
Migräne. J Neurol (Berlin). 1972;201:279–292.
42. Pithova B. Clinico-EEG correlation in migraine. Electroencephalogr
Clin Neurophysiol. 1983;55:31P(abst.).
463
43. Smyth VOG, Winter AL. The EEG in migraine. Electroencephalogr
Clin Neurophysiol. 1964;16:194–202.
44. Jonkman EJ, Lelieveld MHJ. EEG computer analysis in patients
with migraine. Electroencephalogr Clin Neurophysiol. 1981;52:
652–655.
45. Drake ME, Huber SJ, Pakalnis A, et al. Computerized EEG spectral
analysis in migraine and tension headaches. J Clin Neurophysiol.
1987;4:301.
46. Golla FL, Winter AL. Analysis of cerebral responses to flicker in
patients complaining of episodic headache. Electroencephalogr Clin
Neurophysiol. 1959;11:539–549.
47. Slater KH. Some clinical and EEG findings in migraine. Brain.
1968;91:85–98.
48. Simon RH, Zimmerman A, Tasman A, et al. Spectral analysis of
photic stimulation in migraine. Electroencephalogr Clin Neurophysiol.
1982;53:270–276.
49. Kugler J. Elektroenzephalographie und Beziehungen zur Epilepsie.
In: Soyka D, ed. Die MigrŠne. Dusseldorf: Labaz; 1979:51–62.
50. Scollo-Lavizzari G. Das Elektroenzephalogramm bei der Migräne.
Schweiz Rdsch Med (Praxis). 1975;64:234–237.
51. Westmoreland B. EEG in the evaluation of headaches. In: Klass DW,
Daly DD, eds. Current Practice of Clinical Electroencephalography.
New York: Raven Press; 1978;381–394.
52. Schoenen J, Jamart B, De Pasqua V, et al. Mapping of EEG and
auditory event-potentials in migraine. Electroencephalogr Clin
Neurophysiol. 1990;75:S134(abst.).
53. Gorman MJ, Welch KMA. Cerebral blood flow and migraine. In:
Phillips JW, ed. The Regulation of Cerebral Blood Flow. Boca Raton:
CRC Press; 1993:399–410.
54. Soysal A, Atay T, Ozturk M, et al. Pattern reversal visual evoked
potentials and EEG in migraine with and without visual aura.
J Neurol Sci (Turkish). 2001;18:1–8.
55. Farkas V, Szeg L, Kohlhéb O. Hemiplegic migraine in childhood.
Differential diagnosis and EEG aspects. Electroencephalogr Clin
Neurophysiol. 1985;61:4P–5P(abst.).
56. Bradshaw P, Parsons M. Hemiplegic migraine: a clinical study.
J Med. 1965;34:65–85.
57. Degen R, Degen HE, Palm D, et al. Die Migraine hémiplégique im
Kindesalter. Dtsch Med Wochenschr. 1980;105:640–645.
58. Heron JR. Migraine and cerebrovascular disease. Neurology
(Minneapolis). 1966;16:1097–1104.
59. Rosenbaum HE. Familial hemiplegic migraine. Neurology
(Minneapolis). 1960;10:164–170.
60. Isler W. Akute Hemiplegien und Hemisyndrome im Kindesalter.
Stuttgart: Thieme; 1969.
61. Müller D, Müller J. Die familiäre hemiplegische Migräne. Z Artzl
Fortibild. 1977;71:763–767.
62. Martens de Nordhout A, Timsit-Bertheir M, Timsit M, et al.
Contingent negative variation in headache. Ann Neurol. 1986;19:
78–80.
63. Ahmed I. Contingent negative variation in migraine: effect of beta
blocker therapy. Clin Electroencephaologr. 1999;30:21–23.
64. Lérique-Koechlin A, Mises J. L’EEG dans une manifestation parox-
ystique non-épileptique de l’enfant: la migraine. Electroencephalogr
Clin Neurophysiol. 1964;16:203–204.
65. Kinast M, Lueders H, Rothner AD, et al. Benign focal epileptiform
discharges in childhood migraine (BFEDC). Neurology (New York).
1982;32:1309–1311.
66. Kellaway P, Crawley JW, Kagawa N. Paroxysmal pain and auto-
nomic disturbances of cerebral origin. A specific electroclinical
syndrome. Epilepsia (Amsterdam). 1960;1:466–483.
67. Moore MT. Paroxysmal abdominal pain. A form of symptomatic
epilepsy. JAMA. 1945;129:1233–1240.
464 Part III ■ Clinical EEG: General
68. Seshia SS, Reggin JD, Stanwich RS. Migraine and complex seizures
in children. Epilepsia (New York). 1985;26:232–236.
69. Koehler B. Benign paroxysmal vertigo of childhood: a migraine
equivalent. Eur J Pediatr. 1980;134:149–151.
70. Bickerstaff ER. Basilar artery migraine. Lancet. 1961a;1:15–17.
71. Bickerstaff ER. Impairment of consciousness in migraine. Lancet.
1961b;2:1057–1059.
72. Caplan LR. Migraine and vertebrobasilar ischemia. Neurology.
1991;41:55–61.
73. Lapkin ML, French JH, Golden GS, et al. The electroencephalogram
in childhood basilar artery migraine. Neurology (Minneapolis). 1977;
27:580–583.
74. Camfield PR, Metrakos K, Andermann F. Basilar migraine,
seizures and severe epileptiform EEG abnormalities. Neurology
(Minneapolis). 1978;28:584–588.
75. Panayatopoulos CP. Basilar migraine, seizures and severe epilep-
tic EEG abnormalities. Neurology (Minneapolis). 1980;30:
1122–1125.
76. Parain D, Samson-Dollfus D. Electroencephalograms in basilar
artery migraine. Electroencephalogr Clin Neurophysiol. 1984;58:
392–399.
77. Gastaut H. A new type of epilepsy: benign partial epilepsy of child-
hood with occipital spike-waves. Clin Electroencephalogr. 1982;13:
13–22.
78. Walker MC, Smith SJM, Sisodiya SM, et al. Case of simple partial
status epilepticus in electrophysiological, and magnetic resonance
imaging characteristics. Epilepsia. 1995;36:1233–1236.
79. Ramelli GP, Sturzenegger M, Donati F, et al. EEG findings during
basilar migraine attacks in children. Electroencephalogr Clin
Neurophysiol. 1998;107:374–378.
80. Cernibori A, Bouquet F. Loss of consciousness during basilar
artery migraine attack in childhood: EEG and clinical studies.
Electroencephalogr Clin Neurophysiol. 1984;58:72P(abst.).
81. Ganji S, Hellman S, Stagg S, et al. Episodic coma due to acute basi-
lar artery migraine: correlation of EEG and brain stem auditory
evoked potentials. Clin Electroencephalogr. 1993;24:44–48.
82. Breslau N, Chilcoat HD, Andreski P. Further evidence on the
link between migraine and neuroticism. Neurology. 1996;47:
663–667.
83. Alvarez WC. Migraine plus epilepsy. Neurology (Minneapolis). 1959;
9:487–491.
84. Giardina M, Sideri S. Migraine and epilepsy. Electroencephalogr
Clin Neurophysiol. 1985;60:71P(abst.).
85. Basser LS. The relation of migraine and epilepsy. Brain. 1969;92:
285–300.
86. Lance JW, Anthony M. Some clinical aspects of migraine. A
prospective survey of 500 patients. Arch Neurol (Chicago). 1966;
15:356–361.
87. Lees R, Watkins SM. Loss of consciousness in migraine. Lancet.
1963;2:647–650.
88. Hess R. Epilepsie und Kopfschmerzen. A. EEG-EMG. 1977;8:
125–136.
89. Andermann F, Lugaresi E, eds. Migraine and Epilepsy. Boston:
Butterworth; 1987.
90. Andermann F, Andermann E. Migraine and epilepsy with special
reference to the benign epilepsies of childhood. In: Degen R,
Dreifuss FE, eds. Benign Localized and Generalized Epilepsies of
Early Childhood. Amsterdam: Elsevier; 1992:207–214.
91. Niedermeyer, E. Epilepsy Guide. Baltimore: Urban & Schwarzenberg;
1983.
92. Niedermeyer, E. The Epilepsies. Baltimore: Urban & Schwarzenberg;
1990.
93. Niedermeyer, E. Migraine-triggered epilepsy. Clin Electroencephalogr.
1993;24:37–43.
94. Jacobs J, Goadsby PJ, Duncan JS. Use of sumatriptan in post-ictal
migraine headache. Neurology. 1996;47:1104.
95. Ottman R, Lipton RB. Is the comorbidity of epilepsy and migraine
due to shared genetic susceptibility? Neurology. 1996;47:918–924.
96. Sbrascini S, Bassi P. Headache and slow hypersynchronization of
the EEG during hyperventilation. Electroencephalogr Clin
Neurophysiol. 1983;55:3P(abst.).
97. Raskin NH. The hypnic headache syndrome. Headache. 1988;28:
534–536.
98. Evers S, Goadsby PJ. Hypnic headache. Neurology. 2003;60:
905–909.
99. Heyck H, Hess R. Vasomotorische Kopfschmerzen als Symptom
larvierter Epilepsien. Schweiz Med Wochenschr. 1955;85:573–575.
100. Laplante P, Saint-Hilaire JM, Bouvier G. Headache as an epilep-
tic manifestation—two cases studied with depth electrodes.
Electroencephalogr Clin Neurophysiol. 1982;54:36P–37P(abst.).
101. Olesen J, Tfelt-Hansen P, Welch DMA, eds. The Headaches. New
York: Raven Press; 1993.
 Part IV Clinical EEG in Epilepsy and Related Disorders
Seizures and Epilepsies in the
Preterm and Term Neonate
EMMA LAURETA, ELI M. MIZRAHI, AND SOLOMON L. MOSHÉ
NEONATAL SEIZURES
Historical Aspects
Initial investigations and attempts to characterize and classify
neonatal seizures began in the 1950s and 1960s, as neonatal
electroencephalography (EEG) began to emerge and well before
video-EEG recordings became established as diagnostic tools
(1–4). Early investigators relied on clinical observation or
observation supplemented by interictal EEG recordings (5,6).
Later, bedside EEG recordings of infants who were experiencing
seizures combined with cinematography, improved correlation
of observed clinical phenomenon with electrographic activity
(7–10).
These pioneering French investigators observed that seizures
in the neonate had unique clinical features that distinguished
them from seizures generated by more mature brains. For
example, generalized tonic–clonic seizures were found not to
occur (5,10–12). The more common focal clonic seizures were
often asynchronous if they occurred bilaterally and did not
spread in typical Jacksonian sequence (2,8–10).
Also, in the 1950s and 1960s, paroxysmal phenomenon with
minimal motor manifestations began to be recognized as clini-
cal seizures. These included paroxysmal ocular and oro-buccal-
lingual movements, repetitive limb movements resembling
pedaling, rowing or swimming movements, changes in skin
color, respiration, and other autonomic changes (5,7). They
were initially characterized as “anarchic or atypical” and later
various other terms have been used to describe these seizures
including “slight,” “minimal,” “subtle,” or “motor automatisms”
(9,10,13–16). Many of the so-called subtle seizures were subse-
quently found not to be associated with simultaneous electro-
graphic discharges. This was also observed to be true for other
seizures, such as those characterized by generalized tonic pos-
turing. Over the years, this has led to controversy over their
pathophysiology. Observing that these events were similar to
certain reflex behaviors, Mizrahi and Kellaway postulated that
these were primitive brainstem reflexes “released” by forebrain
depression, rather than true epileptic events (17,18).
These observations and controversies have led to the evolu-
tion of current classification systems of neonatal seizures,
which are primarily based on clinical semiology, electroen-
cephalographic features, the temporal relationship of clinical
and EEG features, and presumed pathophysiology. They have
also led to several issues regarding diagnosis and management
of neonatal seizures especially those unaccompanied by promi-
nent clinical manifestations.
CHAPTER
25
Terminology
In the term infant, the neonatal period is defined as the first 28
days of life. In the preterm infant, this period extends to the 44
completed weeks gestational age. It is a period of increased vul-
nerability to anoxic and metabolic stress, and most seizures dur-
ing this period are acute reactive seizures, as opposed to
manifestations of neonatal epilepsy––although the latter may
occur with early neonatal seizures being the first of a chronic
condition (19). However, the term “epileptic” is used in current
classification systems to denote presumed underlying pathophys-
iology––seizures generated by hypersynchronous cortical neu-
ronal discharges as demonstrated by temporally related EEG
changes. This is in contrast to “nonepileptic” seizures that occur
without any EEG correlate and are of unclear pathophysiology,
most likely based in reflex physiology. Focal clonic, focal tonic
events, some myoclonic events and spasms are most often epilep-
tic. Generalized tonic posturing and motor automatisms are
often purely clinical. Seizures recorded in the absence of clinical
activity are termed “electrical only or electrographic seizures.”
These may occur in infants not treated with antiepileptic drugs
(AEDs) with encephalopathy or those with electroclinical
seizures who have been treated with AEDs with subsequent
“decoupling” of electrical and clinical manifestations (17).
Epidemiology
Reported incidence of neonatal seizures range from 1.8 to 5.5
per 1000 newborns (20–23). The National Collaborative
Perinatal Project, which studied infants of 54,000 pregnant
women between 1959 and 1966, reported an incidence of
5.1/1000 live births (24). More recent population-based studies
in Fayette County, Kentucky; Newfoundland, Canada; and
Harris County, Texas have reported lower incidences rates,
between 1.8/1000 and 2.5/1000 live births (20–23). Differences
in target population, periods of surveillance, methods used in
case identification, and use of clinical versus clinical and EEG
criteria in the identification of cases likely account for the vari-
ability in reported rates (25). The incidence of neonatal seizures
is higher in premature and low birth weight infants, as well as
infants cared for in intensive care units (23). Neonatal seizures
are most common in the first week of life (20,26).
Pathophysiology
Seizures occur more frequently in the neonatal period than at any
other time in life. Basic science research has shown various fac-
tors underlie this increased susceptibility to seizures, based pri-
marily on relatively enhanced excitation in the immature brain
465
466 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
(27–32). For example, in the immature brain, potassium tends to
accumulate in the extracellular space, secondary to decreased
Na , K-ATPase activity, and immature enzyme systems. This
leads to the development of a hyperexcitable state and decreased
seizure threshold (27). Excitation may also be enhanced by a rel-
ative increase in the density of glutamate receptors and NMDA-
gated channels. On the other hand, synapses and receptors for
inhibitory neurotransmitters may be less abundant in the devel-
oping brain. Also, GABA, the main inhibitory neurotransmitter
in the CNS, may have excitatory functions early on (33–35). In
animal models, the substantia nigra pars reticularis, which in
adults is involved in the control of seizures, has been shown to
amplify seizures in immature animals. These and other factors,
not only lower the overall seizure threshold, but likely, also allow
seizures to spread more readily (25).
Long-term consequences of neonatal seizures on brain devel-
opment are unclear and likely affected by multiple factors, such
as seizure burden, age of the infant when seizures occur, and
underlying etiology. Despite the increased susceptibility to
seizures, studies using animal models/rodents provide evidence
to suggest that the immature brain is more resistant to seizure-
induced hippocampal injury (28). However, in rats with a preex-
isting lesion, the risk of status epilpeticus (SE)–induced injury is
significantly elevated (36). Also severe recurrent seizures can
lead to impaired synaptogenesis, myelination, and brain growth.
Galanopoulou has shown that three SE in P4-6 rat pups may
upregulate the expression of the chloride cotransporter KCC2 in
immature neurons with still depolarizing GABAA receptor
responses, which may prevent the maintenance of the GABA-
mediated effects on cell maturation (37,38). This effect differs
from those described in adults and may therefore underlie and
partially explain the age-specific consequences of SE.
There are also studies that suggest that exposure to currently
available AEDs used to control seizures during this period, may
have similar detrimental effects causing increased apoptosis or
even on occasion promote seizures. In each case, the cost bene-
fit ratio of the effects of seizures or AEDs must be determined,
although ideally age-specific and even gender-specific treat-
ments should be developed (see the following section).
Clinical and EEG Features of Neonatal Seizures
Characterization and Classification of Neonatal Seizures
Traditional classification schemes for seizures have been diffi-
cult to apply to neonatal seizures that are generally fleeting
events occurring in infants with limited repertoire of normal
behavior to begin with. More than one seizure type can occur
in a particular infant.
The two most widely accepted classification schemes for
neonatal seizures are by Volpe and by Mizrahi and Kellaway. Each
characterizes seizures based on semiology of the most prominent
movement. In Volpe’s classification scheme, these neonatal
seizures are classified as (i) clonic, (ii) tonic, (iii) myoclonic, and
(iv) subtle (15,39). Mizrahi and Kellaway’s classification scheme,
developed from findings on their EEG-polygraphic monitoring
studies, not only considers semiology, but emphasizes the associ-
ation of clinical behaviors with presumed pathophysiology based
on the presence or absence of simultaneous EEG changes (16,17).
In their classification scheme, neonatal seizures are broadly cate-
gorized as (i) epileptic events—when associated with consistent
electrocortical signature, (ii) nonepileptic events— when unac-
companied by time-synchronized EEG changes, and (iii) electri-
cal seizures—when EEG seizure discharges are not associated
with obvious clinical manifestations (17).
The types of seizures found to be most consistently associ-
ated with EEG changes include focal clonic, focal tonic, some
myoclonic seizures, and spasms. Seizures not consistently asso-
ciated with an electrographic signature include generalized
tonic, some myoclonic, and motor automatisms. Mizrahi and
Kellaway used the term “motor automatisms” to describe
seizures consisting of a variety of paroxysmal behaviors such as
oral-buccal- lingual movements, ocular signs, progression
movements, and complex purposeful movements. Each clinical
seizure type and their associated EEG abnormalities are
described in detail below.
Focal Clonic: These seizures consist of rhythmic, usually slow
(approximately one to three jerks per second) repetitive move-
ments of the face, proximal or distal arm or leg muscles, or axial
structures. The movements may be unifocal (confined to one
part of the body); hemiconvulsive (involving one side of the
body); or multifocal (involving multiple body parts on both
sides). Clonic seizures that involve both sides of the body simul-
taneously are usually asynchronous unlike true generalized
clonic seizures in older individuals. Seizures may alternate
between sites within a particular seizure, or may migrate from
one site to another, not necessarily spreading in traditional
Jacksonian sequence. Focal clonic seizures are usually associ-
ated with focal brain lesions such as cerebral infarctions but can
also occur with more diffuse neuropathologic processes. By far,
they are the type of seizures most consistently associated with
electrographic correlate.
Focal Tonic: Sustained posturing of a limb, or unilateral flex-
ion of the trunk characterize these seizures that can also be
accompanied by sustained conjugate deviation to one side. Like
focal clonic seizure, tonic seizures are usually associated with
synchronized EEG discharges. They occur less frequently than
focal clonic seizures.
Generalized Tonic: These seizures are characterized by sus-
tained bilateral extension or flexion of the limbs or trunk,
resembling decerebrate or decorticate posturing. Unlike their
more focal counterpart, these seizures are generally not associ-
ated with EEG correlate and are classified as nonepileptic.
Because they tend to occur in association with other automatic
behavior in obtunded infants with diffuse central nervous
pathologic processes, Mizrahi and Kellaway proposed that such
seizures are “brainstem release phenomenon” from disinhibi-
tion of forebrain structures in clinically depressed infants.
Myoclonic: Irregular, erratic twitching, or contractions of
muscle groups involving the face, limbs, or trunk are termed
myoclonus. These can be either epileptic or nonepileptic.
Myoclonic movements associated with EEG changes are also
called “cortical myoclonus,” while myoclonus with no EEG
 Chapter 25 ■ Seizures and Epilepsies in the Preterm and Term Neonate
correlate is termed “sub-cortical myoclonus.” The latter term
implies that more caudal structures in the brainstem or spinal
cord responsible for generating the movements. The seizures
can be focal, usually involving muscles of one upper extremity;
or they may be generalized or multifocal. Generalized
myoclonus consists of bilateral symmetric flexion jerks of the
limbs or trunk. Multifocal or “fragmentary” myoclonus is char-
acterized by brief asynchronous twitching of different muscle
groups. Of the three types, fragmentary myoclonus is the type
least associated with EEG changes, while two thirds of general-
ized myoclonus have EEG correlate.
Spasms: Spasms characterized by bilateral flexor, extensor, or
mixed flexion and extension contractions of limb and truncal
muscles occur rarely in the neonate. Like spasms in older
infants, they can occur in clusters and can be more frequent
upon awakening. On EEG, they are associated with a general-
ized attenuation of the background or a high-voltage slow wave.
Spasms are not included in Volpe’s classification scheme.
Motor Automatisms: Also called “subtle seizures,” paroxysmal
changes in behavior or autonomic functions with minimal
motor manifestations occur frequently in the newborn. These
consist of various irregular and disconjugate ocular move-
ments, eye-opening, chewing, oral-buccal movements, and
peculiar extremity movements such as pedaling, stepping, box-
ing, swimming movements. Mizrahi and Kellaway found these
behaviors to be inconsistently associated with EEG seizure and
classify them as nonepileptic. However, similar behaviors, in
preterm, and some full term infants have been demonstrated by
other investigators to be ictal phenomena associated with
simultaneous abnormal EEG discharges (40). Similarly, in older
individuals, paroxysmal automatic behaviors are known to
occur without surface EEG correlate as manifestations of psy-
chomotor seizures or postictal behavior (41). Studies in neona-
tal animals have also shown that epileptiform discharges can
occur in deep limbic structures with minimal or no propaga-
tion to the cortical surface (30). These observations raise the
possibility that subtle seizures in neonates may be true epileptic
events with generators in deeper limbic structures if concomi-
tant electrographic seizure activity is the only criteria. These
observations, however, do not exclude the possibility that some
of these events behavior may be reflex behaviors.
467
Autonomic Signs: A variety of paroxysmal changes in auto-
nomic activity such as alterations in breathing, heart rate, blood
pressure, salivation, sweating, color changes, have also been
described as manifestations of subtle seizures (42–44). However,
in isolation, they rarely occur as epileptic phenomena. When
they are accompanied by ictal EEG correlates, they occur in asso-
ciation with other clinical behavioral or motor phenomenon. In
a study by Watanabe et al., 14 of the 21 infants with apneic
seizures with electrical EEG correlate exhibited other subtle phe-
nomenon such as eye-opening, “staring,” deviation of the eyes
and mouth movements (45). Bradycardia was also found to be
more common in “convulsive apnea” versus “nonconvulsive”
apnea in another study (46).
Ictal EEG Correlates
Electrical seizures are more common in term than in preterm
infants. When they occur in infants less than 34 to 35 weeks,
they are shorter in duration (23). Onset is usually focal, usually
in the central and temporal regions (Fig. 25.1). Less common
sites of origin are the frontal and occipital regions. Seizures can
arise from more than one area, propagating asynchronously
from different foci (Fig. 25.2). Frequency, morphology, and
voltage of the discharge may vary within the same seizure, or
from one seizure to the next in a given infant. The seizure dis-
charge can consist of repetitive spikes, sharps or slow waves, or
a combination of different waveforms. Seizures may remain
localized to a specific area, slowly spread to involve contiguous
regions, abruptly involve one of hemisphere, or migrate to the
other hemisphere (25).
Electrographic seizures in the absence of clinical activity in
severely encephalopathic neonates are characterized by back-
ground EEG activity that is usually depressed and undifferenti-
ated. Some are lower in voltage, longer in duration, and show
little tendency to spread or change. These have been referred to
as “seizure discharges of the depressed brain” (26) (Fig. 25.3).
Alpha seizure discharges are characterized as paroxysmal 8 to
12 Hz alpha activity in the central or temporal regions in
encephalopathic newborns and suggest a poor prognosis
(45,47,48) (Fig. 25.4).
When neonates are treated for seizures with antiepileptic
medications, persistent electrical seizure activity can be found
Figure 25.1 Central onset of electrical seizure
activity. Rhythmic sharp waves arise in the left
central region and remain confined to that region,
associated with a clinic focal clonic seizure involv-
ing the right hand. The EEG was recorded from a
40-week C.A. infant with a left frontal lobe infarc-
tion. (Reprinted with permission from Mizrahi
EM, Hrachovy RA, Kellaway P. Atlas of Neonatal
Electroencephalography. 3rd ed. Philadelphia:
Lippincott Williams & Wilkins; 2004:250.)
468 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Figure 25.2 Multifocal electrical seizure activ-

ity. Seizure discharges occur simultaneously but
asynchronously from the two hemispheres in this
39-week C.A. infant with hypoxic-ischemic
encephalopathy. There is initial rhythmic,
moderate-voltage, sharp-wave activity in the
right centrotemporal region and then an inde-
pendent discharge arising from the left temporal
region characterized by sharp- and slow-wave
activity with complex morphology. No clinical
seizures accompanied these discharges.
(Reprinted with permission from Mizrahi EM,
Hrachovy RA, Kellaway P. Atlas of Neonatal
Electroencephalography. 3rd ed. Philadelphia:
Lippincott Williams & Wilkins; 2004:250.)

long after clinical seizures have stopped. This “decoupling” of
electrical from clinical activity, first described by Mizrahi and
Kellaway, has been demonstrated by other investigators (17). In
a study by Clancy and coworkers, 79% of 393 electrographic
seizures recorded were not accompanied by clinical seizure
activity. Seizures were the result of diverse etiologies and most
infants had received one or more AED prior to the EEG record-
ings (49). Similar seizures without clinical correlates were also
noted by Scher, Bye and Flanagan, and Murray to occur fre-
quently (23,40,50).
Interictal EEG Correlates
Although the documentation of simultaneous EEG discharge
during clinical events is useful in the diagnosis of seizures in the
neonate, the finding of sharp-wave transients on interictal
records does not aid in the diagnosis of seizures, or predict
epileptogenicity. Unlike in adults, transient sharp waves, fre-
quently found in the central and temporal regions are a nonspe-
cific finding that can be seen in pathologically or as normal
finding in term and preterm infants (26,51,52). Early studies
suggested that positive rolandic sharp waves were most specific
for intraventricular hemorrhage (53–55), but subsequently have
been more closely associated with periventricular leukomalacia.
However, in the setting of neonatal seizures, assessment of
the interictal EEG background is still clinically useful. It can
suggest possible etiologies for the seizures, help in determining
an infant’s degree of cerebral dysfunction, risk for persistent
seizures, and prognosis for long-term outcome.
EEG findings correlating with diffuse cerebral dysfunction
include a depressed undifferentiated background, suppression-
burst pattern, multifocal sharp transients, and dysmaturity.
These abnormalities, although not typically etiologic specific,
suggest etiologies which result in generalized or multifocal brain
injury, such as hypoxia, acute metabolic disturbances, inborn
errors of metabolism, infection, and bilateral brain hemorrhage.
On the other hand, persistent asymmetries of the background
and persistently lateralized sharp transients may indicate focal
structural brain abnormalities such as focal cerebral infarction
or areas of dysgenesis as underlying etiologies. The absence of
focal features does not rule a focal structural pathology (25).
The EEG obtained within the first 24 hours can help in
determining prognosis and risk for further seizures. Prospective
studies in infants presenting with seizures in the neonatal
period have found that the background EEG can be used for
prognostic purposes to predict survival and long-term outcome
(25,56–58). A normal EEG suggests a good prognosis while the

Figure 25.3 Seizure discharge of the depressed

brain. Low-voltage, rhythmic, monomorphic,
slow sharp waves on the left persist virtually
unchanged throughout the recorded seizure. No
clinical seizures were present. The background
activity is depressed and undifferentiated. The
infant is 39-week C.A. with hypoxic-ischemic
encephalopathy. (Reprinted with permission
from Mizrahi EM, Hrachovy RA, Kellaway P.
Atlas of Neonatal Electroencephalography. 3rd
ed. Philadelphia: Lippincott Williams & Wilkins;
2004:250.)
 Chapter 25 ■ Seizures and Epilepsies in the Preterm and Term Neonate
presence and persistence of diffuse abnormalities on serial
EEGs suggest poor outcome.
EEG background may also predict persistent seizures in the
acute period as well as in the postneonatal period. Laroia and
coworkers, studying 51 infants at risk for neonatal seizures, found
that a normal or immature background strongly predicted the
absence of electrographic seizures in subsequent 18- to 24-hour
period of video-EEG monitoring, while background abnormali-
ties predicted the occurrence of electrographic seizures within
that same time frame (59). Among the 27 survivors of neonatal
seizures followed for a mean of 31 months by Clancy and
coworkers, postneonatal epilepsy occurred in 68% (13 of 19) of
patients with moderately or markedly abnormal EEG back-
grounds but in only 25% (2 of 8) without (60). Ortibus and col-
leagues however, did not find background abnormalities to be a
strong predictor of seizures past the neonatal period (56).
Diagnosis and Treatment of Acute Seizures
Etiology
Major etiologies for acute seizures in the neonate include (i)
neonatal encephalopathy and hypoxic-ischemic encephalopa-
thy; (ii) intracranial hemorrhage and infarctions; (iii) infec-
tions; (iv) metabolic disturbances; (v) congenital structural
brain lesions; and (vi) drug withdrawal (25,61). The term
neonatal encephalopathy has gained favor in the literature over
recent years over the more specific term, hypoxic-ischemic
encephalopathy, as oftentimes, a direct causal relationship
between hypoxia and encephalopathy is hard to establish.
Metabolic disturbances causing seizures are important to
identify as many acute disturbances are potentially treatable with
etiology-specific therapy. These include electrolyte abnormalities
such as hypocalcemia, hypomagnesemia, and hypoglycemia.
Rare inborn errors of metabolism such as urea cycle defects,
aminoacidurias, and organic acidurias can cause neonatal
seizures refractory to AED treatment. These usually require
dietary manipulation to control. Pyridoxine-dependent
seizures can produce severe seizures, including spasms that may
respond to pyridoxine.
Systemic and CNS infections can be secondary to postna-
tally acquired bacterial and viral agents, or congenital, in-utero
infections.
469
Figure 25.4 Alpha seizure discharge. A seizure
discharge in the left temporal region is characteri-
zed by sinusoidal rhythmic 10 to 11 Hz activity
that evolved from rhythmic sharp wave activity
in that region. There is also an independent
seizure discharge in the left occipital region
appearing as semiperiodic sharp waves. The
background activity is depressed and undifferen-
tiated in this 38 week C.A. infant with meningi-
tis. (Reprinted with permission from Mizrahi EM,
Hrachovy RA, Kellaway P. Atlas of Neonatal
Electroencephalography. 3rd ed. Philadelphia:
Lippincott Williams & Wilkins; 2004:250.)
Common acquired structural brain lesions include intracra-
nial hemorrhage and cerebral infarctions. Focal areas of cerebral
dysgenesis and more diffuse anomalies from major congenital
brain malformations, such as in lissencephaly, can produce
seizures as well in the neonatal period. The presence of a struc-
tural brain lesion can be suggested by focal seizures, electro-
graphically arising from the same region, or from asymmetries
in voltage or frequency of background activities. Neuroimaging
is essential when such abnormalities are found on EEG.
Drug withdrawal or intoxication is another cause of tran-
sient seizures in the newborn. History of maternal drug use or
of administration of medications during delivery is suggestive
of the diagnosis. The affected neonate can appear depressed and
exhibit symptoms such as jitteriness that can be confused with
seizures.
When infectious, metabolic, and structural causes of seizures
are excluded, neonatal encephalopathy secondary to hypoxic
injury is often implicated, especially if accompanied by other sys-
temic signs of anoxia such as persistently low Apgar scores, sys-
temic acidosis, and multiorgan involvement. Grading schemes
and criteria used to diagnose neonatal encephalopathy vary and
it is difficult to assess the true proportion due to asphyxia (62,63).
Perlman and colleagues looked at various clinical and laboratory
markers within the first hour of life that identify newborn infants
at high risk for seizures secondary to hypoxia ischemia. In their
study of 96 infants, significant relationships with seizures were
found with the following variables: 5-minute Apgar score of 5 or
less, the need for intubation in the DR, umbilical cord arterial pH
of 7.00 or less, and base deficit of –14 mEq/L (64). In 49 infants
monitored with continuous EEG, Murray and colleagues found
similar positive predictive value for seizures for low Apgar scores,
systemic acidosis, and the need for intubation.
Surgery for congenital heart disease in the neonatal period
is a procedure that carries increased risk for seizures.
Electrographic seizures were found in 11% of 183 post-op
infants monitored by video-EEG in a study by Clancy and col-
leagues (65) (Table 25.1).
Differential Diagnosis
The differential diagnosis of neonatal seizures includes physio-
logic and nonphysiologic events that may be abnormal or a
470 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Tabl e 2 5 . 1 neonatal behavioral repertoire. Apnea or changes in autonomic

functions are rarely epileptic in origin when they occur in the
Most Frequently Identified Etiologies of Neonatal absence of other autonomic or motor manifestations (45).
Seizures a
Treatment
Hypoxia-ischemia Seizures are often the first sign of neurologic dysfunction in the
Intracranial hemorrhage newborn. Prolonged seizures can adversely affect an infant’s
homeostasis and protracted or frequent seizures have been
Intraventricular associated with increased morbidity (74). Thus, seizures in the
Intracerebral neonate are considered to be urgent medical conditions that
Subdural require prompt evaluation and treatment.
Management begins with stabilization of the infant with par-
Subarachnoid ticular attention given to maintenance of airway, breathing, and
Infection––CNS circulation. Establishing intravenous access for medications is of
Meningitis prime importance since this is the most reliable route for achiev-
ing immediate therapeutic blood levels of anticonvulsants.
Encephalitis Although determining etiology should not delay the treat-
Intrauterine ment of seizures, in certain circumstances, etiology-specific ther-
Infarction apy alone is required to control seizures. This is particularly true
with seizures secondary to metabolic disturbances such as hypo-
Metabolic glycemia, hypocalcemia, hypomagnesemia, hypo- or hyperna-
Hypoglycemia tremia, hyperammonemia, or pyridoxine-dependent seizures.
Hypocalcemia Similarly, antibiotics and antiviral agents are used empirically
when infectious causes are suspected, until ruled out by CSF and
Hypomagnesemia blood cultures (61).
Chromosomal anomalies The current practice is to treat highly suspicious clinical
Congenital abnormalities of the brain behavior as seizures with anticonvulsants even without initial
EEG confirmation. Focal clonic and tonic movements correlate
Neurodegenerative disorders well with electrographic seizures in the neonate. On the other
Inborn errors of metabolism hand, motor automatisms, myoclonus, bilateral tonic stiffening,
Benign neonatal convulsions as well as movements that are highly stimulus sensitive, corre-
late poorly with electrographic seizures. Therefore, it is ideal to
Benign familial neonatal convulsions confirm their epileptic pathophysiology with EEG before con-
Drug withdrawal or intoxication tinuing treatment with anticonvulsants.
When initiating antiepileptic therapy, physiologic factors
a
Listed in relative order of frequency. Not listed is “unknown” etiology, which is unique to the neonatal period have to be taken into considera-
encountered in approximately 10% of cases. (From Mizrahi EM, Kellaway P. tion. These may influence the choice, route of administration,
Diagnosis and Management of Neonatal Seizures. 1st ed. Philadelphia: Lippincott
and dosing of medications. The oral route of administration for
Williams & Wilkins; 1998.)
initiation of AED therapy is not appropriate. Absorption of
medication is decreased due to immaturity of the GI system and
variation of normal neonatal behavior. Infants withdrawing from from early variations in gastric pH (75). Intramuscular absorp-
in-utero drug exposure or who have suffered anoxic or acute tion of medications is also erratic and inconsistent due to
metabolic insults can have tremulous or jittery movements, as increased water content and smaller muscle mass in the new-
well as exaggerated clonus (66,67). Unlike seizures, these move- born. The same factors also lead to higher risk of local injury
ments are very stimulus sensitive and can be suppressed by repo- and muscle necrosis with this route of administration (76).
sitioning or restraint. Physiologic myoclonus is a common sleep Therefore, the most dependable route for administering medica-
phenomenon mistaken for seizures. Some infants have a particu- tions remains intravascular, and establishing access is critical
larly exaggerated form of it that persists for the first few months early on.
but is benign. Because of its occurrence in sleep it is called benign The half-lives and free drug concentration of many medica-
neonatal sleep myoclonus (68). Infants with hyperekplexia, a tions are increased due to a combination of factors. These
condition caused by a mutation in genes encoding for inhibitory include a greater proportion of total body water in the neonate,
glycine receptor, also tend to have prominent myoclonus in sleep decreased albumin concentration and plasma binding of drugs,
or upon arousal. In addition, they have excessive startle response decreased drug metabolism, and excretion from immature
triggered by minimal tactile or auditory stimulus, as well as hepatic and renal systems (19). This scenario makes anticonvul-
episodes of generalized tonic stiffening (68–73). As mentioned sant drug levels highly variable, especially when the infant is
earlier, subtle sudden or repetitive mouth or limb movements systemically ill and receiving other medications. Close monitor-
can raise the suspicion for seizures, but can be part of normal ing of drug levels is necessary especially early on in the course
 Chapter 25 ■ Seizures and Epilepsies in the Preterm and Term Neonate
of treatment that needs to be individualized. Medication
dosages often need adjustment over time, as the infant’s physi-
ologic systems mature.
Antiepileptic Therapy AED therapy is begun with an initial
intravenous bolus given to achieve targeted therapeutic levels of
the drug. Subsequent smaller boluses of the same medication
can be given to control persistent clinical seizures. If seizures
continue despite titration to maximal levels or clinical toxicity,
a second agent is added. The ideal is to control both clinical and
electrographic seizures; however, it is controversial to aggres-
sively pursue purely electrographic seizures that often require
high dosages of potentially toxic AED levels for control.
Two medications most frequently used as first-line agents
are phenobarbital and phenytoin/fosphenytoin (77), although
phenobarbital is the most favored. Studies have shown them to
be equally effective in controlling seizures, although overall
effectiveness only reaches 50% (78).
Phenobarbital is more widely used as a first-line agent. The
usual loading dose is 20 mg/kg IV, with initial maintenance doses
of 3 to 4 mg/kg/day in two divided doses. Titration of the dose to
achieve levels of up to 40 mcg/mL may be necessary in refractory
cases; however, further control of seizures is hardly achieved
above this maximal therapeutic level (79). Common toxic effects
at this level include lethargy, central respiratory depression with
apnea, bradycardia, and hypotension. Apparent half-life after 5 to
7 days is 100 hours. It is greater in premature infants and in those
suffering from impaired hepatic metabolism and renal clearance
from hypoxic-ischemic encephalopathy.
Phenytoin is used after phenobarbital, although the prodrug
form, fosphenytoin is the preferred agent due to its better side
effect profile. A loading dose of 15 to 20 mg PE/kg IV is recom-
mended with subsequent maintenance doses of 3 to 8 mg
PE/kg/day in two to three divided doses. Average therapeutic
range is 10 to 20 mcg/mL. Because of highly variable metabo-
lism and oral absorption, half-life ranges from 40 to 100 hours.
Due to nonlinear kinetics, small dose changes near maximal
therapeutic levels can lead to acute toxicity, therefore close
monitoring and an individualized approach to therapy is neces-
sary (80–82). Common side effects include arrhythmia and
local irritation with phenytoin.
There is little consensus in the choice of AEDs to treat
seizures that are refractory to phenobarbital and phenytoin.
Medications reported to have been used in such scenarios
include benzodiazepines (79,83–88), valproic acid (89), primi-
done (90), carbamazepine (91,92), anesthetic agents (87,93–96),
and some of the newer antiepileptic agents such as lamotrigine
(97), vigabatrin (98), topiramate, and levetiracetam (99) admin-
istered nasogastrically.
Lorazepam at a dose of 0.05 mg/kg and repeated up to a total
dose of 0.15 mg/kg (85) and diazepam at 0.25 mg can be
administered as IV boluses, with subsequent additional boluses
given for persistent seizures (19). Many intensivists prefer the
use of midazolam given as an infusion of 0.05 to 0.4 mg/kg/hr
after an initial bolus of 0.1 to 0.3 mg/kg. Its short half-life lends
to ease of titration, and discontinuation of the drug once
seizures are controlled. Published data on its effectiveness show
471
contradictory results. In a nonrandomized retrospective study
of midazolam in the setting of refractory neonatal seizures,
electrographically confirmed seizures were rapidly controlled in
13 of 13 nonresponders to phenobarbital/phenytoin treated
with midazolam (88). However, a prospective trial looking at its
effectiveness as a first-line agent found it to be ineffective (87).
Valproic acid is more widely used in Europe, but because of
its potential to cause fatal hepatotoxicity in this age group, it is
rarely used in the United States. In a small study, lidocaine was
found be effective in controlling seizures in 50% of infants with
refractory seizures (89).
More recently there has been interest in the use of lidocaine
as an AED for otherwise refractory neonatal seizures. There is a
narrow therapeutic range, with reports of dose-related and
reversible cardiac arrhythmias, with infusions (95).
There is no universally accepted protocol for duration of
treatment of the acute period. Duration of treatment ranges
from 2 weeks to 1 year (100–103). Regardless of duration, most
clinicians obtain an EEG prior to gradually weaning off AEDs,
looking for persistent subclinical seizures (77).
Prognosis
Neonatal seizures are associated with significant incidence of
mortality and morbidity. Reported mortality rates range from
24% to 30%, with decreasing incidence over the years from
improved neonatal and obstetric care. The risk of neurodevel-
opmental sequelae is also high but varies depending on etiology
and degree of brain injury. Abnormalities in the neurologic
examination and developmental delay are reported in about
half of survivors. Cerebral palsy and mental retardation have
been reported in about a third (56,104,105). Prognosis was bet-
ter for term than for preterm infants (104). Epilepsy occurs in
20% to 30%, usually with onset within the first 6 to 9 months
of life (23,56,57,60,100,104).
Etiology is the most important determinant of overall out-
come in neonates who experience seizures (4,13,25,67,106–108).
Infants whose seizures are secondary to severe or diffuse brain
insults such as hypoxic-ischemic encephalopathy, intraventricu-
lar hemorrhage, or congenital cerebral malformations have
poorer prognosis compared to those whose seizures are second-
ary to transient causes, such as hypoglycemia, hypocalcemia,
and drug withdrawal. This observation holds true for the risk of
developing both neurodevelopmental sequela and postneonatal
epilepsy. The highest risk of epilepsy is reported in the setting of
cerebral dysgenesis and infections and severe intraventricular
hemorrhage (13,56,58,104).
Certain clinical characteristics of the seizures themselves
have been found to correlate with outcome. Mizrahi and
Kellaway found that focal tonic and focal clonic seizures, typi-
cally associated with focal brain lesion, predicted better prog-
nosis. Myoclonic seizures, motor automatisms, and generalized
tonic posturing on the other hand were more likely to be asso-
ciated with diffuse brain involvement and a worse prognosis
(109). These observations have been supported by later investi-
gators (104,105).
The duration of seizures and their responsiveness to medica-
tions also determine prognosis. Seizures that are prolonged
472 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
(56,74), frequent (110), or require multiple drugs to control
(104) have been associated with poorer prognosis.
EEG can have prognostic significance as well. As mentioned
earlier, the presence EEG background abnormalities, is associ-
ated with poorer outcome. A normal EEG or mildly abnormal
EEG, on the other hand, predicted normal neurologic develop-
ment in more than 80% in a recent study (58). The presence of
interictal epileptiform activity is not as clearly related to out-
come or postneonatal seizures. Rowe and coworkers found
epileptiform abnormalities on interictal EEG to correlate with
adverse outcomes, but not to be as significant a predictor of out-
come as background rhythms (111). Ortibus found the number
of negative sharp waves in the temporal region to correlate with
both outcome and postneonatal seizures (56). In Clancy’s study,
PNE occurred in only 27% (3 of 11) of patients without spikes
or sharp waves on postnatal EEGs performed at the age of 3
months but in 100% (3 of 3) of patients with spikes or sharp
waves (60). Electroencephalographic seizures, whether associ-
ated with clinical manifestations have been correlated with
higher mortality (up to 50% in a study by Scher and coworkers)
and increased risk of developmental abnormalities (23,111).
NEONATAL EPILEPSY SYNDROMES
Four neonatal onset seizure disorders are recognized by the
International League Against Epilepsy (ILAE). Two of these are
relatively benign syndromes characterized by transient seizures
and good prognosis neurodevelopmental outcome. The two
others do not have similar favorable prognosis and present with
seizures that persist beyond the neonatal period and progressive
neurologic deterioration.
Benign Syndromes
Benign Familial Neonatal Seizures (BFNS)
As its name implies, the syndrome of benign familial neonatal
seizures (BFNS) is a familial epileptic disorder that occurs in the
neonatal period, and runs a benign course with good prognosis.
The name was coined by Bjerre and Corelius in 1968 (112); how-
ever, the first familial case was described by Rett and Teubel 4
years earlier (113). It was first included in the ILAE classification
of epilepsy as a syndrome in 1989. In recent years, it has been
identified as a CNS channelopathy caused by mutations in the
genes encoding potassium channels (114–118). BFNS is a rare
disorder that may be under recognized. Ronen and colleagues
found 5 familial cases out of 34,615 live births in their popula-
tion-based study of neonatal seizures in Newfoundland (104).
Seizures in affected infants begin by 2 to 8 days of life, with
80% appearing on the second or third day (19). Onset of seizures
can be delayed up to 3.5 months especially in premature infants
(104,119). The seizures that are brief, typically 1 to 2 minutes, can
occur multiple times a day during the vulnerable period, with
most seizures remitting by 16 months of life (104,119,120).
The diagnosis is suspected when typical seizures occur
without obvious precipitants in an otherwise normal newborn
with a family history of similar seizures in the neonatal period.
It requires exclusion of other causes of seizures by normal
maternal and delivery history and laboratory and neuroimag-
ing studies (121).
Clinical seizures are described to have clonic or tonic com-
ponents and can be partial or generalized with asymmetric
motor activity. Myoclonic seizures or spasms have not been
described. Apnea is a frequent component (122). Seizures
described in kindred of 69 affected individuals consisted of
mixed seizure types, starting with tonic posture, ocular symp-
toms, apnea, and other autonomic features, and progressing to
clonic movements and motor automatisms. The ictal EEG pat-
tern was characterized by generalized suppression of amplitude
on onset (123). Hirsch also reported generalized onset with
bilateral, symmetrical flattening of the EEG in three infants.
This was associated with simultaneous tonic motor activity and
apnea, and followed by bilateral discharge with associated
vocalizations, chewing, and focal or generalized clonic activity
(124). However, Bye recorded focal onset seizures with and
without secondary generalization in one neonate (125), and
Plouin and Anderson suggested that the electroclinical presen-
tation is closer to partial epilepsies (121).
The interictal EEG is nonspecific and may be normal or
abnormal (119). Some abnormalities reported in the literature
include interictal sharp transients, the theta pointu alternant
pattern (126–129). Centrotemporal spikes have also been
reported in one patient who later on developed one clinical
rolandic seizure at the age of 4 years (130).
Prognosis is generally good with resolution of seizures by 1
to 6 months and normal neurologic outcome in the majority of
cases. Mental retardation and learning disabilities do not occur
at an increased rate compared to the general population (120).
However, epilepsy develops in 11% to 14% of cases with 5%
risk of febrile seizures (121,122).
Seizures are generally treated with antiepileptics because of
their frequency even though they are generally self-limited.
Phenobarbital is most commonly used although phenytoin and
valproate have been used in refractory cases (121). Novel
approaches to treatment with antiepileptic and anti-inflamma-
tory medications that target KCNQ2/KCNQ3 channels are
being investigated in animal models (131–133).
Benign Neonatal Convulsions (BNCs)
Navelet and Dehan were the first to describe nonfamilial cases of
self-limited idiopathic seizures in the newborn that ran a benign
course (134). It was later recognized as an epileptic syndrome by
the ILAE and given the name “benign neonatal convulsions.”
Emphasizing its unclear etiology, the term “benign idiopathic
neonatal convulsions” has also been used in the literature. Because
of the tendency of the seizures to occur on the fourth or fifth day
of life, the term “fifth day fits” has also been commonly used.
Although the exact incidence is unknown, BNCs are sug-
gested to represent 2% to 7% of seizures in the neonate (121).
Affected infants are usually term or near term and are products
of uneventful pregnancies and deliveries. Males are more
affected than females (121).
Like the familial form, its diagnosis requires the exclusion of
other transient causes for seizures that present with the typical
 Chapter 25 ■ Seizures and Epilepsies in the Preterm and Term Neonate
EEG abnormalities, such as hypocalcemia, subarachnoid hemor-
rhage, and neonatal meningitis (121). Plouin proposed the fol-
lowing diagnostic clinical criteria: (i) Apgar score greater than
seven at 1 minute, (ii) typical interval between birth and
seizures onset (4 to 6 days), (iii) normal neurologic examination
before seizures and interictally, (iv) normal laboratory findings
(metabolic studies, neuroimaging, and CSF analysis), (v) no
family history of neonatal seizures or postneonatal epilepsy
(135). The final criterion distinguishes this syndrome from the
familial form.
The etiology and pathogenesis of the seizures are unknown.
One intriguing study found rotavirus in the stool of 95% of 19
“fifth day fitters,” compared to 40% in the control group (136).
Another study found low levels of zinc in the CSF of affected
infants (43). In four cases, a mutation in the gene encoding for
KCNQ2 (a mutation frequently found in the familial form of
BNCs) was identified (117). However, no consistent links have
been found to metabolic, infectious, or genetic causes.
Typical seizures in affected infants consist of brief focal,
clonic activity typically lasting 1 to 3 minutes. The seizures tend
to occur in clusters beginning on the fourth or seventh day of
life. In 80% of cases, the seizures occur on the fourth or fifth
day. Multiple clusters of seizure can occur in a day and can lead
to status epilepticus. Average duration of seizures is 20 hours.
Although the syndrome is classified as a generalized epilepsy
syndrome, focal clonic seizures are more common (121,137).
These may or may not be associated with apnea. Focal tonic
seizures are rare although Guerra reported one patient with
fifth day fits presenting with generalized tonic–clonic seizures
and generalized electrical discharges on video-EEG (138). Ictal
EEG patterns are nonspecific and can be focal or more general-
ized, characterized by rhythmic spikes or rhythmic slow waves.
The seizure discharge can arise from any area but is most com-
monly seen in the rolandic region (121).
The interictal EEG is more characteristic, although nonspe-
cific. Plouin reviewed EEG patterns described in 101 of 278
cases (121). In 60% of cases, a theta pointu alternant pattern
was identified consisting of discontinuous, nonreactive, 4 to
7 Hz theta activity that frequently alternates between hemi-
spheres and is intermixed with sharp activities. The pattern can
be present in wakefulness as well as quiet sleep. This is clearly an
interictal pattern as it can persist days after cessation of clinical
seizures. Other abnormal patterns that have been described
include centrotemporal sharps, discontinuous patterns, and
focal or multifocal nonspecific abnormalities. The EEG is nor-
mal in 10% of patients (121).
The course of BNC is considered generally benign, although
favorable outcome needs confirmation by more extensive stud-
ies (121). Seizures usually remit by 48 hours of onset (139).
Some argue against treating these self-limited seizures, but
most clinicians treat acute seizures with antiepileptics––usually
phenobarbital and phenytoin. Although the exact incidence of
postneonatal epilepsy is unknown, seizures rarely occur later in
life. Mild psychomotor retardation has been described in some
survivors; however, prognosis for neurodevelopmental out-
come is considered to be generally favorable (121).
473
Catastrophic Epilepsy Syndromes
Two so-called catastrophic epilepsy syndromes occur in the
neonatal period and have been classified as generalized sympto-
matic epilepsies by the ILAE. Unlike the previously discussed
benign disorders, these epilepsy syndromes are invariably associ-
ated with abnormal mentation, psychomotor retardation, and
persistent, difficult to control seizures, often consisting of various
types. It is thought that in these syndromes, the epileptiform
abnormalities themselves, and not just the underlying disorder,
contribute to the eventual neurologic deterioration. Thus, they
join the syndromes of infantile spasms, Lennox–Gastaut, Dravet,
and electrical status epilepticus during sleep under the category
epileptic encephalopathies.
Early Myoclonic Encephalopathy
The syndrome was first described by Aicardi and Goutieres
in 1978, who used the term neonatal myoclonic encephalopa-
thy (140). Other subsequent nomenclatures have included
“myoclonic encephalopathy with neonatal onset” (141), “neona-
tal epileptic encephalopathy” (142), and “early myoclonic epilep-
tic encephalopathy” (143).
In this syndrome, erratic and fragmentary myoclonic seizures
occur early on, often within the first few days or months of life.
Myoclonic movements can affect the face, proximal, or distal
limb muscles. The jerks can be restricted to a few muscles or can
affect multiple muscles asynchronously, shifting erratically from
one area to another. Generalized massive myoclonus is less com-
mon. Frequency of seizures is variable but can be almost contin-
uous. Partial seizures follow the appearance of the erratic
myoclonus. These seizures can be characterized by focal clonic
movements of a limb or subtle eye-deviation with respiratory
changes. Tonic spasms are the last type of seizures to appear,
usually between 3 to 4 months, although earlier onset has been
described.
The interictal EEG is highly abnormal and is characterized
by brief, 1 to 3 seconds, bursts of high amplitude slow waves
with intermixed spikes and sharp waves. Burst alternate with
periods of suppression that can last up to 10 seconds. This burst
suppression pattern is more pronounced in sleep and can later
on evolve into atypical hypsarrhythmia. The EEG can be nor-
mal at the onset of the seizures, and repetitive EEGs may be
necessary for the diagnosis (144). Partial seizures are often
accompanied by electrographic seizure activity but the erratic
myoclonus usually has no EEG correlate.
The neurologic examination is abnormal from the onset and
affected infants are often lethargic, hypotonic, and have psy-
chomotor delays. Developmental regression has also been
described (143). The mortality rate is high (up to 50%) and
many die within the first 2 years of life (145).
EME is a rare disorder. Of 2220 cases of epilepsy in children
under 10 years of age, 2 were classified as having EME (preva-
lence rate of 0.1%) by Oka and colleagues in their population-
based survey in Okayama, Japan (146). Various etiologies have
been associated with this syndrome, suggesting that EME may
represent a broad spectrum of clinical entities (147). Inborn
errors of metabolism are commonly reported. Cases associated
474 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
with nonketotic hyperglycinemia (145,147–149), methyl-
malonic acidemia (148), propionic acidemia (148,150), molyb-
denum cofactor deficiency (151), and pyridoxine-dependency
(147) have been reported. Structural brain abnormalities have
also been described but are less common than in the other cat-
astrophic neonatal epilepsy syndrome––early infantile epileptic
encephalopathy (152–154).
Early Infantile Epileptic Encephalopathy
Also known as Ohtahara syndrome, after Ohtahara who first
described the disorder in 1976 (155), early infantile epileptic
encephalopathy is another catastrophic epilepsy syndrome with
neonatal onset. In fact, intrauterine onset has been suggested
(153). Certainly, it occurs very soon after birth, usually within
the first month of life (156). The diagnosis is suspected when
the typical newborn infant presents with tonic spasms and has
a suppression-burst pattern on EEG. Often, a structural brain
abnormality is found on neuroimaging studies, and the infant
does poorly with severe psychomotor retardation and refrac-
tory seizures.
The most characteristic seizures in this syndrome are tonic
spasms, which can occur up to 300 times a day (157). They can
be lateralized, or generalized but asymmetric. Other seizure
types such as partial motor seizures and hemiconvulsions
occur in about a third of patients (156). Although generalized
tonic–clonic seizures and rare myoclonic seizures can develop
later, the erratic myoclonus, which characterizes early myoclonic
encephalopathy, does not occur in this syndrome (158).
The characteristic suppression-burst pattern found on EEG
is an interictal pattern, although some have observed clinical
seizures to correlate with the bursts (159). The pattern consists
of bursts of high-voltage (150 to 350 V) slow waves intermin-
gled with spikes, alternating with periods of suppression of
electrical activity. The bursts of spikes and polyspikes can last 1
to 6 seconds, while the periods of suppression are usually 2 to 5
seconds long (145,156), although they can last up to 18 seconds
(160). The interval measured from beginning to beginning of
each burst ranges from 5 to 10 seconds. The pattern can involve
only one hemisphere, or be more widespread, involving both
hemispheres symmetrically or asymmetrically. Asynchrony is
not common, although has been described in Aicardi syn-
drome (156). Focal epileptiform discharges have also been
described during the periods of suppression (161). Ohtahara
stressed that the distinguishing features of suppression-burst
pattern in Ohtahara syndrome were its consistent appearance
in both waking and sleeping states and regular appearance of
periodicity (157). This is in contrast to the suppression-burst
pattern in EME, which can be present only during sleep, or be
enhanced by it.
The suppression-burst pattern can evolve into hypsarrhyth-
mia between 3 and 6 months of age, coincident with the devel-
opment of more typical infantile spasms. This occurred in 8 of
16 cases followed by Ohtahara (156,157,162). Later transition
into slow-spike and wave morphology (characteristic of
Lennox–Gastaut syndrome) occurred in two survivors in the
group. This early transition into hypsarrhythmia is less seen in
EME wherein the suppression-burst pattern can persist into
childhood after a transient evolution into hypsarrhythmia in
middle to late infancy (154,157).
EEG during tonic spasms principally shows desynchroniza-
tion with or without evident rapid activity (156). Using video-
EEG with polymyographic recordings from deltoid muscles,
Fusco and coworkers found each burst to be associated with
tonic contraction of variable duration in all but one of the eight
infants in their study (159). They also recorded partial seizures
clinically consisting of tonic eye-deviation, unilateral clonic
contractions, or subtle or subclinical phenomena. There was no
particular localization of ictal discharges on the EEG. Yamatogi
described partial seizures to arise from a fixed focus and to be
followed by tonic spasms in series (156).
Both congenital and acquired structural brain lesions have
been described in affected infants. In Ohtahara’s original series
of 16 cases, 11 were found to have structural abnormalities. Four
had cerebral dysgenesis, two had Aicardi syndrome, and there
were one case each of hemimegalencephaly, lissencephaly, poren-
cephaly, hydrocephalus, and subacute diffuse encephalopathy
(157). Schlumberger identified six cases of hemimegalen-
cephaly in his series of eight patients. One had Aicardi syn-
drome and another had dentato-olivary dysplasia (158). Other
associated brain malformations reported in the literature
include diffuse cerebral migrational disorders (163) and
polymicrogyria (164).
Metabolic causes are rare but single cases of Leigh
encephalopathy (165), nonketotic hyperglycinemia (160),
cytochrome c oxidase deficiency (166), pyridoxine-dependency,
and carnitine palmitoyltransferase deficiency (159) have been
reported.
Seizures are typically refractory to conventional anticonvul-
sant medications, and affected infants generally do poorly.
Ohtahara reported ACTH to be effective in two cases in the
acute period and in four cases during the transition to West syn-
drome (157). However, ACTH used in four of Clarke’s cases was
not effective (160). He reported only transient improvement of
seizures in two out of three infants given the ketogenic diet, and
in three out of four given chloral hydrate. Cases responding to
vitamin B6 (159), clonazepam and acetazolamide (157), zon-
isamide (156,167), and vigabatrin (98) have been described.
Surgical treatment with hemispherectomy or focal resection of
lesions has been reported to improve seizures in patients with
megalencephaly and cortical dysplasias (168–170).
Treatment of the seizures does not appear to prevent relent-
less neurologic deterioration. Subsequent brain atrophy is fre-
quently found on imaging studies. In 16 cases followed by
Ohtahara in 1992, all 8 survivors were profoundly mentally
retarded and physically handicapped. Two had persistent
seizures and six had persistent epileptiform EEG abnormalities
at time of follow-up (156).
REFERENCES
1. Burke JB. The prognostic significance of neonatal convulsions.
Arch Dis Child. 1954;29(146):342–345.
2. Craig WS. Convulsive movements occurring in the first 10 days of
life. Arch Dis Child. 1960;35:336–344.
 Chapter 25 ■ Seizures and Epilepsies in the Preterm and Term Neonate
3. Keen JH. Significance of hypocalcaemia in neonatal convulsions.
Arch Dis Child. 1969;44(235):356–361.
4. McInerny TK, Schubert WK. Prognosis of neonatal seizures. Am J
Dis Child. 1969;117(3):261–264.
5. Schulte FJ. Neonatal convulsions and their relation to epilepsy in
early childhood. Dev Med Child Neurol. 1966;8(4):381–392.
6. Massa T, Niedermeyer E. Convulsive disorders during the first
three months of life. Epilepsia. 1968;9(1):1–9.
7. Cadilhac J, Passouant P, Ribstein M. Convulsions in the newborn:
EEG and clinical aspects. Electroencephalogr Clin Neurophysiol. 1959;
11:604.
8. Harris R, Tizard JP. The electroencephalogram in neonatal convul-
sions. J Pediatr. 1960;57:501–520.
9. Passouant P, Cadilhac J. EEG and clinical study of epilepsy during
maturation in man. Epilepsia. 1962;3:14–43.
10. Dreyfus-Brisac C, Monod N. Electroclinical studies of status
epilepticus and convulsions in the newborn. In: Kellaway P,
Petersen I, eds. Neurological and Electroencephalographic
Correlative Studies in Infancy. New York: Grune and Stratton;
1964:250–272.
11. The EEG society: At the annual general meeting held on October 4,
1958, the following officers were elected: Chairman: Dr. J. R.
Davidson; Hon. Treasurer: Dr. W. A. Cobb; Hon. Secretary: Dr. S. L.
Sherwood; Hon. Foreign Secretary: Dr. W. Grey Walter. The follow-
ing were elected to the council: Dr. J. A. V. Bates, Dr. M. Jefferson,
Dr. P. Bradley, Dr. G. Pampliglione, Dr. G. D. Dawson, Dr. J. A.
Simpson. Summaries of papers read at the centres in London and
Edinburgh during 1958 Proceedings Secretary: Dr. G. Parr 68,
Compton road, Winchmore hill, London, N. 21. Electroencephalogr
Clin Neurophysiol. 1959;11(3):601.
12. Ribstein M, Walter M. Convulsion in the first month of life. Rev
Neurol (Paris). 1958;99(1):91–99.
13. Rose AL, Lombroso CT. A study of clinical, pathological, and elec-
troencephalographic features in 137 full-term babies with a long-
term follow-up. Pediatrics. 1970;45(3):404–425.
14. Lombroso CT. Seizures in the newborn. In: Vinken PJ, Bruyn GW,
eds. Handbook of Clinical Neurology. New York: Elsevier; 1974:
189–218.
15. Volpe JJ. Neonatal seizures. N Engl J Med. 1973;289(8):413–416.
16. Mizrahi EM. Electroencephalographic/polygraphic/video moni-
toring in childhood epilepsy. J Pediatr. 1984;105(1):1–9.
17. Mizrahi EM, Kellaway P. Characterization and classification of
neonatal seizures. Neurology. 1987;37(12):1837–1844.
18. Kellaway P, Hrachovy RA. Electroencephalography. In: Swainman
KF, Wright FS, eds. The Practice of Pediatric Neurology. 2nd ed. St.
Louis: Mosby; 1982:96.
19. LaJoie J, Moshe SL. Neonatal seizures and neonatal epilepsy
syndromes. In: Devinsky O, Westbrook LE, eds. Epilepsy and
Developmental Disabilities. Boston: Butterworth-Heinemann; 2002.
20. Lanska MJ, Lanska DJ, Baumann RJ, et al. A population-based
study of neonatal seizures in Fayette County, Kentucky. Neurology.
1995;45(4):724–732.
21. Ronen GM, Penney S, Andrews W. The epidemiology of clinical
neonatal seizures in Newfoundland: a population-based study.
J Pediatr. 1999;134(1):71–75.
22. Saliba RM, Annegers JF, Waller DK, et al. Incidence of neonatal
seizures in Harris County, Texas, 1992–1994. Am J Epidemiol. 1999;
150(7):763–769.
23. Scher MS, Aso K, Beggarly ME, et al. Electrographic seizures in
preterm and full-term neonates: clinical correlates, associated brain
lesions, and risk for neurologic sequelae. Pediatrics. 1993;91(1):
128–134.
475
24. Holden KR, Mellits ED, Freeman JM. Neonatal seizures. I.
Correlation of prenatal and perinatal events with outcomes.
Pediatrics. 1982;70(2):165–176.
25. Mizrahi EM, Kellaway P. Diagnosis and Management of Neonatal
Seizures. Philadelphia: Lippincott-Raven; 1998.
26. Kellaway P, Hrachovy RA. Status epilepticus in newborns: a per-
spective on neonatal seizures. In: Delgado-Escueta AV, Wasterlain
CG, Treiman DM, et al., eds. Status Epilepticus. New York: Raven
Press; 1983:93.
27. Swann JW, Moshe SL. Developmental issues in animal models. In:
Engel J, Pedley TA, eds. 2nd ed. Epilepsy: A Comprehensive Textbook.
New York: Raven Press; 1997:467–480.
28. Sperber EF, Moshe SL. The effects of seizures on the hippocampus
of the immature brain. In: Engel J, Morrell F, eds. Brain Plasticity
and Epilepsy: The Legacy of Frank Morell. San Diego: Academic
Press; 2001.
29. Moshe SL, Sperber EF, Velisek L. Critical issues of developmental
seizure disorders. Physiol Res. 1993;42(3):145–154.
30. Moshe SL. Epileptogenesis and the immature brain. Epilepsia.
1987;28(suppl 1):S3–S15.
31. Moshe SL, Albala BJ. Maturational changes in postictal refractori-
ness and seizure susceptibility in developing rats. Ann Neurol. 1983;
13(5):552–557.
32. Moshe SL, Albala BJ, Ackermann RF, et al. Increased seizure
susceptibility of the immature brain. Brain Res. 1983;283(1):
81–85.
33. Mueller AL, Chesnut RM, Schwartzkroin PA. Actions of GABA in
developing rabbit hippocampus: an in vitro study. Neurosci Lett.
1983;39(2):193–198.
34. Ben-Ari Y, Khazipov R, Leinekugel X, et al. GABAA, NMDA and
AMPA receptors: a developmentally regulated ‘menage a trois’.
Trends Neurosci. 1997;20(11):523–529.
35. Michelson HB, Wong RK. Excitatory synaptic responses mediated
by GABAA receptors in the hippocampus. Science. 1991;253(5026):
1420–1423.
36. Scantlebury MH, Ouellet PL, Psarropoulou C, et al. Freeze lesion-
induced focal cortical dysplasia predisposes to atypical hyperther-
mic seizures in the immature rat. Epilepsia. 2004;45(6):592–600.
37. Galanopoulou AS. Dissociated gender-specific effects of recurrent
seizures on GABA signaling in CA1 pyramidal neurons: role of
GABA(A) receptors. J Neurosci. 2008;28(7):1557–1567.
38. Galanopoulou AS. GABA receptors as broadcasters of sexually differ-
entiating signals in the brain. Epilepsia. 2005;46(suppl 5):107–112.
39. Volpe JJ. Neonatal seizures: current concepts and revised classifica-
tion. Pediatrics. 1989;84(3):422–428.
40. Murray DM, Boylan GB, Ali I, et al. Defining the gap between elec-
trographic seizure burden, clinical expression and staff recogni-
tion of neonatal seizures. Arch Dis Child Fetal Neonatal Ed. 2008;
93(3):F187–F191.
41. Devinsky O, Kelley K, Porter RJ, et al. Clinical and electroen-
cephalographic features of simple partial seizures. Neurology. 1988;
38(9):1347–1352.
42. Watanabe K, Hara K, Miyazaki S, et al. Electroclinical studies of
seizures in the newborn. Folia Psychiatr Neurol Jpn. 1977;31(3):
383–392.
43. Goldberg HJ, Sheehy EM. Fifth day fits: an acute zinc deficiency
syndrome? Arch Dis Child. 1982;57(8):633–635.
44. Lou HC, Friis-Hansen B. Arterial blood pressure elevations during
motor activity and epileptic seizures in the newborn. Acta Paediatr
Scand. 1979;68(6):803–806.
45. Watanabe K, Hara K, Miyazaki S, et al. Apneic seizures in the new-
born. Am J Dis Child. 1982;136(11):980–984.
476 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
46. Fenichel GM, Olson BJ, Fitzpatrick JE. Heart rate changes in con-
vulsive and nonconvulsive neonatal apnea. Ann Neurol. 1980;
7(6):577–582.
47. Knauss TA, Carlson CB. Neonatal paroxysmal monorhythmic
alpha activity. Arch Neurol. 1978;35(2):104–107.
48. Willis J, Gould JB. Periodic alpha seizures with apnea in a new-
born. Dev Med Child Neurol. 1980;22(2):214–222.
49. Clancy RR, Legido A, Lewis D. Occult neonatal seizures. Epilepsia.
1988;29(3):256–261.
50. Bye A, Flanagan D. Electroencephalograms, clinical observations
and the monitoring of neonatal seizures. J Paediatr Child Health.
1995;31(6):503–507.
51. Mizrahi EM, Hrachovy RA, Kellaway P. Atlas of Neonatal
Electroencephalography. 3rd ed. Philadelphia: Lippincott Williams
& Wilkins; 2003.
52. Scher MS, Bova JM, Dokianakis SG, et al. Positive temporal sharp
waves on EEG recordings of healthy neonates: a benign pattern of
dysmaturity in pre-term infants at post-conceptional term ages.
Electroencephalogr Clin Neurophysiol. 1994;90(3):173–178.
53. Cukier F, Andre M, Monod N, et al. Contribution of EEG to the
diagnosis of intraventricular hemorrhages in the premature infant.
Rev Electroencephalogr Neurophysiol Clin. 1972;2(3):318–322.
54. Clancy RR, Tharp BR. Positive rolandic sharp waves in the elec-
troencephalograms of premature neonates with intraventricular
hemorrhage. Electroencephalogr Clin Neurophysiol. 1984;57(5):
395–404.
55. Novotny EJ Jr, Tharp BR, Coen RW, et al. Positive rolandic sharp
waves in the EEG of the premature infant. Neurology. 1987;37(9):
1481–1486.
56. Ortibus EL, Sum JM, Hahn JS. Predictive value of EEG for out-
come and epilepsy following neonatal seizures. Electroencephalogr
Clin Neurophysiol. 1996;98(3):175–185.
57. Bye AM, Cunningham CA, Chee KY, et al. Outcome of neonates
with electrographically identified seizures, or at risk of seizures.
Pediatr Neurol. 1997;16(3):225–231.
58. Tekgul H, Gauvreau K, Soul J, et al. The current etiologic profile
and neurodevelopmental outcome of seizures in term newborn
infants. Pediatrics. 2006;117(4):1270–1280.
59. Laroia N, Guillet R, Burchfiel J, et al. EEG background as predictor
of electrographic seizures in high-risk neonates. Epilepsia. 1998;
39(5):545–551.
60. Clancy RR, Legido A. Postnatal epilepsy after EEG-confirmed
neonatal seizures. Epilepsia. 1991;32(1):69–76.
61. Co JP, Elia M, Engel J Jr, et al. Proposal of an algorithm for diag-
nosis and treatment of neonatal seizures in developing countries.
Epilepsia. 2007;48(6):1158–1164.
62. Nelson KB, Broman SH. Perinatal risk factors in children with seri-
ous motor and mental handicaps. Ann Neurol. 1977;2(5):371–377.
63. Leviton A, Nelson KB. Problems with definitions and classifications
of newborn encephalopathy. Pediatr Neurol. 1992;8(2):85–90.
64. Perlman JM. Summary proceedings from the neurology group on
hypoxic-ischemic encephalopathy. Pediatrics. 2006;117(3 pt 2):
S28–S33.
65. Clancy RR, Sharif U, Ichord R, et al. Electrographic neonatal
seizures after infant heart surgery. Epilepsia. 2005;46(1):84–90.
66. Parker S, Zuckerman B, Bauchner H, et al. Jitteriness in full-term
neonates: prevalence and correlates. Pediatrics. 1990;85(1):17–23.
67. Volpe JJ. Neurology of the Newborn. Philadelphia: W.B. Saunders;
2001.
68. Coulter DL, Allen RJ. Benign neonatal sleep myoclonus. Arch
Neurol. 1982;39(3):191–192.
69. Giacoia GP, Ryan SG. Hyperekplexia associated with apnea and
sudden infant death syndrome. Arch Pediatr Adolesc Med. 1994;
148(5):540–543.
70. Zhou L, Chillag KL, Nigro MA. Hyperekplexia: a treatable neuro-
genetic disease. Brain Dev. 2002;24(7):669–674.
71. Tijssen MA, Vergouwe MN, van Dijk JG, et al. Major and minor
form of hereditary hyperekplexia. Mov Disord. 2002;17(4):
826–830.
72. Bakker MJ, van Dijk JG, van den Maagdenberg AM, et al. Startle
syndromes. Lancet Neurol. 2006;5(6):513–524.
73. Andrew M, Owen MJ. Hyperekplexia: abnormal startle response
due to glycine receptor mutations. Br J Psychiatry. 1997;170:
106–108.
74. Pisani F, Cerminara C, Fusco C, et al. Neonatal status epilepticus
vs. recurrent neonatal seizures: clinical findings and outcome.
Neurology. 2007;69(23):2177–2185.
75. Heimann G. Enteral absorption and bioavailability in children in
relation to age. Eur J Clin Pharmacol. 1980;18(1):43–50.
76. Guillet R, Morselli PL. Pharmacokinetics of anticonvulsants in the
neonate. In: Wasterlain CG, Vert P, eds. Neonatal Seizures. New
York: Raven Press, Ltd.; 1990:257–267.
77. Boer HR, Gal P. Neonatal seizures: a survey of current practice.
Clin Pediatr (Phila). 1982;21(8):453–457.
78. Painter MJ, Scher MS, Stein AD, et al. Phenobarbital compared
with phenytoin for the treatment of neonatal seizures. N Engl J
Med. 1999;341(7):485–489.
79. Gal P, Toback J, Boer HR, et al. Efficacy of phenobarbital monother-
apy in treatment of neonatal seizures––relationship to blood levels.
Neurology. 1982;32(12):1401–1404.
80. Dodson WE. Special pharmacokinetic considerations in children.
Epilepsia. 1987;28(suppl 1):S56–S70.
81. Dodson WE. Antiepileptic drug utilization in pediatric patients.
Epilepsia. 1984;25(suppl 2):S132–S139.
82. Bourgeois BF, Dodson WE. Phenytoin elimination in newborns.
Neurology. 1983;33(2):173–178.
83. Andre M, Boutroy MJ, Bianchetti G, et al. Clonazepam in neona-
tal seizures: dose regimens and therapeutic efficacy. Eur J Clin
Pharmacol. 1991;40(2):193–195.
84. Sheth RD, Buckley DJ, Gutierrez AR, et al. Midazolam in the treat-
ment of refractory neonatal seizures. Clin Neuropharmacol. 1996;
19(2):165–170.
85. Maytal J, Novak GP, King KC. Lorazepam in the treatment of
refractory neonatal seizures. J Child Neurol. 1991;6(4):319–323.
86. Claassen J, Hirsch LJ, Emerson RG, et al. Continuous EEG moni-
toring and midazolam infusion for refractory nonconvulsive sta-
tus epilepticus. Neurology. 2001;57(6):1036–1042.
87. Boylan GB, Rennie JM, Chorley G, et al. Second-line anticonvul-
sant treatment of neonatal seizures: a video-EEG monitoring
study. Neurology. 2004;62(3):486–488.
88. Castro Conde JR, Hernandez Borges AA, Domenech Martinez E,
et al. Midazolam in neonatal seizures with no response to pheno-
barbital. Neurology. 2005;64(5):876–879.
89. Gal P, Oles KS, Gilman JT, et al. Valproic acid efficacy, toxicity, and
pharmacokinetics in neonates with intractable seizures. Neurology.
1988;38(3):467–471.
90. Sapin JI, Riviello JJ Jr, Grover WD. Efficacy of primidone for
seizure control in neonates and young infants. Pediatr Neurol.
1988;4(5):292–295.
91. MacKintosh DA, Baird-Lampert J, Buchanan N. Is carbamazepine
an alternative maintenance therapy for neonatal seizures? Dev
Pharmacol Ther. 1987;10(2):100–106.
 Chapter 25
92. Hoppen T, Elger CE, Bartmann P. Carbamazepine in phenobar-
bital-nonresponders: experience with ten preterm infants. Eur J
Pediatr. 2001;160(7):444–447.
93. Koren G, Butt W, Rajchgot P, et al. Intravenous paraldehyde for
seizure control in newborn infants. Neurology. 1986;36(1):108–111.
94. Hellstrom-Westas L, Svenningsen NW, Westgren U, et al.
Lidocaine for treatment of severe seizures in newborn infants. II.
Blood concentrations of lidocaine and metabolites during intra-
venous infusion. Acta Paediatr. 1992;81(1):35–39.
95. Hellstrom-Westas L, Westgren U, Rosen I, et al. Lidocaine for
treatment of severe seizures in newborn infants. I. Clinical effects
and cerebral electrical activity monitoring. Acta Paediatr Scand.
1988;77(1):79–84.
96. Malingre MM, Van Rooij LG, Rademaker CM, et al. Development
of an optimal lidocaine infusion strategy for neonatal seizures.
Eur J Pediatr. 2006;165(9):598–604.
97. Barr PA, Buettiker VE, Antony JH. Efficacy of lamotrigine in
refractory neonatal seizures. Pediatr Neurol. 1999;20(2):161–163.
98. Baxter PS, Gardner-Medwin D, Barwick DD, et al. Vigabatrin
monotherapy in resistant neonatal seizures. Seizure. 1995;4(1):
57–59.
99. Silverstein FS, Ferriero DM. Off-label use of antiepileptic drugs
for the treatment of neonatal seizures. Pediatr Neurol. 2008;
39(2):77–79.
100. Hodson A. A case for long-term treatment with anticonvulsants.
Pharmacotherapy. 1985;5(5):274–277.
101. Gal P. Anticonvulsant therapy after neonatal seizures––how long
should it be continued? I. A case for early discontinuation of
anticonvulsants. Pharmacotherapy. 1985;5(5):268–273.
102. Hellstrom-Westas L, Blennow G, Lindroth M, et al. Low risk of
seizure recurrence after early withdrawal of antiepileptic treat-
ment in the neonatal period. Arch Dis Child Fetal Neonatal Ed.
1995;72(2):F97–F101.
103. Guillet R, Kwon JM. Prophylactic phenobarbital administration
after resolution of neonatal seizures: survey of current practice.
Pediatrics. 2008;122(4):731–735.
104. Ronen GM, Buckley D, Penney S, et al. Long-term prognosis in
children with neonatal seizures: a population-based study.
Neurology. 2007;69(19):1816–1822.
105. Brunquell PJ, Glennon CM, DiMario FJ Jr, et al. Prediction of
outcome based on clinical seizure type in newborn infants.
J Pediatr. 2002;140(6):707–712.
106. Hill A, Volpe JJ. Seizures, hypoxic-ischemic brain injury, and
intraventricular hemorrhage in the newborn. Ann Neurol. 1981;
10(2):109–121.
107. Bergman I, Painter MJ, Hirsch RP, et al. Outcome in neonates
with convulsions treated in an intensive care unit. Ann Neurol.
1983;14(6):642–647.
108. Lombroso CT. Prognosis in neonatal seizures. Adv Neurol. 1983;
34:101–113.
109. Mizrahi EM. Neonatal seizures: the need for precise clinical char-
acterization. Ann Neurol. 1987;21(3):312–313.
110. Legido A, Clancy RR, Berman PH. Neurologic outcome after
electroencephalographically proven neonatal seizures. Pediatrics.
1991;88(3):583–596.
111. Rowe JC, Holmes GL, Hafford J, et al. Prognostic value of the
electroencephalogram in term and preterm infants following
neonatal seizures. Electroencephalogr Clin Neurophysiol. 1985;
60(3):183–196.
112. Bjerre I, Corelius E. Benign familial neonatal convulsions. Acta
Paediatr Scand. 1968;57(6):557–561.
■ Seizures and Epilepsies in the Preterm and Term Neonate 477
113. Rett A, Teubel R. Neugeborenen krampfe im rahmen einer
epileptisch belasten familie. Wiener Klinische Wochenschr.
1964;76:609–613.
114. Leppert M, Singh N. Benign familial neonatal epilepsy with
mutations in two potassium channel genes. Curr Opin Neurol.
1999;12(2):143–147.
115. Leppert M, Anderson VE, Quattlebaum T, et al. Benign familial
neonatal convulsions linked to genetic markers on chromosome
20. Nature. 1989;337(6208):647–648.
116. Singh NA, Charlier C, Stauffer D, et al. A novel potassium chan-
nel gene, KCNQ2, is mutated in an inherited epilepsy of new-
borns. Nat Genet. 1998;18(1):25–29.
117. Claes LR, Ceulemans B, Audenaert D, et al. De novo KCNQ2
mutations in patients with benign neonatal seizures. Neurology.
2004;63(11):2155–2158.
118. Ryan SG, Wiznitzer M, Hollman C, et al. Benign familial neona-
tal convulsions: evidence for clinical and genetic heterogeneity.
Ann Neurol. 1991;29(5):469–473.
119. Miles DK, Holmes GL. Benign neonatal seizures. J Clin
Neurophysiol. 1990;7(3):369–379.
120. Zonana J, Silvey K, Strimling B. Familial neonatal and infantile
seizures: an autosomal-dominant disorder. Am J Med Genet.
1984;18(3):455–459.
121. Plouin P, Anderson VE. Benign familial and nonfamilial neonatal
seizures. In: Roger J, Bureau M, Dravet C, et al., eds. Epileptic
Syndromes in Infancy, Childhood and Adolescence. 4th ed.
Montrouge, France: John Libbey Eurotext Ltd.; 2005:3–15.
122. Proposal for classification of epilepsies and epileptic syndromes.
commission on classification and terminology of the interna-
tional league against epilepsy. Epilepsia. 1985;26(3):268–278.
123. Ronen GM, Rosales TO, Connolly M, et al. Seizure characteristics
in chromosome 20 benign familial neonatal convulsions.
Neurology. 1993;43(7):1355–1360.
124. Hirsch E, Velez A, Sellal F, et al. Electroclinical signs of
benign neonatal familial convulsions. Ann Neurol. 1993;34(6):
835–841.
125. Bye AM. Neonate with benign familial neonatal convulsions:
recorded generalized and focal seizures. Pediatr Neurol. 1994;
10(2):164–165.
126. Dobrescu O, Larbrisseau A. Benign familial neonatal convul-
sions. Can J Neurol Sci. 1982;9(3):345–347.
127. Camfield PR, Dooley J, Gordon K, et al. Benign familial neonatal
convulsions are epileptic. J Child Neurol. 1991;6(4):340–342.
128. Pettit RE, Fenichel GM. Benign familial neonatal seizures. Arch
Neurol. 1980;37(1):47–48.
129. Cronin C, Cosgrove J. Benign familial neonatal convulsions. Ir
Med J. 1986;79(11):325–326.
130. Maihara T, Tsuji M, Higuchi Y, et al. Benign familial neonatal
convulsions followed by benign epilepsy with centrotemporal
spikes in two siblings. Epilepsia. 1999;40(1):110–113.
131. Peretz A, Degani N, Nachman R, et al. Meclofenamic acid and
diclofenac, novel templates of KCNQ2/Q3 potassium channel
openers, depress cortical neuron activity and exhibit anticonvul-
sant properties. Mol Pharmacol. 2005;67(4):1053–1066.
132. Rundfeldt C, Netzer R. The novel anticonvulsant retigabine
activates M-currents in Chinese hamster ovary-cells transfected
with human KCNQ2/3 subunits. Neurosci Lett. 2000;282(1–2):
73–76.
133. Wickenden AD, Yu W, Zou A, et al. Retigabine, a novel anticon-
vulsant, enhances activation of KCNQ2/Q3 potassium channels.
Mol Pharmacol. 2000;58(3):591–600.
478 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
134. Navelet Y, D’Allest AM, Dehan M, et al. Are convulsions on the
fifth day of life a distinct clinical and electrophysiological entity?
(author’s transl.). Rev Electroencephalogr Neurophysiol Clin. 1977;
7(3):366–370.
135. Plouin P. Benign familial neonatal seizures and benign idiopathic
neonatal seizures. In: Engel J, Pedley TA, eds. Epilepsy: A
Comprehensive Textbook. Philadelphia: Lippincott-Raven
Publishers; 1997:2247–2255.
136. Herrmann B, Lawrenz-Wolf B, Seewald C, et al. 5th day convul-
sions of the newborn infant in rotavirus infections. Monatsschr
Kinderheilkd. 1993;141(2):120–123.
137. Watanabe K, Miura K, Natsume J, et al. Epilepsies of neonatal
onset: seizure type and evolution. Dev Med Child Neurol. 1999;
41(5):318–322.
138. Guerra MP, Wilson GA, Boylan GB, et al. An unusual presenta-
tion of fifth-day fits in the newborn. Pediatr Neurol. 2002;26(5):
398–401.
139. Pryor DS, Don N, Macourt DC. Fifth day fits: a syndrome of
neonatal convulsions. Arch Dis Child. 1981;56(10):753–758.
140. Aicardi J, Goutieres F. Neonatal myoclonic encephalopathy
(author’s transl.). Rev Electroencephalogr Neurophysiol Clin.
1978;8(1):99–101.
141. Cavazzuti GB, Nalin A, Ferrari F, et al. Encefalopatia epilettica ad
insorgenza neonatale. Clin Pediatr. 1978;60:239–246.
142. Martin HJ, Deroubaix-Tella P, Thelliez PH. Encéphalopathie
épileptique néonatale à bouffées périodiques. Rev EEG
Neurophysiol. 1981;11:397–403.
143. Dalla Bernardina B, Dulac O, Fejerman N, et al. Early myoclonic
epileptic encephalopathy (E.M.E.E.). Eur J Pediatr. 1983;140(3):
248–252.
144. Ozyurek H, Turanli G, Aliefendioglu D, et al. Repetitive EEG
recordings are necessary for the diagnosis of early myoclonic
encephalopathy. Neurol India. 2005;53(2):235–237.
145. Aicardi J, Ohtahara S. Severe neonatal epilepsies with suppres-
sion-burst pattern. In: Roger J, Bureau M, Dravet C, et al., eds.
Epileptic Syndromes in Infancy, Childhood and Adolescence. 4th
ed. Montrouge, France: John Libbey Eurotext Ltd.; 2005:39–50.
146. Oka E, Ishida S, Ohtsuka Y, et al. Neuroepidemiological study of
childhood epilepsy by application of international classification
of epilepsies and epileptic syndromes (ILAE, 1989). Epilepsia.
1995;36(7):658–661.
147. Wang PJ, Lee WT, Hwu WL, et al. The controversy regarding
diagnostic criteria for early myoclonic encephalopathy. Brain
Dev. 1998;20(7):530–535.
148. Lombroso CT. Early myoclonic encephalopathy, early infantile
epileptic encephalopathy, and benign and severe infantile
myoclonic epilepsies: a critical review and personal contribu-
tions. J Clin Neurophysiol. 1990;7(3):380–408.
149. Dalla Bernardina B, Aicardi J, Goutieres F, et al. Glycine
encephalopathy. Neuropadiatrie. 1979;10(3):209–225.
150. Vigevano F, Maccagnani F, Bertini E, et al. Encefalopatia mio-
clonica precoce associata ad alti livelli di acido propionico nel
siero. Boll Lega Ital Epil. 1982;39:181–182.
151. Aukett A, Bennett MJ, Hosking GP. Molybdenum co-factor defi-
ciency: an easily missed inborn error of metabolism. Dev Med
Child Neurol. 1988;30(4):531–535.
152. Spreafico R, Angelini L, Binelli S, et al. Burst suppression and
impairment of neocortical ontogenesis: electroclinical and neu-
ropathologic findings in two infants with early myoclonic
encephalopathy. Epilepsia. 1993;34(5):800–808.
153. du Plessis AJ, Kaufmann WE, Kupsky WJ. Intrauterine-onset
myoclonic encephalopathy associated with cerebral cortical dys-
genesis. J Child Neurol. 1993;8(2):164–170.
154. Djukic A, Lado FA, Shinnar S, et al. Are early myoclonic
encephalopathy (EME) and the Ohtahara syndrome (EIEE)
independent of each other? Epilepsy Res. 2006;70(suppl 1):68–76.
155. Ohtahara S, Ishida T, Oka E, et al. On the specific age dependent
epileptic syndrome: the early-infantile epileptic encephalopathy
with suppression-bursts [in Japanese]. No To Hattatsu. 1976;8:
270–280.
156. Yamatogi Y, Ohtahara S. Early-infantile epileptic encephalopathy
with suppression-bursts, Ohtahara syndrome; its overview refer-
ring to our 16 cases. Brain Dev. 2002;24(1):13–23.
157. Ohtahara S, Yamatogi Y. Ohtahara syndrome: with special refer-
ence to its developmental aspects for differentiating from early
myoclonic encephalopathy. Epilepsy Res. 2006;70(suppl 1):
S58–S67.
158. Schlumberger E, Dulac O, Plouin P. Early-infantile epileptic syn-
drome(s) with suppression-burst: nosological considerations. In:
Roger J, Bureau M, Dravet C, et al., eds. Epileptic Syndromes in
Infancy, Childhood and Adolescence. Montrouge, France: John
Libbey & Company Ltd.; 1992:35–42.
159. Fusco L, Pachatz C, Di Capua M, et al. Video/EEG aspects of
early-infantile epileptic encephalopathy with suppression-bursts
(Ohtahara syndrome). Brain Dev. 2001;23(7):708–714.
160. Clarke M, Gill J, Noronha M, et al. Early infantile epileptic
encephalopathy with suppression burst: Ohtahara syndrome.
Dev Med Child Neurol. 1987;29(4):520–528.
161. Al-Futaisi A, Banwell B, Ochi A, et al. Hidden focal EEG seizures
during prolonged suppressions and high-amplitude bursts in
early infantile epileptic encephalopathy. Clin Neurophysiol. 2005;
116(5):1113–1117.
162. Ohtahara S, Ohtsuka Y, Yamatogi Y, et al. The early-infantile
epileptic encephalopathy with suppression-burst: developmental
aspects. Brain Dev. 1987;9(4):371–376.
163. Miller SP, Dilenge ME, Meagher-Villemure K, et al. Infantile
epileptic encephalopathy (Ohtahara syndrome) and migrational
disorder. Pediatr Neurol. 1998;19(1):50–54.
164. Tominaga I, Kaihou M, Kimura T, et al. Early infantile epileptic
encephalopathy (Ohtahara syndrome) with poly-microgyria. Rev
Neurol (Paris). 1993;149(10):532–535.
165. Tatsuno M, Hayashi M, Iwamoto H, et al. Autopsy case of Leigh’s
encephalopathy with wide lesions in central nervous system and
early infantile epileptic encephalopathy with burst suppression.
No To Hattatsu. 1984;16(1):68–75.
166. Williams AN, Gray RG, Poulton K, et al. A case of Ohtahara syn-
drome with cytochrome oxidase deficiency. Dev Med Child Neurol.
1998;40(8):568–570.
167. Ohno M, Shimotsuji Y, Abe J, et al. Zonisamide treatment of early
infantile epileptic encephalopathy. Pediatr Neurol. 2000;23(4):
341–344.
168. Pedespan JM, Loiseau H, Vital A, et al. Surgical treatment of an
early epileptic encephalopathy with suppression-bursts and focal
cortical dysplasia. Epilepsia. 1995;36(1):37–40.
169. Komaki H, Sugai K, Maehara T, et al. Surgical treatment of early-
infantile epileptic encephalopathy with suppression-bursts associ-
ated with focal cortical dysplasia. Brain Dev. 2001;23(7):727–731.
170. Hmaimess G, Raftopoulos C, Kadhim H, et al. Impact of early
hemispherotomy in a case of Ohtahara syndrome with left
parieto-occipital megalencephaly. Seizure. 2005;14(6):439–442.

Seizures and Epilepsy in
Infants to Adolescents
DOUGLAS R. NORDLI, JR., JAMES J. RIVIELLO, JR., AND ERNST NIEDERMEYER
BASIC CONSIDERATIONS
Definition and Prevalence
There is no disease named “epilepsy.” Rather, epileptic seizures
are abnormal reactions of the brain caused by a large number
of diseases. The entire brain or parts of it may be involved; the
extent of involvement largely determines the seizure type.
Epilepsy originates in the brain, but systemic disorders and
resulting metabolic or toxic effects may cause secondary
encephalopathies, and thus, secondary epileptic seizures.
Genetic predisposition also plays a very important role.
The prevalence of epileptic seizures has been estimated to
be between 0.5% (core group) and 5% (fringe group with at
least one seizure during life). More precise figures of Juul-
Jensen and Foldspang (1) lie between 1.27% and 2.44%.
Gumnit (2) estimates that core and fringe group may add up
to 9% of the population.
An overview of various epilepsy prevalence studies (3) shows
fairly consistent figures: Mumbai, India 3.7; Rochester, Minnesota,
6.66; Denmark, 6.9; Poland, 7.8; Nigeria, 5.3; Mississippi, 6.78; and
China, 4.57 (prevalence per 1000). Similar figures were presented
by Greulich and Gerber (4) and Hauser (5).
Introduction and Historical Remarks
Electroencephalography (EEG) has revolutionized the entire field
of epileptology. No one can deny that ingenious and even great
concepts arose in the pre-EEG era; clinical acumen and keen
analysis of the observed phenomena resulted in remarkable
descriptions that have stood the test of time. Indeed, focal motor
seizures, infantile spasms, rolandic seizures, and absence seizures,
to list but a few, were all recognized prior to the development
of EEG. Conventional EEG has opened new vistas and modern
EEG techniques, especially the prolonged EEG monitoring of
people with epilepsy, have made further important epileptologic
contributions. These noninvasive techniques are complemented
by invasive recordings from cortex and deep structures that con-
tinue to provide insights into fundamentals of epilepsy.
We review the role of the EEG in the epilepsies, including the
correlation of the semiology with the electrographic findings
(electroclinical correlation) that are contributory to the EEG
diagnosis of epileptic phenomena in ictal states or the interictal
period in various epileptic conditions.
The interictal paroxysmal phenomena are discussed first.
These discharges represent the basic elements of more complex
ictal patterns.
CHAPTER
26
Principal Differences between Interictal
and Ictal Discharges
A distinction is made between ictal and interictal paroxysmal
activity. This logical differentiation, however, is beset with con-
fusion. It is sometimes difficult or even impossible to distin-
guish between ictal-clinical, ictal-subclinical, and interictal
paroxysmal EEG activity, even using neuropsychological tests.
Computerized testing demonstrates the existence of a gray zone
between interictal and ictal paroxysmal activities (e.g., transient
cognitive impairment).
Problems of Terminology
Definitions of EEG events are indispensable, but they are also
quite difficult to arrive at and are often unsatisfactory. Attempts
to reassess EEG terminology have been made periodically; they
often result in neologistic construction of terms that fail to
receive general acceptance. EEG terminology is filled with
widely used, popular, and, often quite sloppy and inaccurate
terms. Nevertheless, most electroencephalographers know what
is meant by such expressions in the professional jargon.
One could argue that the busy electroencephalographer will
always be satisfied with the nomenclature learned during the
training period, whereas the academicians among electroen-
cephalographers constantly strive for an unobjectionable and
crystal clear terminology free from linguistic flaws and capable
of defining an EEG event from a variety of aspects. This view,
however, is oversimplified. Academia sometimes has a vested
interest in certain terms; it may bitterly resent new terms rec-
ommended by international committees.
A preliminary proposal by the International Federation
of Societies for Electroencephalography and Clinical Neuro -
physiology (6) was followed by a definitive proposal (7) and
a glossary of terms used by clinical electroencephalogra-
phers (8). This glossary was updated in 1999 (9). The hand-
book catalogue of Dutertre (10) deserves special admiration.
Unfortunately, the definition of electrophysiologic events can
never do justice to all facets of the phenomenon. The ques-
tion then arises as to whether we must rigidly follow the new
nomenclature. A negativistic attitude must be castigated, but
electroencephalographers should not be pressured to strictly
adhere to newly proposed terms as time may be needed for
these terms to achieve consensus.
As to epileptic discharges, there has been considerable dis-
unity in the ranks of electroencephalographers. The term
epileptic discharge has been criticized because such discharges
479
480 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
may occur in the absence of clinical seizure manifestations or in
individuals who have never had seizures. The same is true for
the widely used term seizure discharge; even paroxysmal dis-
charge, certainly a more cautious and unassuming term, has not
found general acceptance. The term epileptiform discharge is an
elegant solution since it addresses the critical issue: an epilepti-
form discharge demonstrates an underlying epileptic patho-
physiology but its presence alone does not make the diagnosis
of epilepsy. The diagnosis of epilepsy remains a clinical one,
which must be based on actual clinical events.
INTERICTAL EPILEPTIFORM DISCHARGES
Spike (Single or Random Spike)
According to IFSECN (8), a spike is a transient, clearly distin-
guished from the background activity, with pointed peak at con-
ventional paper speed and a duration from 20 to under 70 msec;
the main component is generally negative and the amplitude is
variable.
The distinction from the background activity is based on
wave morphology and amplitude. In many cases, spikes stand
out against the background because of their high voltage. If the
voltage of spikes and background activity is approximately
equal, the faster character (i.e., the shorter duration) of the spike
is its distinctive feature. Is it possible that a spike of 50 msec
duration and moderate amplitude may be embedded in 20 Hz
activity (50 msec wave duration), for instance, in an epileptic
with considerable drug-induced fast activity? Under these cir-
cumstances, the spike activity may be undetectable. In such a
case, a fast paper speed could demonstrate the morphologic
features of the spike (multiphasic and more pointed character
with a dominant negative phase) in contrast with the more
monotonous appearance of the fast waves. In the digital era,
paper speed is reflected in the sweep time.
Spikes have many characteristics in common and yet there are
also remarkable inter- and intraindividual variations (Fig. 26.1)
and inconsistencies between one spike and the next in the same
Figure 26.1 Various wave morphologies of spikes recorded from the
same patient (age 6 years). Upper lead, T3–T5; lower lead, T5–O1. (First
and second tracing from the top, continued in tracings 3 and 4, respec-
tively 5 and 6.) Note multiphasic character of spikes with predomi-
nance of the negative phase and occasional formation of double spikes.
lead (11). Spikes appear to be hypersynchronous events due to
excessive simultaneous neuronal discharge. This view, however,
is not congruent with the desynchronized (low voltage) EEG
aspect of “electrodecremental seizures” or the electroincre-
mental initiation of focal or generalized attacks. Definitions
such as “epilepsy is paroxysmal hypersynchrony” (12) may be
correct on the neuronal level but partly incorrect from the
viewpoint of the “macro-EEG.”
Basic neurophysiologists have been cautioning against the
casual use of terms such as synchronization and desynchronization.
Low-voltage fast rhythms are not necessarily desynchronized and
may be associated with enhancement and synchronization within
intracortical networks.
On the basis of long-term monitoring and computer analy-
sis, Gotman (13) has raised the question as to whether interic-
tal spikes are in reality postictal. Gotman (13) and Gotman and
Marciani (14) observed marked and prolonged increase of
interictal spikes in a patient after a seizure (awake as well as
asleep). Drug levels had little influence on the rate of spike
activity, and there was no increase of spiking prior to a seizure.
The multiphasic character of a single spike warrants partic-
ular emphasis. A sequence of a minor positive (the “positive
prespike”), a major negative, and a second minor positive com-
ponent (Fig. 26.1) is typical in most instances. A slow negative
component may trail the spike discharge and often attain about
the same amplitude as the negative main component of the
spike. This trailing slow component of a single spike should not
be regarded as evidence of a spike-wave complex. Dutertre (10)
in his handbook catalogue shows examples of single spike-wave
complexes that one could just as well describe as compounded
or multiphasic single spikes. From the vantage point of basic
neurophysiology, there is some reason to presume that the trail-
ing slow wave is caused by the same type of hyperpolarization
demonstrated by Pollen et al. (15) in the experimental analysis
of the spike-wave discharge.
The electroencephalographer very seldom finds single spikes
on the scalp with predominant positive component (16,17).
Positive single spikes are more common in electroencephalo-
graphic and depth recordings; on the scalp, predominant posi-
tivity of single spikes raises the question of defective superficial
cortical laminae. Following surgery of cortical arteriovenous
(AV) malformations or after traumatic laceration (18), such
positive spikes may be present occasionally. With dipole forma-
tion (especially in benign rolandic epilepsy), a positive spike
discharge may be found in a moderate distance from the nega-
tive spike focus (19,20).
Spikes represent the basic element of epileptiform activity in
the EEG. A unitarian view that all spikes mean a hidden or overt
epileptic event would be erroneous. The fine semiology of
spikes is extremely important and the EEG interpreter must
consider the following questions:
What is the precise wave morphology?
Where do the spikes occur?
What is the patient’s age?
What is the state of awareness?
Is there any possibility of an artifact of similar appearance?
 Chapter 26 ■ Seizures and Epilepsy in Infants to Adolescents 481

Is there any possibility of a physiologic potential of similar

appearance?
Discrepancies between spikes and behavioral epileptic events
are not uncommon at all. This highly complex situation has
been reinvestigated by Niedermeyer (21).

Wave Morphology
The largest and most pronounced spikes are not necessarily
associated with more serious epileptic seizure disorders. On the
contrary, rolandic spikes in a child age 4 to 10 years are very
prominent; however, the seizure disorder itself is usually quite
benign or there may be no clinical seizures at all. Alternatively,
“small sharp spikes” are indeed marginal spike discharges and
will be discussed separately.

Spatial Distribution
In childhood, occipital spikes are, in general, the most benign
spike discharges, with less than 50% having clinical seizures;
rolandic central–midtemporal–parietal spikes are also quite
benign, whereas frontal spikes or multifocal spikes are more
epileptogenic.

Age
The significance of the age factor is enormous. From the spikes
of an epileptic newborn to a seizure focus of old age, age-deter-
mined varieties of spikes can be distinguished.
Level of Awareness
Random spikes may occur at any state of waking–sleeping cycle
and occur even in rapid eye movement (REM) sleep when bilat-
eral synchronous bursts of spikes or spike waves are usually
suppressed.
Repercussions on Perceptual Function
Shewmon and Erwin (22,23) investigated the impact of spikes
located over the occipital region and its vicinity on visual percep-
tion and reaction time in three subjects with posterior spikes.
Their work shows that the visuomotor reaction time was pro-
longed (or the response was absent) when the task was time
locked to the spike. Furthermore, “spike-induced dysfunction
was most pronounced when either response hand or visual field
of stimulus was contralateral to the spike” (23). In a further study,
Shewmon and Erwin (24) emphasized the significance of the
trailing slow wave that follows a single spike discharge. The per-
ceptive disturbance was particularly pronounced when the task
was presented during the slow wave. These fascinating findings
indicate that even single spikes of rapid duration may be capable
of momentary impairment of the cortical visual function.
Distinction from Similar Physiologic Patterns
This differentiation is particularly important in the case of ver-
tex sharp waves during deep drowsiness and stage 2 of light
non-REM sleep. In childhood (after age 4), these waves may
have a particularly spiky appearance and may be misinterpreted
as paroxysmal spikes.
Physiologic sharp or spiky vertex waves are usually quite eas-
ily distinguished from rolandic spikes in sleep (Fig. 26.2). Even
Figure 26.2 Coexistence of rolandic spikes and physiologic vertex
waves in light sleep (8-year-old boy, attacks of abdominal pain, no
proven epileptic seizures). Right centroparietal spikes with occasional
spread to the left are marked X; typical examples of a vertex wave are
marked O.
in the more uncommon case of epileptiform spike discharges
over the vertex, the differentiation from physiologic vertex
sharp waves is not difficult.
The physiologic nature of the occipital lambda waves and
“lambdoid” activity (positive occipital transients of sleep)
must also be considered. The main component of this pattern
is positive.
Distinction from Artifacts of Similar Appearance
This distinction depends on the electroencephalographer’s
experience and is usually an easy one.
The interpretation of the clinical significance of spikes can
be extremely difficult and depends on the electroencephalogra-
pher’s experience in the art of reading the EEG recording and
also on the clinical understanding of epileptologic problems.
Extensive personal laboratory experience is just as essential as
scientific knowledge in interpreting the EEG.
Are Interictal Spikes Nothing but EEG Signals?
Our earlier view (found in this chapter in previous editions)—
namely that single spikes are an EEG signal and nothing else—
is in need of some correction. This seems to be particularly true
for “negative myoclonus” (ictally suppressed motor activity),
examples of which have been presented by Gambardella et al.
(25) and Werhahn and Noachtar (26) (also see Ref. 27). A single
482 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

spike may also cause changes in cerebral blood flow and increase
of metabolic demands (28), demonstrated with the simultane-
ous use of EEG and functional magnetic resonance imaging
(fMRI).
Even behavioral changes with impaired test performance
may be associated with single interictal spikes (29,30). Further
information is found in the work of Fisch (27).

Sharp Waves
According to IFSECN (8), a sharp wave is a transient, clearly
distinguished from background activity, with pointed peak at
conventional paper speeds and duration of 70 to 200 msec, that
is, more than approximately 1/14 to 1/5 sec. The main compo-
nent is generally negative relative to the other components. The
term is restricted to epileptiform discharges and does not apply
to distinctive physiologic waveforms, such as vertex sharp tran-
sients, lambda waves, and POSTS.
Jasper (31) pointed out that the rising phase of the sharp wave
is of the same order of magnitude as in spikes, but the descending
phase is prolonged. The configuration with a steeper ascending
phase, however, is not always present. Examples of sharp waves
are shown in Figure 26.3.
A very unusual type of repetitive slow sharp wave activity
has been reported in prematurely born infants with intraven-
tricular hemorrhage; these discharges show predominantly
positive polarity and are recorded mainly over the rolandic
region (32,33). The maximum of this activity is sometimes
found over the vertex.
Spikes and sharp waves are neurophysiologically closely
related phenomena; both of them are typical epileptiform dis-
charges and highly suggestive of an epileptic seizure disorder,
although both phenomena may occur in patients without a his-
tory of seizure disorder.
Sharp waves are usually found as random focal discharges;
most anterior temporal spikes are, in a strict sense, sharp waves.
This is also true for most benign rolandic spikes of childhood.
Sharp waves are more seldom found in generalized synchro-
nous bursts where single spikes, spike waves, and polyspikes
predominate.
It has been contemplated that sharp waves on the scalp cor-
respond with spikes in the depth or on the cortex. Combined
depth and scalp recording clearly refutes this view (34). One can
detect spikes as well as sharp waves in depth recordings; a deep
sharp wave usually corresponds with a sharp wave on the scalp
if there is any corresponding scalp activity at all.
The long duration of a sharp wave permits better insight
into the multiphasic character of this potential. A small preced-
ing positive component may be very fast and qualify as a spike;
Figure 26.3 Recurrent sharp wave recorded from a 61-year-old patient.
Upper lead, F3–C3; lower lead, C3–P3. A prominent negative phase is fol-
lowed by a large and relatively slow trailing positive wave.
Figure 26.4 An example of a very slow sharp wave (blunted sharp
wave), maximal over T6, obtained from a 6-year-old boy with cranio-
cerebral trauma at age 4 followed by seizures.
even a small biphasic spike discharge may precede the much
larger sharp wave. Some sharp waves even exceed the maximum
length of 200 msec of the IFSECN definition (8) (Fig. 26.4), also
named “slow sharp waves” (35) or in cases of particularly long
duration, “blunted sharp waves.” Others become very complex.
They consist of a constantly varying number of components;
such compounded sharp waves may occur as the periodic dis-
charges (periodic lateralized epileptiform discharges) described
by Chatrian et al. (36).
It is certainly not incorrect to use the terms spikes and sharp
waves synonymously when discussing the clinical significance
of a focal transient.
Multiple Spike Complex or Polyspike Complex
This discharge type represents a complex of spikes and may also
be called a polyspike complex. In modern terminology (8), the
term multiple spike complex is preferred because “polyspikes” is an
etymologic hybrid. It has been defined as a complex paroxysmal
EEG pattern with close association of two or more diphasic spikes
occurring more or less rhythmically in bursts of variable duration,
generally with large amplitudes (10) (Figs. 26.5 and 26.6).
Polyspike bursts are readily elicited by electrical stimulation
of single depth leads, especially in limbic regions. On the scalp,
however, polyspikes occur mostly as bilateral or generalized
synchronous discharges. Exceptional focal polyspikes are occa-
sionally encountered; these usually have a frontal maximum,
except for occipital accentuation in hypsarrhythmia. Polyspikes
and also polyspike-wave complexes are sometimes associated
with concomitant myoclonus, especially in primary generalized
epilepsy and in photosensitive individuals with this type of
 Chapter 26
Figure 26.5 Polyspikes in light non-rapid eye movement (non-REM)
sleep, especially over the central region, associated with K complexes.
(With permission of AMA.)
seizure disorder. Children with Lennox–Gastaut syndrome may
also show association of polyspikes and myoclonus.
Runs of Rapid Spikes
This pattern has been described as “grand mal discharge”
(37,38), “fast paroxysmal rhythms” (39), “rhythmic spikes” (40),
and “generalized paroxysmal fast activity” (41). It is seen only in
sleep and occurs in older children, adolescents, and younger
adults (Fig. 26.7). It consists of bursts of spike discharges at a rate
from 10 to 25 Hz, usually generalized but with a well-defined
maximum over frontal regions; it may even be confined to the
frontal leads. The voltage is in the medium to high range, often
exceeding 100 or even 200 V. The discharge rate is in most cases
somewhat irregular. The bursts last for about 2 to 10 seconds;
■ Seizures and Epilepsy in Infants to Adolescents 483
Figure 26.6 The differentiation between polyspikes and drug-induced
fast activity may be difficult. The run of beta waves (center) certainly
falls into the category of fast activity, but the more outstanding and
burst-like run (right) is more likely to represent polyspike activity. A
definitive statement depends on the impressions derived from the entire
record.
bursts of more than 5-second duration are usually associated
with tonic seizures and thus represent an ictal pattern.
The obvious misnomer “grand mal discharges” is based on
certain similarities with the ictal EEG of a generalized tonic–
clonic seizure. In a patient population with generalized
tonic–clonic seizures, runs of rapid spikes as an interictal pat-
tern are a very rare finding. This has been confirmed by the data
of Chayasirisobhon et al. (42). The inappropriateness of the
term grand mal discharge was stressed by Rodin et al. (43), who
thought that this pattern occurred in patients with primary
generalized epilepsy who suffered from more than one type of
seizure and especially from akinetic seizures. As already noted
by Gastaut et al. (40), this discharge is very typical in patients
with Lennox–Gastaut syndrome; it is hardly ever found outside
Figure 26.7 A run of rapid spikes in a 19-year-old epilep-
tic patient (Lennox–Gastaut syndrome). Note anterior
maximum of the discharge. A few slow spike wave com-
plexes are also seen (right temporo-occipital region).
484 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
this clinical entity (44–46). It may be found in very rare
instances of posttraumatic epilepsy with EEG traits of the
Lennox–Gastaut syndrome (47). Brenner and Atkinson (41),
however, insist that runs of rapid spikes (“generalized paroxys-
mal fast activity”) may be found outside the Lennox–Gastaut
syndrome. In our experience, runs of rapid spikes may occur in
all “imitators” of the Lennox–Gastaut syndrome (48,49).
Little is known about the underlying neurophysiologic mech-
anisms of this pattern. It is clearly a frontal lobe discharge, and
the associated tonic seizures are probably arising from the pre-
motor and supplementary motor portions of the frontal lobes.
There could be certain similarities to the electrodecremental
seizure patterns, which also originate most commonly from the
frontal lobe (50). This electrodecremental pattern may actually
be associated with high-frequency discharges (high-frequency
oscillations) that are not detectable at the scalp, giving the false
impression of a decrement in cortical activity (51).
Unfortunately, those electroencephalographers who never
use sleep portions in routine records will never see this fascinat-
ing pattern, which is definitely in need of further basic research.
Spike-and-Slow -Wave Complex
(Classical 3 Hz Spike Wave)
Classical 3 Hz spike-wave complexes are widely known even out-
side the community of electroencephalographers. It is officially
termed spike-and-slow-wave complex (8); this term comprises all
types of spike-wave complexes, which are listed separately
because of markedly differing associated clinical–epileptologic
conditions. The official definition (8) is quite simple: “A pattern
consisting of a spike followed by a slow wave.” Older terms such
as spike-and-dome complex, wave-and-spike complex, and dart-
and-dome complex are seldom used these days and should be
discouraged. The abridged term 3 Hz spike wave is certainly
acceptable, because the omitted word complex is automatically
implied.
Spike-wave complexes of the classical variety (3 Hz) have
been described as the EEG pattern of absence seizures (38).
Since that time, the 3 Hz spike-wave discharge has been equated
with absence epilepsy. Regrettable mistakes have occurred due to
lack of communication between electroencephalographers and
clinicians who initiated treatment with anticonvulsants specific
for an absence seizure due to the use of the term 3 Hz spike-wave
complexes alone. Unfortunately, the term petit mal discharge (37)
has reinforced this type of reflex thinking. Fortunately, elec-
troencephalographers and epileptologists have always cautioned
against the misinterpretation of the term petit mal discharges;
this should be interpreted as “discharges that occur in absences
but also in other conditions.” The work of Silverman (52), Clark
and Knott (53), and Lundervold et al. (54) has clearly shown that
the spike-wave discharge as an interictal event correlates with
the occurrence of petit mal absences only in moderate percent-
age of patients (16% according to Clark and Knott).
The 3 Hz spike-wave discharge is a worthwhile subject for
more detailed discussion. As to wave morphology, there is good
evidence that the spike-wave complex is not simply an associa-
tion of a spike and a slow wave, although the above-mentioned
definition is justifiable as a simplification. The spike-wave com-
plex contains hidden components that can be seen even with
simple visual analysis at 30 mm/sec paper speed; special record-
ing methods will better reveal these extra components. Gastaut
and Hunter (55) and Cohn (56) and Weir (57) have elucidated
the components of this pattern. Most important is a spike of
positive polarity during the descending portion of the slow
wave (57), becoming evident as notching under usual recording
conditions.
On the basis of experimental studies of cortical faradiza-
tion, Jung and Toennies (58) stressed the inhibitory character
of the slow wave following a spike discharge (“Bremswelle”).
Experimentally induced spike-wave-like activity in cats, elicited
by electrical stimulation of thalamic intralaminar structures, was
studied with intracellular cortical microelectrodes (15,59,60).
According to Pollen (59), all components of the spike-wave com-
plex were the “consequence of postsynaptic activity.” This author
points out that “the initial positivity which may precede a nega-
tive spike (depending on the intensity of the stimulation) results
from depolarization of middle and deep lying neurons, whereas
the surface negative spike is the result of EPSP (excitatory postsy-
naptic potential) generation primarily upon apical dendrites of
vertically oriented pyramidal type neurons.” According to Pollen
(59), the long surface negative wave is generated by inhibitory
postsynaptic potentials (IPSPs) predominantly on or near the
soma of similar pyramidal cells. Thus, the slow wave indeed
appears to be related to inhibitory impulses.
In this view, the spike-wave discharge apparently represents an
alternating succession of excitation and inhibition. The clinical
ictal activities are thus constantly curbed by intervening inhibitory
impulses that prevent the discharge from progressing into massive
downward discharges with motor effects (polyspikes with massive
myoclonus) or to a full-blown generalized tonic–clonic seizure.
For this reason, motor manifestations of ictal episodes character-
ized by spike waves are almost always inconspicuous or modest.
Furthermore, spike-wave bursts rarely proceed into generalized
tonic–clonic seizures; such rare exceptions have been demon-
strated by Halász (61) and Niedermeyer (62).
The distinction between classical (3 Hz), slow (1 to 2.5 Hz),
and fast (4 to 5 Hz) spike-wave complexes and the smaller
6 Hz spike-wave discharge is justified on the basis of different
clinical–epileptologic correlates of each spike-wave type; it is
not an example of electroencephalographic hair splitting. The
classical 3 Hz spike-wave discharge is most typical and most
pronounced in children with absence seizures. Clinical absences
are usually present when the burst duration is greater than
5 sec. Thus, shorter bursts are usually subclinical, but numerous
computerized neuropsychological studies have been employed
to demonstrate a clinical component (fluctuations of level of
awareness) during an apparent subclinical spike-wave bursts.
The classical spike-wave complex is, in most cases, easily acti-
vated by hyperventilation, whereas the slow and the fast forms
of this discharge show little or no enhancement.
A classical 3 Hz spike-wave burst does not run exactly at a rate
of 3 Hz. The complexes are faster at the onset of the burst (mostly
around 4 Hz), then slow down to 3.5 and 3 Hz for the main por-
tion, and eventually slow down to 2.5 Hz at the end of the burst.
During a burst, the spike discharges become gradually smaller,
often shrinking to insignificance (Fig. 26.8) as in drowsiness and
sleep.
 Chapter 26
The spatial distribution of the bursts is very typical; the max-
imum almost always lies over the frontal midline region,
whereas a minimum is found over temporal and occipital areas.
The different conclusion by Dondey (63) was based on EEG
recordings in which no midline electrodes were used. Rodin
and Ancheta (64) have confirmed the concept of a frontal mid-
line maximum using brain mapping. Almost all spike-wave
bursts are bilateral synchronous or generalized synchronous;
the synchrony, however, is not perfect (56,65). The age of distri-
bution lies mainly in the range from 4 to 16 years.
Computerized analysis by Lemieux and Blume (66) has also
shed light on the spatial distribution of the spike-wave com-
plex: the field distribution of spikes differed from that of the
ensuing slow waves. Spikes “arose in one frontal region and
propagated to the homologous part of the other frontal region
with a peak-to-peak interval less than 15 msec. Less commonly,
spikes first moved anteriorly within the initiating frontal region
before contralateral propagation occurred.” By contrast, nega-
tive slow waves were more diffuse, more symmetrical in evolu-
tion, and more posteriorly centered than the spikes.
The onset of the spike component in one frontal lobe is con-
gruent with the earlier findings of Cohn (56) and Lüders et al.
(67), who noted asynchronies ranging from 5 to 20 msec. The
frontal lobe is a large structure and subdivisions are necessary.
Onset of spike-wave activity (and, in particular, of the spike
component) over the frontal midline means in reality that a
frontal portion in the immediate vicinity of the interhemi-
spheric fissure is the origin of the discharge; this area belongs to
the supplementary motor zone.
What are the neuropsychological correlates of subclinical
generalized spike-wave bursts and clinical absences with spike
waves? In this context, it should be pointed out that the impact
of apparently subclinical 3 Hz spike-wave bursts might be
stronger than one would expect from routine clinical observa-
tion. With its maximum over the frontal midline (frontal lobe
close to interhemispheric fissure), there might be some impact
on prefrontal lobe functions and, in particular, on the working
memory. This concept of working memory, introduced by
Baddeley (68) and enormously refined by Fuster (69,70),
implies a system of constant perceptory afferences to the pre-
frontal dorsolateral region whence motor impulses emanate,
resulting in readiness (set) for motor action. This system could
■ Seizures and Epilepsy in Infants to Adolescents 485
Figure 26.8 Petit mal absence with generalized synchro-
nous spike-wave complexes. The enormous amplitude of
the discharges necessitates considerable lowering of the
gain. The frontal voltage maximum is evident. Note grad-
ual decline of the spike component at the end of attack,
along with slight slow down of the frequency to about
2.5 Hz.
be blocked for a moment by a few (even by a single) spike waves
and interfere with the working memory. In a clinical 3 Hz
spike-wave absence, blocking of the working memory might be
the most likely explanation for the mild disturbance of con-
sciousness and its immediate full recovery with the termination
of the attack (71). No “true disturbance of consciousness”—
epileptic or nonepileptic—can recover immediately. In other
words, the function of working memory is simply suspended
for the duration of the spike waves.
Due to the uniqueness of primary generalized epilepsy with
3 Hz spike waves in the human (72), animal models for this
response have great limitations. The arrest reaction of the monkey
(73) produced by alumina cream injection into the fronto-orbital
cortex has some resemblance to the human absence epilepsy; this
cannot be said about the sudden freeze of motility noted in the cat
stimulation of the cat’s mesial anterior thalamus (74); there is no
spike-wave activity associated with this type of arrest reaction.
According to Sperling and Skolnick (75), there is a general
decrease of cerebral blood flow (133 xenon method) during
human generalized spike-wave activity, which, however, is less
pronounced in frontal regions than in the parietal lobes.
Transcranial magnetic stimulation of the motor cortex in
patients with primary generalized epilepsy and 3 Hz spike
waves produced motor evoked potentials of significantly
reduced size when the cortical stimulus was time locked to the
slow wave component of the spike-wave complex (but slightly
reduced or unchanged when the stimulus was time locked with
the spike) (76). This underscores the special inhibitory char-
acter of the slow wave component, which has been known since
the work of Jung and Toennies (58).
Spike-Wave Complex (Slow 1 to 2.5 Hz)
After the demonstration of the 3 Hz spike-wave complex and its
relationship to the absence seizure (38), the Gibbses and William
G. Lennox were struck by the occurrence of slow spike-wave
complexes in a much different type of patients with seizures
other than the petit mal absence. The classical 3 Hz spike-wave
complex was termed petit mal discharge. Consequently, the slow
spike-wave complex was termed petit mal variant discharge. This
term was used first by Gibbs et al. (77). In that study, hypo-
glycemia was induced in order to demonstrate an increase of
classical 3 Hz spike-wave activity in the hypoglycemic state; an
486 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 26.9 Generalized slow spike-wave complexes (mostly around
2 Hz) in a child with severe epileptic seizure disorder (Lennox–Gastaut
syndrome).
increase of slow spike-wave activity, however, was not elicited
with this activation method.
The wave morphology of this pattern varies considerably. In
the majority, the complex consists of a rather slow spike (accord-
ing to the definition, a sharp wave, lasting 70 msec or longer) and
a slow wave. In a sizable number of cases, true spikes (60 msec or
less duration) are followed by a slow wave (Fig. 26.9).
The spatial distribution is, in most instances, quite similar to
that of the 3 Hz spike-wave complex. In the vast majority of the
cases, the bursts are bilateral or generalized synchronous with
imperfect synchrony and a frontal midline maximum is the rule
in these discharges. Lateralization or occasional focal occur-
rence is sometimes observed, usually in children with severe
residual brain damage in certain areas where parenchymatous
destruction abolishes the spike-wave discharge.
Generalized slow spike-wave discharges are often quite pro-
longed. In some children or adolescents, the entire sleep por-
tion (light and moderately deep non-REM sleep) consists of
unabated generalized slow spike-wave activity. The diagnosis of
an “electrical status epilepticus” may or may not be justified in
such cases; there are usually no behavioral or polygraphic char-
acteristics to suggest an ongoing ictal event. In some cases, the
slow spike-wave activity is found only in non-REM sleep.
Although the classical 3 Hz spike-wave complex is seldom
seen before the age of 4 and almost never before 3.5 years, its slow
counterpart appears much earlier, sometimes before the age of 6
months. At this early age, the frontal maximum may not be read-
ily demonstrable.
The slow spike-wave complex is almost always associated
with a severe and often uncontrollable type of childhood
epilepsy (seldom with onset between ages 11 and 20 years) called
Lennox–Gastaut syndrome (40,44,46,77,78). Most of these chil-
dren have a mixed seizure disorder and mental retardation.
Thus, the distinction between slow (1 to 2.5 Hz) and classical
(3 Hz) spike waves is of great clinical significance. One has to
keep in mind, however, that many children or adolescents with
slow spike waves also show series of 3 Hz or even 4 Hz spike-
wave complexes; the presence of these spike-wave types is obvi-
ously of no significance in such cases and what counts is the slow
type. On the other hand, patients with classical 3 Hz spike-wave
complexes and a usually benign type of epileptic seizure disor-
der almost never show slow spike waves except when a 3 Hz or
4 Hz spike-wave burst may slow down to 2.5 Hz at the end.
Spike-Wave Complex (Fast 4 Hz, 4 to 5 Hz)
This pattern is closely related to the classical 3 Hz spike-wave com-
plex; both discharge types are most commonly lumped together.
A differentiation, however, is justified on clinical grounds.
Gastaut (79) deserves credit for an outline of the distinctive
features between the 3 Hz and the 4 Hz (or 4 to 5 Hz) spike-
wave discharge. According to his investigations, the fast spike-
wave burst is usually of shorter duration (1 to 3 sec), it occurs
in patients older than 15 years, the bursts are always subclinical,
and the associated seizure disorder is usually characterized by
myoclonic jerking, grand mal attacks, or a combination of both
seizure types, whereas absence seizures are quite uncommon.
Paroxysmal flicker responses are common in such patients. A
positive family history of epileptic seizures disorder is fre-
quently found in this group of patients.
The 4 Hz spike-wave discharge is spatially distributed in the
same manner as the 3 Hz spike-wave discharge; the frontal mid-
line maximum is quite prominent.
The 6 Hz Spike-and-Slow -Wave Complex
(Phantom Spike and Wave, Wave and Spike
Phantom)
This is now considered a pattern of uncertain clinical signifi-
cance, or a normal variant (marginal paroxysmal pattern),
described by Gibbs and Gibbs (37): “The spike in this pattern
has a strong positive component but the entire wave complex
looks in general like a miniature reproduction of the 3 per sec-
ond spike and wave of petit mal.” Because of the positive spike
component, Silverman (80) stressed the similarity to the 6 Hz
component of the 14 and 6 Hz positive spike burst and reported
a transition from the former to the latter with deepening
drowsiness. Marshall (81) introduced the term wave and spike
phantom, which has found but limited acceptance.
The clinical correlates of 6 Hz spike-wave complex have been
elucidated by Thomas (82), Gibbs and Gibbs (83), Hughes et al.
(84), and Thomas and Klass (85). It is usually a pattern of adult-
hood but may also occur in adolescents and children. About
50% to 60% of the patients suffer from indubitable epileptic
seizures, mostly generalized tonic–clonic seizures; the remain-
der show a history of syncopal attacks, posttraumatic states, or
(as emphasized by Small in Ref. 86) psychiatric problems.
The 6 Hz spike-wave complex is an uncommon but not rare
pattern (about 0.5% to 1% in a central EEG laboratory).
The discharge may be recorded in waking state, drowsiness,
and light non-REM sleep; light drowsiness appears to be the
optimal recording condition (Fig. 26.10).
Investigations regarding the spatial distribution of this
pattern have led to an important dichotomy. Hecker et al. (87)
 Chapter 26
Figure 26.10 A short run of 6 Hz spike waves, posterior type,
recorded in a 52-year-old woman with a history of head injury 2 years
earlier and subsequent headache, dizziness, and memory loss. There
was computed tomography (CT) scan evidence of cortical atrophy.
pointed out a distinction between a frontal versus an occipital
accentuation. The frontal type is most commonly associated
with epileptic seizure disorders and sometimes found in combi-
nation with other epileptiform discharges. The occipital types
are found predominantly in patients with no evidence of epilep-
tic seizure disorder.
In a study of 839 patients with 6 Hz spike-wave complexes,
Hughes (88) placed particular emphasis on the distinction
between frontal and occipital maximum. He summarized the
principal features of both forms by the acronyms WHAM and
FOLD:
W: waking record F: females
H: high in amplitude O: occipital
A: anterior L: low in amplitude
M: males D: drowsy record
An interesting sidelight of this observation is the uneven
gender distribution, which is rather unique; no other EEG
abnormalities show any gender preference.
According to Hughes (89), patients with anterior 6 Hz spike
waves are more likely to have epileptic seizures, and those with
posterior discharges tend to have neuroautonomic disturbances.
■ Seizures and Epilepsy in Infants to Adolescents 487
Kocher et al. (90) were struck by the occurrence of 6 Hz
spike-wave complexes in the abstinence or withdrawal phase of
drug-dependent individuals. Tharp (91) produced the dis-
charge by intravenous diphenhydramine (50 mg) in normal
subjects. Hecker et al. (87) found that the occipital form of the
6 Hz spike-wave complex is often related to drug dependence
(hypoanalgesics, barbiturates) and withdrawal. All this under-
scores the need for a distinction of the two forms. Thus, the
frontal form will have to be regarded as a truly epilepsy-related
paroxysmal pattern, whereas the occipital form finds itself on
the fringe of paroxysmal discharge types.
Patterns of Uncertain Clinical Significance,
or Benign EEG Variants, Benign Epileptiform
Variants, and Marginal Patterns
The just-mentioned 6 Hz spike-wave discharge is considered a
“benign EEG variant and pattern,” which include patterns with
an “epileptiform morphology.” According to Westmoreland
(92), the 6 Hz spike-wave discharge is not a reliable indicator of
seizures (also see Refs. 85,93). Other patterns listed under the
heading of benign patterns are small sharp spikes, 14 and 6 Hz
positive bursts, wicket spikes, and the rhythmic temporal theta
bursts of drowsiness (psychomotor variant pattern).
According to Westmoreland, these patterns “have an epilep-
tiform appearance but are considered normal or benign vari-
ants of EEG activity or activity of uncertain significance.” These
patterns occur in the healthy individual and there is no doubt
that various EEG abnormalities may occur in someone who
functions normally (48,49,94). Although these may be associ-
ated with potential epileptogenicity, they should be interpreted
with caution, especially if the only finding on the EEG.
Westmoreland (92) suggests that “critical evaluation of the var-
ious types of EEG activity is important.” From a single labora-
tory in London, Ontario, benign epileptiform variants occurred
in 3.4% of 1183/35,249 recordings (95).
Rudimentary Spike-Wave Complex
(“Pseudo Petit Mal Discharge”)
Gibbs and Gibbs (83) described a paroxysmal discharge that
consists of generalized or nearly generalized high-voltage 3 to
4 Hz waves “with a poorly developed spike in the positive
trough between the slow waves, occurring in drowsiness only.”
This pattern is most prominent over parietal areas and was
found only in infancy and early childhood with hypnagogic
hypersynchrony. This hypogognic hypersynchrony can occur in
generalized bursts rather than in prolonged stretches. These
bursts, sometimes with a frequency as slow as 3 to 4 Hz, may
contain small spike elements that justify a term like rudimen-
tary spike waves. The Gibbsian term pseudo petit mal discharge
suggests a relationship to spike-wave discharges and absence
epilepsy. However, this pattern should not be equated to a spe-
cific epileptiform discharge and a transition into a classical or
other spike-wave pattern does not occur (83). A history of
febrile illness is very often found in these children with rudi-
mentary spike waves and a history of febrile convulsions is also
quite common (83,96).
488 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Small Sharp Spikes
Small sharp spikes, also known as benign epileptiform tran-
sients of sleep (BETS), were delineated by Gibbs and Gibbs
(37,83). It is the most inconspicuous paroxysmal discharge and
hence is easily overlooked. The main negative and positive com-
ponents are of about equally spiky character (Fig. 26.11), it is
fairly widespread but seen chiefly over temporal and frontal
areas, either shifting from side to side or synchronously firing.
The pattern is almost exclusively found in drowsiness and/or
light non-REM sleep. It is essentially a pattern of adulthood
with a peak between ages 30 to 60 years; occurrence in adoles-
cence and old age is somewhat less common. It is virtually
absent in the first 10 years of life.
Koshino and Niedermeyer (97) found a prevalence of 1.36%
with a peak of 2.9% between ages 30 and 40. In this study, two
thirds (67.4%) of the patients had a history of epileptic seizures;
this underscores the paroxysmal character of the discharge.
From Ontario, small sharp spikes occurred in 1.85% of EEG
recordings (95).
Small sharp spikes sometimes are precursors of typical ante-
rior temporal spikes or sharp waves, which, in such patients,
simply occur somewhat later in drowsiness or sleep. This asso-
ciation was also noted by Gibbs and Gibbs (83). In a depth EEG
study by Westmoreland et al. (98), an extensive generator area
of small sharp spikes was found. Among people without
epilepsy, the discharges may occur in patients with cerebrovas-
cular disorder, syncopal attacks, and psychiatric problems.
Small (86) and Small et al. (99) have placed special emphasis on
the occurrence of this spike discharge in a psychiatric popula-
tion and especially in patients with manic-depressive illness.
Figure 26.11 Small sharp spikes (51-year-old patient).
Note the subtle character and moderate voltage of the
discharge; also note its predominance in the left
nasopharyngeal lead. There is evidence of spread into
T3, as well as into the right nasopharyngeal lead. The
left section was recorded in the waking state (transition
to earliest drowsiness); the middle and right sections
were recorded in sleep. (Reproduced with permission
from Koshino Y, Niedermeyer E. The clinical signifi-
cance of small sharp spikes in the electroencephalo-
gram. Clin Electroencephalogr. 1975;6:131–140.
It is interesting to note that Gibbs and Gibbs (83) found a
rather high prevalence of small sharp spikes in presumed nor-
mal adult control subjects, reaching 7.9% in the range from 40
to 49 years. Reiher and Klass (100), Lebel et al. (101), White et
al. (102), Klass and Westmoreland (93), and Gutrecht (103)
have stressed the nonspecific nature of this pattern and do not
regard its occurrence as an abnormality. By contrast, Low et al.
(104) as well as Hughes and Gruener (105) have pointed out
that small sharp spikes are not a normal or meaningless pattern.
The data of these authors clearly show that this pattern indi-
cates “a moderate degree of epileptogenicity” (105). This has
been strongly confirmed by Saito et al. (106) and also by Molaie
et al. (107). This is another indication of the inherent epilepto-
genic potentialities in so-called benign patterns.
Needlelike Occipital Spikes of the Blind
Spike discharges of a particularly fast and needlelike character
develop over the occipital region in most congenitally blind
children (83,108,109). These spikes are completely innocuous,
unrelated to epileptic seizure disorder, and probably due to a
state of functional deafferentation (110). The discharges disap-
pear during childhood or adolescence.
The 14 and 6 Hz Positive Bursts
(Fourteen and Six Positive Bursts)
In the IFSECN (8), “Fourteen and Six Hz Positive Bursts,” are
defined as: “Bursts of arch shaped waves at 13 to 17 Hz and/or 5
to 7 Hz, most commonly at 14 and/or 6 Hz, seen generally over
the posterior temporal and adjacent areas of one or both sides of
the head during sleep. The sharp peaks of its component are
 Chapter 26
Figure 26.12 Various examples of 14 Hz and 6 Hz positive spikes
(underlined). Note the posterior predominance for this pattern and
shifting asymmetrics. Also note the sometimes-blurred distinction
between the 14 and 6 components, due to notch formation. The record-
ing was obtained from a 12-year-old patient; montages to ipsilateral ear.
positive with respect to other regions.” For these bursts, the ampli-
tudes are generally below 75 V, the pattern is best demonstrated
by referential recording using contralateral ear lobes or other
remote reference leads, and the clinical significance is controver-
sial. The synonymous term ctenoids was introduced by Lombroso
et al. (111) with regard to the comblike shape. Contrary to
Dutertre’s (10) statement, “mitten” is not a synonymous term.
The 14 and the 6 component may be observed independ-
ently. According to Gibbs and Gibbs (83), the 6 Hz component
prevails in early childhood and adulthood, while the 14 Hz
component is more prominent in older children and adoles-
cents. Hughes (112,113) emphasized the harmonic character
of the major frequencies and preferred the term 14 and 7 Hz
positive spikes (Fig. 26.12). From Ontario, 14 and 6 positive
bursts occurred in 0.52% of EEG recordings (95).
This pattern was described first by Gibbs and Gibbs (114) and
regarded as “evidence of thalamic and hypothalamic epilepsy.”
The following 15 years produced exciting correlations to auto-
nomic nervous system dysfunctions and behavioral disorders
(115–126). Enthusiasm flagged when it became clear that this
pattern is too often found in healthy individuals (111,127). Little
(128) even presumed that this pattern “is more likely to be a sign
of health than of disease.”
The “14 and 6 pattern” may be confined to the regions lying
beneath a skull defect (129) and hence could denote a “hidden
pattern” that is not detectable from the scalp unless it reaches
major proportions.
The pattern occurs most commonly in children after age 4
and adolescents and declines in adulthood; Gibbs and Gibbs
(83,130) recorded the 14 and 6 Hz activity even in infants. Its
occurrence in the waking state is exceptional; it typically occurs
in deep drowsiness and very light non-REM sleep, although
deep sleep might be more conducive in very young children
(83). Okada et al. (131) noted 14 and 6 Hz positive bursts in
association with simultaneous negative spikes over the frontal
area (suggestive of dipole formation).
■ Seizures and Epilepsy in Infants to Adolescents 489
Little is known about origin and neurophysiologic discharge
demonstrated in depth recordings at the thalamic as well as the
basal ganglia level (132); the discharge was also repeatedly
recorded in the amygdaloid complex.
Much of the controversy about this pattern has arisen from
differences of recording techniques and the use of routine sleep
tracings and, in particular, the length of the sleep portion. The
prevalence used to range from 10% to 30% at the age of 10 to 25
years, but there has been an unexplained marked decrease of the
prevalence of “14 and 6” Hz discharge according to Gibbs (133).
We interpret the 14 and 6 Hz positive burst itself as “within
normal limits,” unless it is extremely frequent and pronounced;
then it is regarded as minimally abnormal. It appears to be a
marginal pattern of no or very little epileptologic significance
and proven cases of epileptic seizures very seldom have a 14 and
6 Hz positive burst as the only significant finding.
However, the 14 and 6 Hz positive burst may occur in the
advanced states of metabolic encephalopathies, especially in
hepatic coma (134,135). Drury (136) reported periodic 14 and
6 Hz positive bursts in 7/121 children with hepatic coma, mostly
in association with Reye’s syndrome. All cases had a severe back-
ground EEG abnormality; these bursts were activated by stimu-
lation in four children, and disappeared on follow-up recordings
in six. The “14 and 6” pattern may also occur as a transient toxic
effect of diphenhydramine hydrochloride (Benadryl).
Rhythmic Temporal Theta Burst of Drowsiness,
Rhythmic Midtemporal Discharges, or
Psychomotor Variant
This pattern was widely known as “psychomotor variant dis-
charge” (37,83) because of certain similarities to rhythmical
ictal activity occurring in psychomotor seizures (temporal lobe
seizures, complex partial seizures). This term has been discour-
aged by IFSECN (8), which recommended the term rhythmical
theta bursts of drowsiness. The term rhythmic midtemporal dis-
charge of Lipman and Hughes (137) is also widely used.
The pattern consists of long runs of rhythmical activity in the
range of 5 to 6.5 Hz with a maximum over the midtemporal
region, often with considerable spread into posterior temporal,
anterior temporal, and occipital areas. The theta activity shows a
well-defined negative sharp component (Fig. 26.13) that infers
the epileptiform character of the discharge. These trains of sharp
theta waves may occur in a unilateral, bilateral shifting, or syn-
chronous distribution type. The duration of a single run usually
exceeds 10 seconds and may reach 1 minute or more. Very often,
the first run is noted in early drowsiness; in deep drowsiness and
light sleep, the pattern tends to disappear. According to Egli et al.
(138), patients with rhythmical theta bursts show shortened
periods of REM sleep and also, to a lesser degree, of slow sleep
in nocturnal sleep studies at the expense of log drowsy periods
with the rhythmic theta pattern. Hughes (139) has emphasized
that the rhythmic midtemporal discharge is not limited by the
drowsy state and may be very active in wakefulness.
This pattern is rare. Egli et al. (138) found a prevalence of
0.1% in a large laboratory, and Santoshkumar et al. (95) in 0.12%
of EEG recordings. Although occurring mainly in younger or
middle-aged adults, it is also seen in adolescents and children.
490 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 26.13 Psychomotor variant discharge of the right midtemporal
region (T4). Tracing was obtained from a 21-year-old woman. The
record had low-voltage character; this necessitated the use of high gain.
The clinical significance is not clear. Despite its strongly epilepti-
form appearance, its epileptogenic properties seem to be very
low; most patients have no history of clinical seizures, although
Lipman and Hughes (137) found seizures in 36% of their
patients. Personality disorders and some autonomic nervous
system dysfunctions are common in patients with this pattern
(83,137,140). This view is shared by Eeg-Olofsson and Petersén
(1982; cited in Boutros et al. (141)), who also studied the
genetic transmission of this pattern and wondered about possi-
ble autosomal-dominant inheritance.
Subclinical Rhythmic EEG Discharge
of Adults (SREDA)
This pattern was originally described by Naquet et al. (142) as a
“paroxysmal discharge of the temporo–parietooccipital junction,”
facilitated or triggered by temporary hypoxic conditions. In spite
of certain epileptiform features and its rhythmical sharp charac-
ter, the possibility of an epileptic (ictal) event has been strongly
de-emphasized. Herranz and Lopez (143) have described this pat-
tern as “subclinical paroxysmal activity” in 31 patients and
stressed its rarity; it has an incidence of 0.02% to 0.045%. From
Ontario, SREDA occurred in 0.07% of EEG recordings (95). The
authors have seen this pattern once in a child.
Westmoreland and Klass (144,145) and Miller et al. (146)
deserve particular credit for the elucidation of this discharge
type. These authors disagree with the view of Naquet et al.
(142) concerning the major role of cerebrovascular insuffi-
ciency and hypoxia, although the average age of their patients
was 61 years. The term “SREDA” has been proposed by Miller et
al. (146). O’Brien et al. (147) analyzed this pattern in seven
recent patients and only one of them showed evidence of cere-
brovascular disorder.
This pattern occurs mainly in the waking state or in light
drowsiness. The onset may be fairly abrupt with widespread
sharp rhythmical theta (4 to 7 Hz) and occasionally with delta
activity. In other cases, the onset is gradual with a few single sharp
discharges that increasingly develop rhythmical character, first at
delta, then at theta frequency. As to the spatial distribution, a
maximum of this discharge is usually found over the centropari-
etal region and especially over the vertex. Hyperventilation may
induce this pattern.
Periodic or Quasiperiodic Discharges
These rhythmically (periodic), or nearly rhythmically (“quasi-
periodic”) recurring discharges are of an epileptiform charac-
ter, but are not usually associated with chronic epilepsy. Periodic
discharges occur with a specific interval duration and quaiperi-
odic discharges occur with a varying interval duration. These are
an important EEG feature of a severe ongoing central nervous
system (CNS) disorder. They are hence disease-suggestive and
sometimes virtually disease-specific, rather than suggestive of
epilepsy in general. A review of periodic activities was presented
by Gaches (148) as well as by Bauer and Pieber (149). The peri-
odic character of these patterns remains enigmatic. This rhyth-
mical firing is markedly different from the predominantly
random character of interictal seizure discharges (spikes, sharp
wave, etc.). There is no satisfactory model to explain the perio-
dicity of the discharges.
Periodic discharges are always of large amplitude, mostly in the
range of 100 to 300 V. Morphologically, these may be simple
sharp waves, usually with a duration exceeding 150 msec, but may
be a compounded waveform. Periodic discharges may be focal,
widely scattered, or generalized synchronous. A fine tabulated
review of periodic discharges has been presented by Spehlmann
(150). Periodic discharges are differentiated by the interval: either
long-interval or short-interval periodic discharges.
Periodic Complexes in Subacute Sclerosing
Panencephalitis (SSPE)
SSPE is now exceptionally rare in the developed world following
measles immunization. Periodic complexes dominate the second
state of SSPE, a prolonged phase in which the clinical diagnosis is
usually made. This type of complex discharge almost never
occurs in other clinical conditions and is hence almost disease-
specific. It is a very reliable and highly contributory diagnostic
finding in this disease. The discharge duration varies from 0.5 to
3 seconds and are typically formed by two or more waves with
mean amplitude of 500 V (100 to 1000 V) (150, reaching as
high as 1400 V (11)). The enormous discharge amplitude might
be confused with movement artifact, which is made more con-
fusing by the associated myoclonus or sudden loss of tone.
The periodicity develops with disease progression. The earlier
stages are usually dominated by diffuse 1 to 3 Hz activity of fairly
rhythmical character, but the compounded discharge may be
aperiodic. Sometimes, the periodic discharge is present in the
earliest stage of the disease and may persist to the fatal outcome;
more often, this pattern disappears in the terminal or third stage.
The discharge is more impressive in the waking state (Fig.
26.14). The elements of the discharge are variable; a giant slow
wave is usually mixed with several sharp waves. The discharge is
generalized, with a maximum over the frontocentral areas and
vertex. The frequency of the complexes ranges from 4 to 16 Hz
(151). This is an example of a long-interval periodic discharge.
Accompanying myoclonus is fairly synchronous with the
discharge; the motor activity may precede or trail the discharge
 Chapter 26 ■ Seizures and Epilepsy in Infants to Adolescents 491

Figure 26.14 Periodic discharges in subacute sclerosing panencephalitis (SSPE). A: At an earlier state with some preser-
vation of other activities. B: At a more advanced state (recorded from two different patients, ages 8 and 9 years).

in a range of 200 to 800 msec. There may be minor interhemi-
spheric asynchronies (15 msec) and an earlier appearance in
parietooccipital areas than in frontocentral areas, where the dis-
charge shows its most impressive voltage (152).
The background activity between the complexes is quite
variable. The background may be normal with posterior alpha
rhythm; however, this is exceptional; severely disordered back-
ground activity is most commonly found. Slow activity prevails
and numerous spikes are sometimes recorded, especially over
anterior areas. Spike-wave complexes may also be present; even
clinical absence seizures with classical 3 Hz spike-wave com-
plexes have been observed (153). A case of acute fulminant
SSPE with only 2 months’ duration was reported by Gökcil et
al. (154); the EEG showed very pronounced discharges.
Periodic discharges mimicking SSPE have been reported by
Giovanardi Rossi et al. (155) in children with myoclonic juve-
nile epilepsy of nonprogressive character. These discharges,
however, were confined to ictal episodes. Most unusual is the
observation of Prier et al. (156): convincing periodic complexes
of the aforementioned type in an indubitable case of multiple
sclerosis (age 15 years).
The origin of the periodic SSPE discharge is controversial.
Radermecker and Poser (157) have proposed a deep thalamic or
mesencephalic-reticular origin; this view has been supported by
Cobb (158,159), who performed depth EEG recordings, and by
Lombroso (160), who studied direct current (DC) potential
changes during the periodic discharge. Other authors favor the
idea of a cortical origin (161–163).
In an autopsy-confirmed case of SSPE, Martinovic (164)
demonstrated periodically occurring generalized high-voltage
bursts firing at a rate of 11 to 13 Hz with a spiky contour. There
was a repetition interval of 27 to 28 seconds. This most unusual
pattern was gradually replaced by the conventional type of peri-
odic discharges.
Periodic Complexes in Herpes Simplex
Encephalitis
In this extremely severe but not invariably fatal necrotizing
encephalitis, slow repetitive spikes have been noted by
Radermecker (165). These observations were subsequently con-
firmed by Millar and Coey (166), Perier et al. (167), Carmon et
al. (168), Rawls et al. (169), Adams and Jennett (170), Upton and
492 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Gumpert (171), Gaches and Arfel (172), Elian (173), Kugler et al.
(174), and Cobb (175). The discharge was also found in experi-
mental disease inoculated rabbits by Gupta and Seth (176).
In an earlier stage, local mostly unilateral temporal polymor-
phic delta waves are the most striking feature, but soon large sharp
waves emerge over the most affected region. These discharges usu-
ally fire at intervals from 2 to 4 seconds. This is a short-interval
periodic discharge. The discharge may be quite slow and exceed
1000 msec; consider that a sharp wave is defined as having a dura-
tion from 70 to 150 msec. The amplitudes are in the range from
100 to about 500 V. Similar periodic sharp discharges have been
demonstrated in neonatal herpes simplex encephalitis (177). In
the course of the disease, the originally regional discharge tends to
become generalized synchronous; asynchronous bilateral dis-
charges may also occur. Predominantly occipital localization of
the periodic discharges has been reported by Bergey et al. (178).
In contrast with the long-interval periodic complexes found
in SSPE, the periodic discharge of herpes simplex encephalitis
has a short interval, with the longest intervals of 4 seconds or
less; according to Gaches (148), this criterion is always met in
the herpes simplex discharge. In some cases, the periodic dis-
charge can be detected only when almost daily repeat records
are carried out. This explains the fact that periodic discharges
are not reported in all cases of herpes simplex encephalitis.

Periodic Discharges in Creutzfeldt–Jakob Disease
The clinical and neuropathologic features of this subacute dis-
ease were established by the work of Creutzfeldt (179), Jakob
(180), and Kirschbaum (181,182). A very similar syndrome
known as spongiform encephalopathy (183) and a special vari-
ant described by Heidenhain (184) were eventually found to be
the same disease. Although originally presumed to be a degen-
erative CNS disorder, a slow virus was eventually demonstrated
as the causative agent (185,108). These investigators transmit-
ted this virus to the chimpanzee with intracerebral inoculation
and produced a similar invariably fatal disease. It is interesting
to note that chimpanzees with the full clinical picture of
Creutzfeldt–Jakob disease showed prominent delta activity in
their EEG but neither spikes nor periodic discharges (186).
The periodic discharge consists of a sharp wave or a sharp
triphasic complex of 100 to 300 msec duration with a repetition
rate of 0.5 to 2 Hz or intervals of 500 to 2000 msec. This is also a
short-interval periodic discharge. The discharges occur against a
severely disordered background of activity, mostly in generalized
synchrony. Some cases show unilateral onset with discharges over
one hemisphere or lateralized to one hemisphere for several days
or weeks (187). According to Tariska et al. (188), the periodic dis-
charges may even remain unilateral in the course of the disease.
The periodic activity usually shows a maximum over the ante-
rior region except for the Heidenhain form, which has a posterior
maximum (Fig. 26.15); in this special form, cortical blindness is a
common feature. In sleep, the periodic discharges tend to disap-
pear. The absence of periodic discharges after 10 weeks of illness
militates against Creutzfeldt–Jakob disease (189). EEG studies of
Chiofalo et al. (190) are based on the observation of 27 cases. The
differential diagnosis of the combined clinical and electroen-
cephalographic picture of this disease is quite complex (191).
Figure 26.15 Periodic discharges in Creutzfeldt–Jakob disease. Note
the mild lateralization of the recurrent sharp waves to the left. The 68-
year-old patient was demented, with psychotic features; there were long
tract signs, fasciculation, and frequent myoclonus. Myoclonic jerking
did not alter the tracing.
Myoclonus is a typical clinical feature of this disease. The peri-
odic patterns may or may not be associated with myoclonus (192).
In addition to periodic discharges, cyclic changes may also
occur, and the periodic activity may get temporarily replaced by
diffuse slow activity in the delta and theta range (193).
The periodic EEG discharge of Creutzfeldt–Jakob disease may
be caused by other conditions in rare instances. Hypothyroidism
is mentioned as a treatable cause of this pattern.
It has to be emphasized that the periodic discharges of
Creutzfeldt–Jakob disease (i) do not occur in every case of
this disease (even with numerous repeat-recordings) (194) and
(ii) may occasionally be found in other disorders.
Periodic Lateralized Epileptiform
Discharges (PLEDs)
Chatrian et al. (36) gave an extensive account of periodic later-
alized or focal discharges that may occur in a variety of acute
neurologic conditions. This pattern is most often associated
with acute cerebral infarctions but may also occur in neoplastic,
inflammatory, and epileptic disorders, especially status epilepti-
cus. Periodic discharges occurring in herpes simplex encephali-
tis are discussed earlier in this chapter. PLED activity, however,
may also occur in other types of encephalitis—among others,
in infectious mononucleosis encephalitis, according to Aminoff
et al. (195). As to acute vascular pathology, watershed-type
 Chapter 26
infarctions are the most common structural substratum in
patients with PLED. Markand and Daly (196), Dauben and
Adams (197), and Bauer et al. (198) presented further impor-
tant work on this pattern, which also plays a major role in
experimental cerebral embolism (199).
PLEDs may be simple and large sharp waves or complex
(compounded) discharges with mixed spiky and slower ele-
ments. The amplitudes are usually around 100 to 300 V, but
may occasionally be much higher. The firing rate may be as fast
as 3 Hz or as slow as 12/min (36) (Fig. 26.16).
The discharge is found over the maximally involved area, and
the local background activity is almost always severely disor-
dered. The focal maximum is customarily located over the
boundary zone between middle and posterior cerebral arteries
and thus over the posterior temporal region and its immediate
vicinity. Watershed infarctions are seldom found between the
territories of middle and anterior cerebral arteries, with a maxi-
mum of PLED activity over the superofrontal region.
The term PLED implies lateralization. It is possible, however,
that PLED activity is generated independently over both hemi-
spheres. De la Paz and Brenner (200) have investigated the bilat-
eral type of PLED referred to as “BI-PLED” (also “bi-PLED”).
Although strokes were found to be the leading cause in the
PLED group, CNS infection and epileptic seizure disorders,
especially status epilepticus, predominated in the BIPLEDs
group. Bilateral synchronous (generalized synchronous) peri-
odic activity should not be listed as BIPLED in view of the orig-
inal definition of Chatrian et al. (36). This distinction, however,
is not generally observed (201). Filley et al. (202) have broken
down their large material of PLEDs into typical and “indeter-
minate” discharges. The term cerebral bigeminy has been used
by Aldrich and Pugh (203) for bilaterally independent PLED
activity. Even TRI-PLEDs, arising from three different areas,
have been reported (left frontal, left posterotemporal, and right
posterotemporal) in an 85-year-old patient with renal
encephalopathy (204).
PLEDs are often associated with simultaneous focal motor
twitching in the contralateral face or fingers, hand, arm, leg,
foot, etc. This underscores the paroxysmal character of the pat-
tern. It is usually a temporary pattern that changes into other
abnormalities within 1 to 2 days. Patients with PLEDs are in
most cases acutely ill and often show a history of a variety of
mixed problems, such as cerebral arteriosclerosis plus renal
insufficiency or chronic alcoholism or diabetes mellitus. The
possibility of an underlying neoplasm must also be considered.
Patients with PLEDs are commonly older adults; this pat-
tern occasionally occurs in young adults and children (205).
Andriola (206) noted PLEDs as well as BIPLEDs in 12 children
with various types of acute CNS disease; those with bilateral
activity (BIPLED) expired in the acute state. Ritaccio and
March (207) reported the association of BIPLEDs with com-
plex partial status epilepticus. Silbert et al. (208) described
another variant of PLEDs with independent ipsilateral peri-
odic lateralized discharges. A clinical curiosity appeared to be
the observation of Chabolla et al. (209): a young man suffering
from multiple sclerosis (MS) with exacerbations ushered in by
complex partial status epilepticus associated with an enhanc-
■ Seizures and Epilepsy in Infants to Adolescents 493
ing right frontal MRI lesion and PLEDs in the EEG. Another
report on PLEDs in MS (right temporal PLEDs and MRI
lesion) has been presented by Gandelman-Marton et al. (210).
We have seen PLEDs in a child with hemiplegic migraine fol-
lowing a seizure.
A very thoughtful study of PLEDs deserves special attention:
a critical review by Pohlmann-Eden et al. (211). These authors
consider PLEDs as “an EEG signature of a dynamic pathophys-
iologic state in which unstable neurobiologic processes create
an ictal–interictal continuum, with the nature of the underlying
neuronal injury, the patient’s preexisting propensity to have
seizures, and the coexistence of any acute metabolic derange-
ments all contributing to whether seizures occur or not.”
Pohlmann-Eden et al. have strongly reemphasized the domi-
nant etiologic role of strokes in the generation of PLEDs. From
the evidence of functional neuroimaging studies, these authors
feel that PLEDs “might reflect a key pattern for focal hyperex-
citability in the penumbra zone of ischemic stroke.”
A nonlateralized variant of PLEDs are periodic epileptiform
discharges occurring in the midline (PEDIM), as cited by
Westmoreland et al. (212). These midline discharges are also
presumed to arise from watershed areas: anterior and middle,
middle, and posterior cerebral arteries.
Periodic Discharges in Acute Cerebral Anoxia
Repetitive simple or compounded sharp waves in generalized
synchrony may occur against a flat (or at least seemingly flat)
background of activity in patients with acute cerebral anoxia.
Myoclonus is often found in association with this type of activ-
ity (213–215). In most cases, these discharges are probably
aborted bursts in a suppression-burst pattern. True periodic
discharges, however, may also occur (Fig. 26.17).
Periodic Discharges of Other Etiologies
Periodic discharges may occasionally occur in other conditions,
but such observations are rare. The presence of such discharges,
whether regional or generalized, merely underscores the severity
of the basic conditions. Figure 26.18 demonstrates periodic dis-
charges over the right posterior quadrant in a 15-year-old girl
with a history of Dandy–Walker syndrome, numerous shunt
operations, shunt infections, and a severe epileptic seizure disor-
der. These rhythmical discharges also show the tempo of propa-
gation over the right posterior quadrant of the cerebrum.
SPIKES AND OTHER PAROXYSMAL
DISCHARGES IN HEALTHY CHILDREN
Focal or generalized paroxysmal discharges may occur in appar-
ently healthy individuals in all age ranges. These findings may be
termed as “false positives,” but the EEG abnormalities are real
and such spikes are evidence of an underlying epileptic patho-
physiology that may or may not subsequently become manifest.
Let us contemplate the indubitable fact that a complete medical
evaluation will yield certain abnormalities in almost every
healthy individual. What the electroencephalographer needs in
such cases is a commonsense philosophy as a basis for a wise
interpretation. There are many examples of seemingly irrelevant
494 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Figure 26.16 A: Periodic lateralized epilep-

tiform discharge (PLED) in a 70-year-old
patient, widespread over left hemisphere
with left midtemporal maximum. Note the
poorly developed background of activity.
Cerebrovascular accident associated with
twitching of right arm. Patient lethargic and
obtunded. B: PLED over right central–
midtemporal region, sometimes also involv-
ing the right parietal area. Note the well-
preserved physiologic activities of waking
and light drowsiness. Also note the complex
and multiphasic character of the discharge.
Patient is 68 years old with cerebrovascular
disease with right centroparietal low-density
area in CT scan (infarction).
 Chapter 26
Figure 26.17 Comatose patient (acute cerebral anoxia), age 57.
Constantly repetitive compounded spike discharges against a virtually
flat background of activity. These discharges were associated with occa-
sional generalized myoclonus.
and yet unmistakably present abnormalities that the prudent,
seasoned physician will integrate into a holistic view of the indi-
vidual. Similarly, EEG does not differ from the rest of medicine.
The EEG evaluation of comparatively large healthy popula-
tions usually shows a certain percentage of abnormalities.
Thorner (216) showed paroxysmal discharges in 0.3% of 1100
flying cadets; Gibbs et al. (217) found epileptic abnormalities in
0.9% of normal adult controls; Harty et al. (218) observed such
abnormalities in 3% of candidates for medical service; and
Williams (219) noted “larval epileptic disturbances” in 9% of
healthy controls. A study of 682 Air Force applicants showed
paroxysmal changes in 2.6%; about one fourth were focal and
the rest were of nonfocal character (220).
Zivin and Ajmone Marsan (221) and Chatrian (222) have
contemplated the clinical significance of spikes in healthy per-
sons. Above all, the interpretation must take into consideration
age. In childhood, the occurrence of central–midtemporal (also
parietal) spikes is associated with overt seizures in only 50% to
70% of the cases; this pertains mainly to the age from 3 to 12
years. In occipital spikes (mainly age 3 to 5 years), the epilepto-
genicity is even lower. Major studies on this subject based on
large healthy populations were performed by Nekhorocheff
(223), Kellaway and Fox (224), Brandt and Brandt (225), Corbin
and Bickford (226), Trojaborg (227), and Eeg-Olofsson et al.
(228). In general, “benign” focal spikes (such as seen in benign
rolandic epilepsy) outnumber generalized synchronous bursts
of spikes or spike waves by a slight to considerable margin.
The study in children by Cavazzuti et al. (229) demands spe-
cial attention. EEG recordings were performed on 3726 children
from 6 to 13 years of age without evidence of seizure or neuro-
logic deficits. Paroxysmal discharges were found in 131 cases
■ Seizures and Epilepsy in Infants to Adolescents 495
Figure 26.18 Waking record of a 15-year-old girl with Dandy–Walker
syndrome in combination with a brainstem lipoma. The patient suffers
from a severe epileptic seizure disorder; she had been treated with
numerous shunt operations. Against a diffuse background of disorgan-
ized slow activity, a periodic discharge of “blunted” sharp character is
very prominent over the right posterior quadrant. Both bipolar and
unipolar derivations show a focal origin in the right posterotemporal lead
(T6) with prominent spread into right occipital (O2) and right parietal (P4)
regions. This spread is so slow that the traveling character of the periodic
discharges can be gleaned from bipolar and referential derivations.
(3.52%). Generalized, mainly polyspike, wave discharges were
found in 41 cases, midtemporal spikes in 50, centroparietal
spikes in 27, occipital spikes in 2, and bilateral (midtemporal or
parietal) spikes in 11 children. Changes in spike location and
shifts from focal to generalized discharges were rare. The follow-
up period was up to 9 years. The abnormalities disappeared in
most, usually during elementary school age or in early adoles-
cence. Only seven children developed clinical epileptic seizures;
five had generalized discharges, one had midtemporal, and one
had centroparietal spikes. In two subjects there was evidence of
epileptic seizures in the family, while six children had siblings
with spikes, either generalized or rolandic.
Both generalized synchronous (spike wave, polyspike wave)
and rolandic (centroparieto–midtemporal) spikes in nonepilep-
tic children suggest a genetic predisposition if no neurologic
deficit and no history of insult to the CNS are present. In chil-
dren with a history of cerebral palsy and with no seizures but
prominent spiking, the spike activity may herald future
epileptic seizures (83). Even in perfectly healthy children with
spikes, the possibility of future seizures cannot be completely
ruled out. In addition, epileptiform abnormalities may be an
496 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
epiphenomenon associated with an underlying neurologic
disorder, such as attention deficit disorder, headaches, or
behavioral problems, and psychiatric disorders, to name a few.
This section must be capped by a strong plea to clinicians to
refrain from rash and ill-advised statements that a person
without seizures has epilepsy and must be treated because of
spikes in the EEG. There may be need for further medical
attention, and in some circumstances a repeat EEG may be
prudent but anticonvulsive treatment is seldom needed and
indeed is ill-advised in most cases.
Types of Epileptic Seizures
The character of an epileptic seizure is more or less strongly
determined by the chief cerebral area involved. Another impor-
tant factor is the underlying basic epileptic condition. As a
matter of fact, the underlying basic condition may be of para-
mount significance, especially in an infant with infantile
spasms—hypsarrhythmia and jackknife seizures, which are not
found outside this basic epileptic condition, or in a child with
Lennox–Gastaut syndrome and atonic drop attacks, which are
germane to this condition and alien to others (also see Ref. 230).
These basic epileptic conditions are age-determined or age-
dependant; therefore, age is an important factor that is capable of
modifying the character of epileptic seizures. Furthermore, the
nature and extent of an underlying cerebral lesion may be a neg-
ligible factor in the determination of the seizure type in the group
of age-determined epileptic conditions.
Certain types of seizures, such as focal motor and psy-
chomotor seizures, are highly indicative of special cerebral
areas, whereas others are of much less localizing value. The
localization depends heavily on such confirming EEG evidence
as local spikes and recorded ictal episodes and, in some special
cases, on the demonstration of regional changes with depth
electrodes. Structural neurodiagnostic tests are needed for the
demonstration of significant morphologic changes.
Focal seizures may be a prelude to an ensuing grand mal
convulsion. The term aura has been used for such initiating
focal seizures, especially for attacks impinging on afferent sys-
tems (visual, acoustic, olfactory, somatosensory, gustatory, and
so forth).
Simultaneous EEG recording and fMRI is beset with prob-
lems since the EEG is being obtained in a high-field magnet and
during scanning. Baudewig et al. (231) have reported a method
demonstrating simultaneous blood-oxygenation-dependent
(“BOLD”) MRI sequences and EEG activity with very little arti-
factual distortion (in epileptic patients).
Classification of Epileptic Seizures
The classification of epileptic seizures is a thankless job. The
preferences of “splitters” and “lumpers” must be reconciled in
the tedious work of an international committee entrusted with
this task by the International League Against Epilepsy, pro-
gressing from the older classifications of reputed epileptolo-
gists that have sometimes been marred by personal biases. The
most commonly used classification had been proposed by
the Commission on Classification and Terminology (232).
Although widely used and officially recommended (and essen-
tially reiterated by Dreifuss (233)), this classification did not
find universal acceptance. This may have been due to an excess
of terms and the use of cumbersome formulations. Also, in
pediatric practice the classification required several modifica-
tions for practical application. It was not always possible to
differentiate simple from complex focal seizures because
immaturity, limited communication skills, and autistic fea-
tures could prevent the accurate assessment of alteration of
consciousness in some children (234). Infantile spasms were
missing from the classification, as were myoclonic–astatic or
myoclonic–atonic seizures. A breakdown into small ictal
behavioral or ictal-EEG detail may render a classification more
nebulous. There are, of course, limitations for every attempt at
seizure-type classifications. The classification of 1981 could
not accommodate a considerable number of seizures, which
then were listed as unclassifiable because they did not fit the
scheme. A recent recommendation from the ILAE taskforce on
classification suggested a marked simplification of the seizure
classification and tried to address some of these issues (235)
(Table 26.1).
Generalized Tonic–Clonic Seizure (Grand Mal)
Clinical Manifestations
The sequence of clinical manifestations has been masterfully
described by Gowers (236, see also republication in 1964) and is
given in detail below. Further detailed descriptions were given by
Janz (237), Gastaut and Broughton (238), and Karbowski (239).
The generalized tonic–clonic (GTC) seizure lasts about 40 to
70 seconds, or sometimes up to 90 seconds. Initial massive
Tabl e 2 6 . 1
Classification of Seizure Types
Generalized seizures
Tonic–clonic (in any combination)
Absence
Typical
Atypical
Absence with special features
Myoclonic absence
Eyelid myoclonia
Myoclonic
Myoclonic
Myoclonic–atonic
Myoclonic–tonic
Clonic
Tonic
Atonic
Focal seizures
Unknown
Epileptic spasms
 Chapter 26
generalized tonic spasm is sometimes initiated with a cry (“a
wild, harsh, screaming sound,” 236). It is immediately associated
with loss of consciousness (profound coma); the arms are usually
in semiflexion and the legs in extension. After about 10 to 20 sec-
onds, it is supplanted by the clonic phase in which “the vibratory
phenomenon (noted with palpation of the extremely tense mus-
cles) becomes sufficiently prolonged to interrupt completely the
tonic contraction” (238).
This leads to a succession of brief and violent flexor spasms of
the entire body. Accompanying apnea leads to a grayish livid
complexion, while the rhythmical clonic spasms slow down until
a final massive myoclonus marks the end of the seizure.
According to Gastaut and Broughton (238), tongue biting usually
materializes during the clonic phase. Mydriasis, arterial hyper-
tension, and tachycardia accompany the attack. Enuresis usually
occurs at the termination of the seizure (238); there may occa-
sionally be fecal incontinence. The patient is completely flaccid
after the last clonic jerk, after which respiration returns. Only a
few seconds after the beginning of the flaccid, immediate postic-
tal phase, a tonic muscle spasm returns that is most intense and
prolonged in the masseter (238). This postictal trismus tem-
porarily blocks respiratory effect. The respiration becomes regu-
lar in the ensuing recuperative phase, during which the patient
returns to consciousness unless he slips directly into a period of
postictal sleep. A fall caused by a GTC can be traumatizing; even
epidural hematoma has been reported (240).
Tonic–clonic seizures may occasionally show unilateral
predominance of the clinical manifestations. Gastaut and
Broughton (238) feel strongly that these attacks (which are more
common in children and infants) are in fact generalized seizures
with unilateral expression. Do attenuated or mitigated grand
mal seizures exist? The answer is affirmative according to
■ Seizures and Epilepsy in Infants to Adolescents 497
Karbowski (239). These attacks are characterized by shorter
duration and rudimentary tonic or clonic phases, possibly due
to the effect of anticonvulsive medication.
EEG Manifestations
The tonic–clonic seizure is initiated by an abrupt loss of voltage
(desynchronization, electrodecremental period) of a few seconds
duration; there is evidence of very fast (20 to 40 Hz) activity in all
leads. In patients with primary generalized epilepsy, several gen-
eralized bursts of polyspike-wave complexes with massive bilat-
eral myoclonus may precede the phase of desynchronization.
Muscle activity rapidly obscures the recording; the vertex
derivation, however, may remain artifact-free due to the lack of
underlying muscles. Informative recordings can be secured only
from patients with muscle relaxation from curarization and
artificial respiration. Removal of muscle artifact by means of
digital filtering has been achieved by Gotman et al. (241).
After the phase of desynchronization, which may be as short
as 1 to 3 seconds, rhythmical activity at about 10 Hz with rap-
idly increasing amplitude dominates the EEG. Gastaut and
Broughton (238) have laid much stress on this frequency
(“epileptic recruiting rhythm,” 242), which is better discernible
with the use of automatic frequency analysis.
About 10 seconds after the onset of a seizure, slower fre-
quencies are noted, gradually slowing into the theta and delta
range. Once the frequency of 4 Hz is reached, “each slow wave
interrupts the recruiting rhythm, giving rise to polyspikes and
wave complexes, themselves decreasing in frequency” (238).
The clonic activity corresponds with generalized polyspike
bursts at each myoclonic jerk (Fig. 26.19).
The last clonic contraction is followed by postictal flatness
for several seconds (243). Very slow irregular delta activity
Figure 26.19 The EEG correlate of a grand mal
(generalized tonic–clonic) seizure, recorded in a
curarized patient and almost free of superim-
posed muscle activity. The bottom channel of
this continuous recording shows the deflections
of a frequency analyzer (after W. Grey Walter).
The seizure activity is characterized by fast spik-
ing of increasing amplitude during the tonic
phase. There are repetitive bursts of spikes mixed
with persisting weak muscle potentials during
the clonic phase (lower portion). There is unre-
markable EEG activity prior to the seizure; the
termination of the convulsion is characterized
by general voltage depression. (From Gastaut H,
Broughton R. Epileptic Seizures. Springfield, IL:
Charles C Thomas; 1972, with permission.)
498 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
(“postseizure stupor,” 37) then dominates the EEG, with grad-
ual frequency increase into the theta and alpha band; the
appearance of an organized posterior alpha rhythm signals the
return to the waking state.
Prevalence and Age Factor
Tonic–clonic seizures are common at any age, except for the
first year of life, during which they are largely absent (244).
Insufficient myelination of the brain probably precludes them
at this early age. In the research of Gibbs and Gibbs (37), 5598
patients of a population of 11,612 people with epilepsy (48.2%)
had grand mal only; another 3290 patients (28.3%) had grand
mal in combination with other types of seizures.
Neurophysiologic Aspects
One is tempted to regard the tonic–clonic seizure as a standard-
ized maximal and global epileptic response of the brain.
Electroconvulsion studies in the cat, however, have shown that
the cerebellum and the lower brainstem do not fully participate
in the ictal activity (245). The view of a standardized all-out
response of the brain requires some correction. Schmidt and
Wilder (1968) feel that GTCs “vary in their severity and in the
degree to which they incorporate the various tonic and clonic
phases.” Variations of the degree of cerebral cortical participa-
tion in tonic–clonic seizures have been demonstrated in the cat
as well as in the human by Rodin et al. (246). This view has been
supported by the experimental findings of Petsche (247).
This implies that tonic–clonic seizures are graded rather than
maximal cerebral epileptic responses. These variations could
explain differences in the response to electroconvulsive therapy
in psychiatric illness; following the electrically induced GTC, the
EEG may show little or marked slowing, probably depending on
the degree of neuronal participation in the seizure.
Quite different is the view of Gastaut and his coworkers. The
emphasis placed on rhythmical 10 Hz activity during a large
portion of the grand mal attack has been discussed; this “epilep-
tic recruiting rhythm” will follow stimulation only in nonspe-
cific reticular thalamic structures projecting diffusely to both
hemispheres over still uncertain connecting pathways. It was felt
that the motor, autonomic, and EEG phenomena of the tonic
phase could be explained by massive discharge of the thalamic
and subthalamic brainstem reticular structures; this would also
account for the loss of consciousness. In the clonic phase, the
appearance of slow waves has been attributed to cortical
inhibitory systems via thalamic and lower brainstem structures
or, more specifically, via a thalamocaudate circuit branched
from the thalamocorticospinal system. These views have been
laid down by Gastaut et al. (248), Gastaut and Fischer-Williams
(242), and, in condensed form, by Gastaut and Broughton (238).
Absences (Petit Mal)
Terminology
The term petit mal arose from the jargon of physicians and
attendants in the hospitals of Paris early in the 19th century.
According to Temkin (249), Esquirol (250) distinguished more
or less severe epileptic attacks as “grand mal” and “petit mal,”
but his definition of these two terms “is obviously vague.” The
term absence was introduced by Calmei (251). The modern
nomenclature (Gastaut et al., 1970) recommends the term
absence, but the term petit mal is deeply rooted. Petit mal
absence seems to be an acceptable compromise.
Clinical Manifestations
The petit mal absence occurs mainly in children older than 3
years, with a declining incidence throughout adolescence and
early adulthood; persistence beyond middle adult life is very rare.
The attack consists of a sudden lapse of consciousness with
impairment of mental functions. Its usual duration ranges from
5 to 20 seconds; longer absences may occur over 1 to 2 minutes,
but such an unusual length is a sign of a somewhat complicated
epileptic seizure disorder with some degree of automatism-like
ictal behavior. The absence is associated with interruption of
ongoing activity and, due to moderate impairment of con-
sciousness, the patient is unable to see or hear. There is usually
a blank facial expression that contradicts the frequently used
term staring spells; a true stare is more likely to occur in tempo-
ral lobe epileptics with psychomotor seizures. The eyes drift
upward (“star gazing,” 237,252); rhythmical beating of the eye-
lids at 3 Hz is very common and may be the only apparent
motor manifestation of the attack. Marked orofacial move-
ments during the absence are suggestive of a more complex
type with poorer response to therapy.
Petit mal absences show a wide variety of mild-to-moderate
motor accompaniment; rhythmical eye blinking in synchrony
with the spike waves is the most common motor component.
Retropulsion of the head is quite common (“retropulsive petit
mal,” 237); adversive movements and some rhythmically repet-
itive oral motions may also occur. Automatisms occurred in
163/405 (40%) of children and were more likely with hyperven-
tilation and with longer seizures, occurring in only 23% of
spontaneous awake seizures (253,254). The most common clin-
ical manifestations include swallowing, mouthing, chewing,
lip smacking, and lip licking and occurred across all absence
syndromes.
There is good reason to assume that the unique nature of
loss of consciousness during the classical absence is based on a
temporary suspension of the “working memory” due to the
powerful accentuation of spike-wave activity in the frontal cor-
tex (255–257). This explains the unmatched immediate restora-
tion of the working memory along with consciousness at the
end of the absence (no other form of brief loss of conscious-
ness—syncope, for instance—reverses so quickly). As to the
crucial function of working memory, see Ref. 69.
The postural control is grossly maintained, but swaying and
stumbling movements may be noted in the standing patient
(Schmidt and Wilder, 1968). The posture may be altered by sus-
tained or saccadic retroflexion of the head (“retropulsive petit
mal,” 258). Autonomic and especially vasomotor changes are
common during the petit mal absence (259,260).
Impairment of the level of awareness has been the object of
some studies (260–265). A variety of tests, such as rhythmically
pressing a button, have been used to demonstrate the lapse of con-
sciousness (266). Mirsky and Tecce (267) demonstrated the per-
sistence of visual evoked potentials during spike-wave discharges.
 Chapter 26
Absence seizures are divided into typical absence and atypi-
cal absence seizures. Typical absence seizures are classified as
simple or complex; a simple absence has a sudden onset without
motor activity; complex absences have associated motor activity
or autonomic activity. In a typical absence seizure, it is usually
not difficult for an observer to identify the electrographic seizure
onset and the seizure cessation by clinical observation.
The atypical absence seizure has a less clearly defined onset
or cessation, so that it may be difficult to identify the beginning
or the end of the seizure without EEG. The atypical absence
seizure may have more pronounced tone changes, a longer
duration, and are typically associated with other seizure types
and mental retardation. In contrast to the typical absence
seizure, it is usually very difficult for an observer to identify the
electrographic seizure onset by clinical observation alone.
Penry et al. (268) studied 374 absence seizures with video-
EEG in 48 patients; a simple absence occurred in only 9% of
patients; automatisms occurred in at least one episode in 88%,
clonic components occurred in 41%, and seizures lasted less
than 20 seconds in duration in 85%. Automatisms may be per-
severative or occurring de novo, and autonomic abnormalities
may consist of pupillary dilatation, pallor, flushing, salivation,
or incontinence. In a recent study, absence seizures ranged
from 1 to 44 seconds in duration, the average duration was 9.4
seconds, and only 65% lasted between 4 and 20 seconds and
66% had major impairment in awareness; no myoclonia,
except in the face, occurred in 74%, no photic induction
occurred in 83% (269).
The EEG in a typical absence seizure shows a symmetric 3 Hz
generalized spike and wave discharge, which has a faster fre-
quency at the beginning and a slower frequency at the end. The
atypical absence seizure has a frequency between 1.5 to 2.5 Hz
■ Seizures and Epilepsy in Infants to Adolescents 499
(usually slower), may be asymmetric, and usually has a slow
background (see above).
EEG Manifestations
The ictal EEG of the absence seizure is characterized by the gen-
eralized synchronous 3 Hz spike-wave discharge. No typical
absence seizure can occur without this classical pattern, but the
3 Hz spike-wave pattern may occur, usually in bursts of less
than 5-second duration, without an accompanying clinical
absence seizure. This should serve as a stern warning not to
equate generalized 3 Hz spike waves with absence seizures; the
latter cannot occur without the former, but the former fre-
quently occur without clinical absences (52–54) (Fig. 26.20).
The spike-wave discharge is maximal over the frontal mid-
line and may start at a rate of around 4 Hz, quickly slow down
to 3 to 3.5 Hz, and, during the final phase of the attack, slow
down to about 2.5 Hz. Onset and termination are abrupt; the
attacks may be preceded and immediately followed by normal
EEG activity, especially when recorded in the waking–resting
state rather than during hyperventilation or sleep.
There may also be a regularity and organization to the
generalized spike-wave discharges; another term has been frag-
mentation, describing spike-wave discharges that occur in frag-
ments, rather than in a continuous burst.
In the rare cases of the absence seizures of middle or old age,
the spike-wave complexes are somewhat less impressive, with the
spike component being slower and less prominent (Fig. 26.20).
Precipitating Factors
Hyperventilation is an extremely potent activator of the 3 Hz
spike-wave discharge with or without clinical petit mal absences.
Further powerful facilitating mechanisms are non-REM sleep
Figure 26.20 Petit
mal absence, in an
8-year-old patient.
The spike-wave burst
is preceded by an
aborted run of rhyth-
mical posterior slow
3 Hz waves; after
two slow waves, the
attack begins.
500 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
and hypoglycemia. In some cases, intermittent photic stimula-
tion is very effective.
Age Factor and Prevalence
Petit mal absences frequently start at the age of 3 years, are
almost never found prior to the age of 2.5 years, and if so, carry
a poor prognosis (270). Early onset absence seizures have been
associated with the glucose transporter defect (De Vivo syn-
drome) (271). In 34 cases of absence seizures in an age less than
4 years, 4/34 had a Glut-1 deficiency (272). In a 24-hour EEG,
2.5 to 4 Hz spike-wave discharges occurred in 41%, and gener-
alized slowing or attenuation, or no background abnormalities
each occurred in 34% (271). Focal discharges were more com-
mon in younger children, less than age 2 years, whereas gener-
alized discharges were more common in children 2 to 8 years
(271). The EEG may improve when the child is postprandial, so
an EEG before and after feeding may demonstrate a decrease in
epileptiform features (273).
Some of the children with absence epilepsy may have a his-
tory of febrile convulsion in infancy. There is decline of petit
mal epilepsy during the second decade, aside from treatment
effects of drastic reduction of attacks. In some cases, the
absences start around age 9 to 10 years; German schools (see
Ref. 237) have studied the differences in the course of children
with onset earlier (around 4 years) and later (9 to 15 years).
This and the occasional persistence of petit mal into adult life
are discussed below (274,275).
Janz et al. (276) have de-emphasized the significance of the
dichotomy of pyknolepsy and juvenile absences. In 163 patients
with absence epilepsy, Trinka et al. (277) reported that 22%
were in an “overlap” group.
A clinical absence with generalized 3 Hz spike waves was
triggered by tapping the head of a 2½-year-old child. This well-
documented case of DeMarco (278) makes one wonder if such
an unusually early manifestation (and precipitation) belongs in
the category of primary generalized epilepsy.
The prevalence can be gleaned from the figures of Gibbs and
Gibbs (37). Among a total of 11,612 patients with epilepsy, only
335 (2.9%) had absence seizures without other seizure types.
There were 896 patients (7.7%) with absence seizures in combi-
nation with other seizure types; 706 of them (6.1%) had a com-
bination of absence and generalized tonic–clonic seizures.
These figures increase moderately when one deals with a popu-
lation of children and adolescents.
Very prolonged absence-like stages, attacks of petit mal–like
stupor, or petit mal automatisms have been termed petit mal
status (279), whereas the modern terminology recommends the
term absence status. These states are discussed in the section on
status epilepticus. Whereas all of the ictal manifestations of pri-
mary generalized epilepsy tend to occur in children and adoles-
cents, the absence status not only occurs in elderly patients but
may even have its onset in old age.
Neurophysiologic Aspects and Clinical Significance
These subjects are discussed in the section on primary general-
ized epilepsy.
Myoclonic Seizures
Myoclonus is characterized by a rapid involuntary muscle con-
traction, subtle or massive, usually with locomotor effect, gen-
eralized or limited to certain muscular segments, mostly
predominant in flexor muscles, and more pronounced in upper
extremities. The resulting jerk may be synchronous or moder-
ately asynchronous.
Myoclonus may be epileptic or it may represent a dyskinetic
disturbance caused by the breakdown of the motor control sys-
tems of the cerebellum. A combination of epileptic and appar-
ently nonepileptic myoclonus may occur. The term myoclonic
stresses the singular nature (one distinct jerk), whereas clonic
refers to the repetitive type of muscular twitching.
The complexities of the underlying neurophysiologic mech-
anisms can be gleaned from the work of Watson and Denny-
Brown (280), Halliday (281–284), Shibasaki and Kuroiwa (285),
Chadwick et al. (286), Hallett et al. (287), and Kelly et al. (288).
Progress in myoclonus research was sparked by the introduc-
tion of myoclonus-triggered back averaging of the concomitant
EEG (285, also independently developed by Chadwick et al.
(286)). Hallett (289) has summarized the most recent insights
into the physiologic nature of myoclonus. Three major mecha-
nisms have been pointed out: (i) cortical reflex myoclonus,
often occurring in focal motor epilepsy and associated with
giant somatosensory evoked potentials (based on hyperex-
citability of the sensorimotor cortex; (ii) reticular reflex
myoclonus, thought to be due to hyperexcitability of the caudal
brainstem reticular formation; EEG spikes are not time locked
with myoclonus and are maximal over vertex; (iii) primary gen-
eralized epileptic myoclonus, as found in primary generalized
epilepsy with bilateral predominantly frontocentral EEG event
preceding the myoclonus. Hallett’s (289) subdivision of
myoclonic phenomena is strictly physiologic and not etiologi-
cally oriented. The authors make it clear that either cortical
reflex myoclonus or reticular reflex myoclonus occur in clinical
disorders associated with myoclonus (Table 26.2).
The clinical–semiologic details of the myoclonic movements
have been extensively described by Gastaut (290) and cine-
matographically demonstrated by Oller-Daurella and Oller-
Ferrer-Vidal (266,291).
Myoclonic phenomena are found in various epileptic con-
ditions such as primary generalized epilepsy, infantile spasms
with hypsarrhythmia, and Lennox–Gastaut syndrome, and in
degenerative CNS disease, such as Tay–Sachs disease, forms of
CNS lipidosis, essential hereditary myoclonus epilepsy
(Lafora–Unverricht–Lundborg), encephalitis, acute or chronic
renal failure, acute cerebral anoxia, or postanoxic states. More
extensive studies on the clinical significance of myoclonus
have been presented by Weingarten (292), Aigner and Mulder
(293), Gastaut (290), Bauer (294), Kinsbourne and Rosenfield
(295), Sulibhavi and Schneck (296), Farrell and Swanson (297),
Charlton (298), Niedermeyer et al. (299), and Aicardi (291).
It has been pointed out that, in neonates, nonketotic
hyperglycinemia represents an important cause of early
myoclonic epilepsy. In infants and young children, progressive
myoclonic encephalopathies are usually due to metabolic or
 Chapter 26 ■ Seizures and Epilepsy in Infants to Adolescents 501

Tabl e 2 6 . 2

Characteristics of Major Neurophysiologic Types of Myoclonus a

Cortical Reflex Reticular Reflex Primary Generalized

Myoclonus Myoclonus Epilepsy Myoclonus
Clinical setting of Fragment of focal (partial) Fragment of bilateral or Fragment of primary generalized
occurrence epileptic seizures generalized epileptic epilepsy
manifestations (aside
from primary generalized
epilepsy)
Localization of Involves only a few adjacent Myoclonus tends to affect Small bilateral myoclonic, often
myoclonus muscles the whole body fingers only, minimyoclonus
(mimicking tremulousness); also
major whole-body myclonic jerks
Precipitation Spontaneous or induced Spontaneous or induced No extrinsic precipitation except for
(accentuated) by voluntary (accentuate) by voluntary photosensitivity
movement (motor action) movement (motor action)
Bilateral synchrony Asynchronous; usually Usually synchronous, but Synchronous (fairly precise bilateral
of myoclonus unilateral synchrony may be quite synchrony)
imperfect
Somatosensory Enhanced (“giant SSEP”) Unremarkable Unremarkable
evoked potentials
(SSEP)
Focal time-locked Well demonstrable, with EEG event (spike) often Negative event precedes myoclonus,
EEG event preceding spontaneous and associated myoclonus but with bilateral frontocentral
reflex-induced jerks located not time locked; spike maximum, lasting 30–100 msec for
over appropriated motor generalized maximal over major jerks (well time locked) and
cortex region vertex 100–250 msec for minimyoclonus
(less well time locked)
CNS origin of Focal: motor cortex (local ? Caudal brainstem reticular Presumably cortical (frontal,
neurophysiologic hyperexcitability) formation supplementary motor cortex
event as starting point)
Clinical Focal motor (elementary Postanoxic action Primary generalized epilepsy and
significance partial) epileptic seizure, (intention) myoclonus, related epileptic seizure disorders
probably also Creutzfeldt– renal encephalopathy
Jakob disease and others (uremia), others

aModified from Hallett M. Myoclonus: relation to epilepsy. Epilepsia (New York). 1985;11:567–577.

genetic disorders: mitochondrial diseases, gangliosidoses and
ceroid-lipofuscinoses, and in older children, Lafora’s disease
and its variants are in the foreground as the cause of
myoclonus (300). This investigator has also emphasized that a
majority of myoclonic epilepsies do not belong in the group of
neurodegenerative disorders; consequently, their prognosis is
much better.
From the EEG viewpoint, myoclonus is associated with mas-
sive spike discharges and especially with bursts of bilateral or
generalized synchronous polyspikes in patients with primary
generalized epilepsy, and thus with the related photoconvulsive
response to intermittent photic stimulation, and also in patients
with Lennox–Gastaut syndrome. In infantile spasms with hypsar-
rhythmia (West syndrome), the EEG correlate of the myoclonus
is variable and reaches from sudden flattening or desynchroniza-
tion to massive spiking and unaltered ictal recordings. Myoclonus
may occur with or without spiking in degenerative CNS disease.
The term negative myoclonus was introduced by Tassinari (301)
and denotes brief repetitive lapses of postural tone time locked to
spikes over the contralateral central region (302,303). In some
patients, this phenomenon is associated with acute valproate
encephalopathy (valproate stupor). According to Rubboli et al.
(304), epileptic negative myoclonus is accompanied by an
inhibitory frontal spike component.
Table 26.3 shows the EEG correlates in a variety of clinical
conditions with myoclonic seizures.
502 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Tabl e 2 6 . 3
Clinical and EEG Characteristics of Chronic Conditions with Myoclonus
Primary generalized
epilepsy (synonyms:
common generalized,
centrencephalic,
corticoreticular
epilepsy)
Infantile spasms
Lennox–Gastaut
syndrome
Cherry-red spot-
myoclonus syndrome
Essential hereditary
myoclonus epilepsy
(Lafora–Unverricht–
Lundborg)
Benign myclonus
(Hartung) probably
related to
paramyoclonus
multiplex (Friedreich)
Postanoxic
myoclonus
Epileptologic Neurologic or
Age Symptomatology Mental Deficits EEG
Mainly 4–16 Petit mal absences None Principally generalized
years; in case Petit mal–grand mal- or bilateral anterior
grand mal, 10–50 myoclonus synchronous spikes,
years spike waves (3 Hz, 4 Hz),
Grand mal- polyspikes, activated by
myoclonus hyperventilation, flicker,
Absence (petit mal) sleep (non-REM)
status, often with
myoclonus
Age 4 months– Myoclonic, mainly Varying from normal Hypsarrhythmia: high-
2.5 years head nodding findings to forms of voltage output, irregular
(hypsarrhythmia) Jackknifing with cerebral palsy, slowing with massive
tonic component hydrocephaly, spikes, polyspikes, mostly
microcephaly, mental occipital maximum
retardation Asleep: enhanced bursts
with stretches of
depressed voltage
Onset mostly age Numerous minor Varying from normal Slow spike-wave
1–10 years, motor types of or more or less complex (“petit mal
becoming poorly seizures; tonic, severe neurologic variant” after Gibbs and
distinct after atonic–akinetic, deficits, types of coworkers), 1–2.5 Hz; in
age 30 myoclonic, also cerebral palsy, etc; sleep, runs of rapid
psychomotor, grand mental retardation spikes, 10–20 Hz; also
mal petit mal polyspikes; frequently
general slowing
Childhood to Massive myoclonus, Fundi: cherry-red- Positive spikes over
adolescence also facial; grand mal spot; relatively mild vertex in brief bursts
may occur neurologic deficit,
normal intelligence
Onset in Myoclonus Cerebellar Several disorganized,
childhood or Grand mal symptomatology of predominantly slow with
adolescence (the varying degree: numerous single spikes,
earlier the onset, dysmetria, often with needlelike
the more serious intentional tremor, spikes of myogenic
the progression) speech and gait character; few or no
disturbance; polyspikes; sometimes
progressive dementia more rhythmical slow
activity with spikes
Onset in Myoclonus None Mostly normal (a
childhood personal observation of
an abnormal record is an
exception)
Any age Myoclonus, intention Cerebellar deficit, Disorganized, with
or action type, EEG possibly other numerous spikes,
spikes maximal over posterior anoxic bilateral synchronous,
vertex deficits vertex or frontocentral
Dementia
 Chapter 26
Focal Seizures
Focal seizures are discussed in greater detail in the chapter on
epilepsy in adults. A few focal seizures that are particular to
childhood are mentioned here.
Focal Motor Seizures (Rolandic Motor Seizures)
Clonic twitching of the contralateral muscles of the body is the
typical seizure manifestation of a localized discharge within the
precentral gyrus. Bravais (305) studied focal motor seizures in
his thesis. Jackson’s work, however, and the classical electrical
stimulation studies of Fritsch and Hitzig (306), Ferrier (307),
and Luciani (308) have brought the excitability of the human
cortex into the right perspective.
In accordance with the somatotopic arrangement within
the precentral gyrus, the clonic movements are initially limited
to the corresponding area of the body and tend to spread dur-
ing the attack. Such a spread (for instance, from the facial mus-
cles to the leg) is well known as the Jacksonian march.
Such a Jacksonian march does not materialize in many cases
and the attacks remain limited to one body region, usually
thumb, fingers, lips, eyelids, or great toe. The cortical represen-
tation of these functions, according to Gastaut and Broughton
(238), represents a phylogenetic acquisition of humans, who
have been endowed with a particularly large cortical motor
zone. Any region may be involved; even trunk muscles may par-
ticipate. There is no impairment of consciousness during these
attacks. Involuntary lingual movements were described as “lin-
gual seizures” occurring in conjunction with contralateral cen-
trofrontal ictal spiking (309).
A status of constant clonic activity in one muscular region is
known as epilepsia partialis continua or Koshevnikov syndrome.
Bilateral focal motor attacks are very rare. Focal motor seizures
may be caused by an ipsilateral lesion (310), but such cases are
exceptional (paradoxical lateralization).
The ictal EEG is expected to show impressive spiking con-
tralaterally over the involved motor cortex, but such cases with
precise focal EEG documentation are seldom observed.
Widespread EEG changes with desynchronization, spiking,
and more or less rhythmical theta or delta activity may be
noted, or there may be a complete lack of ictal EEG activity,
also caused in part by movement artifact. The discharge may
be subcortical in a three-dimensional involvement of cortical
and deep structures (discussed under rolandic epilepsy), or
cortical spiking may be too small and desynchronized. The dis-
sociation between spiking over the motor cortex and corre-
sponding motor effects has been investigated by Elger and
Speckmann (311) in the rat; spiking in lamina V after local
penicillin gives rise to contralateral twitching without con-
comitant spiking on the cortical surface. With spike activity
in superficial and deep cortical layers, there are accompany-
ing motor effects. These authors also introduced the term
vertical inhibition for a mechanism responsible for the failure
of descending neuronal activity during epicortical epilepti-
form potentials (312). Quiet epileptogenic foci within the
motor cortex can be activated by transcranial magnetic stimu-
lation (313).
■ Seizures and Epilepsy in Infants to Adolescents 503
Adversive Seizures (Versive, Ipsiversive,
or Contraversive Seizures)
Adversive seizures may be limited to conjugated eye move-
ments; there is usually extreme lateral gaze to the side contralat-
eral to the lesion. The eye movement may be tonic (oculogyric)
or clonic (oculoclonic, also called epileptic nystagmus). These
manifestations must be carefully distinguished from nonepilep-
tic nystagmus and oculogyric crises such as brainstem disease
and basal ganglia disease.
More commonly, the adversion consists of deviation of head
and eyes; even the entire body may rotate to one side (usually
contralateral, i.e., opposite side of the focus). Sometimes, the
contralateral arm is raised in tonic extension, with the subject
apparently looking at his raised hand (238). This has been
impressively demonstrated by the illustrations of Aird and
Woodbury (314) and Matthes (315).
These attacks may occur with a loss of consciousness (fron-
topolar region, middle, and superior frontal gyrus) or in a
conscious state (posterior portion of frontal lobe, vicinity of
precentral gyrus) (316,317).
The ictal EEG usually shows more or less rhythmical slowing
with spikes over the affected area and, in most cases, over a large
neighboring region. For this reason, a precise focal EEG diagno-
sis is difficult or impossible.
Adversive seizure phenomena are not limited to the frontal
region; oculoclonic seizures may be frontal as well as occipital.
The occipital origin of combined oculoclonic and adversive
seizures has been clearly shown by Fossas et al. (318). The initi-
ation of grand mal seizures by a brief adversive movement is
quite common and cannot be regarded as a reliable localizing
sign. This has been further corroborated by Robillard et al.
(319), as well as by Ochs et al. (320). Further considerations of
this topic are found in the work of Quesney et al. (321).
Epileptic Spasm/Partial Seizures
A combination of flexor spasms with preceding or subsequent
partial clonic seizures has been described by Pachatz et al.
(322), showing in the EEG a juxtaposition of focal and general-
ized ictal activity. These attacks occur in childhood/adolescence
with cryptogenic or residual-acquired etiologies; their progno-
sis is relatively favorable.
Hypomotor Seizures
This term was introduced by Kallen et al. (323) and pertains
to seizures of infancy and childhood consisting of diminished
motor activity along with an undetermined level of con-
sciousness. These have also been referred to as behavioral
arrest seizures, and may often be accompanied by version of
the eyes (244). The EEG shows typical ictal changes of focal
seizures, often arising from the temporal, parietal, or occipital
lobes.
Seizures Occurring in Infantile Spasms
Only (Jackknifing, Salaam)
These are discussed below (see Infantile Spasms).
504 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Primary Generalized Epilepsy Tabl e 2 6 . 4 A

(“Idiopathic Generalized Epilepsy”)
versus Secondarily Generalized Epilepsy Classification of Epilepsy Syndromes
The term primary generalized epilepsy has fallen out of vogue
Epilepsy syndromes
with the ILAE classification commission, but the concept is an
Neonatal
important one for many neurologists and a suitable substitute
Early myoclonic encephalopathy
term has yet to be officially suggested. A secondary generalized
Ohtahara syndrome
epilepsy, or secondary bilateral synchrony, implies that there is
a focal onset with secondary generalization. Infancy
The entire concept of a primary generalized epilepsy has its Epilepsy of infancy with migrating focal seizures
foundation in the EEG. It implies that the clinical and EEG phe- West syndrome
nomena of the seizures are generalized from the start and that Myoclonic epilepsy in infancy
there is no structural epileptogenic focus. The total lack of Benign infantile epilepsy
structural lesions has been challenged by Meencke and Janz Benign familial infantile epilepsy
(324,325), who found areas of minor dysgenetic changes in Dravet syndrome
autopsy studies of patients with presumed primary generalized Myoclonic encephalopathy in nonprogressive disorders
epilepsy. However, Lyon and Gastaut (326) have minimized the Childhood
significance of microdysgenesis, which may occur in other con- Febrile seizures plus
ditions or in neurologically normal controls. Panayiotopoulos syndrome
Currently, it is suggested that the terms focal and general- Epilepsy with myoclonic–atonic seizures
ized should apply to seizures and not to epilepsies (epileptic Benign epilepsy with centrotemporal spikes
syndromes) (327). For example, focal seizures have occurred Autosomal-dominant nocturnal frontal lobe epilepsy
in several of the generalized epilepsy syndromes, such as juve- Late onset childhood occipital epilepsy
nile myoclonic epilepsy and “benign rolandic spikes” may Epilepsy with myoclonic absences
occasionally develop in childhood absence epilepsy, especially Lennox–Gastaut syndrome
after suppression of the absences and the spike waves with Epileptic encephalopathy with continuous spike and wave
medication. Landau–Kleffner syndrome
Lombroso (328) reported his experience with a mean long- Childhood absence epilepsy
term follow-up of 16 years (up to 22 years) in 58 patients with Epilepsy with eyelid myoclonia
various primary generalized epilepsies. Thirty-two of these
Adolescence
patients had focal features present without structural lesion. He
Juvenile absence epilepsy
hypothesized either microdysgenesis or the development of
Juvenile myoclonic epilepsy
localized hyperexcitability in low threshold structures second-
Epilepsy with generalized tonic–clonic seizures alone
ary to long-term generalized spike-wave activity.
Progressive myoclonus epilepsies
Autosomal-dominant epilepsy with auditory features
EPILEPTIC SYNDROMES Familial temporal lobe epilepsies
The syndromatologic approach to the epilepsies is of fairly
recent origin. It has been recognized over the last few decades
that certain forms of epileptic seizure disorders have special
clinical and EEG characteristics regardless of their polyetiologic (332), Helbig et al. (333), Hirose et al. (334), Ottman et al. (335),
background. These forms represent epileptologic entities or Scheffer et al. (336) or are specified. Epigenetic mechanisms may
epileptic syndromes with important differences in course and also be important (337). These references are used for the genetic
prognosis. The correct diagnosis of these syndromes is there- information for the following epilepsy syndromes, unless other-
fore of considerable practical significance. Furthermore, these wise specified.
syndromes are age-dependent and chiefly occur in certain age The frequency of the epilepsy syndromes is remarkably sim-
ranges, especially infancy and childhood but frequently persist- ilar across series. Most epilepsies can be diagnosed early on,
ing through adolescence into adult life. Accordingly, Berg and but some require follow-up for proper identification (338)
others have presented an age-related approach to the listing of (Table 26.4B).
the various epilepsy syndromes (235) (Table 26.4A).
Teaming modern genomic technologies with improved classi- NEONATAL
fication of seizures and epileptic syndromes is the identification of
the underlying genes, receptors and channels that are associated Benign Neonatal Seizures (BNS)
with seizure types and specific syndromes. The genetic findings There are familial (benign familial neonatal seizures [BFNS])
for the various syndromes come from several sources: Chahine and nonfamilial forms. BFNS typically begin on day 2 or 3 of
and Mikati (329), Lagae (330), Lu and Wang (331), Lucarini et al. life, may be difficult to control initially, but typically resolve
 Chapter 26 ■ Seizures and Epilepsy in Infants to Adolescents 505

Tabl e 2 6 . 4 B

Frequency of Epilepsy Syndromes

Kramer Dura-Trave Berg (N = 613) at Berg (N = 613)

Epilepsy Syndrome (N = 440) (N = 365) Initial Diagnosis After 2 Years
Benign rolandic epilepsy 8% 11.5% 9.6% 10.3%
CEOP 2 9 0.3 0.5
BMEI 0.8 0.2 0.3
CAE 12.3 12.1 12.4
JAE 2.2 2.4 2.8
JME 0.9 0.8 2.0 2.4
GTC on awakening 0.5 0.3 0.3
West syndrome 9 4.1 3.0 3.1
LGS 1.5 0.5 0.7 3.1
Myoclonic–astatic 0.2 6 1.6 1.3
Myoclonic absence 1.9 0.3 0.3
SMEI 1.4 0.2 0.2
ESES 0.2 1.1 0.2 0.2
Acquired Ep aphasia (LKS) 0.2 0.5
Ohtahara syndrome 0.2

within the first several months. In the nonfamilial form,
seizures typically begin during the first week of life, commonly
on Day 5 of life. The clinical manifestations consist of tonic
posturing, automatisms, apnea, or focal or generalized clonic
activity. The seizures may rarely persist into adulthood.
Typically the interictal EEG is normal in both forms but
can have focal or multifocal epileptiform discharges, a discon-
tinuous background, or the theta-pointu alternant pattern.
This theta pattern consists of runs of 4 to 7 Hz activity, with
admixed sharp waves, alternating from side to side. It is not
specific for this disorder. In the few recorded seizures in the
familial variety, seizures start with an electrodecrement for 5
to 19 seconds, followed by 1 to 2 minute bilateral spikes or
sharp waves, the prominence could vary from side to side, and
had no postictal EEG depression (339). Focal seizures (340)
and focal seizures with generalization have also been recorded
(341). In the nonfamilial variety, the ictal EEG shows rhyth-
mic spikes or slow waves that may predominate in the central
regions.
The familial form, BFNS, has been associated with muta-
tions in the KCNQ2 (20q13) and KCNQ3 (8q24) voltage-
gated potassium channels, responsible for the M-current. This
slowly activating current regulates subthreshold neuronal
excitability. Retigabine may be helpful since it blocks the
M-current. The benign neonatal-infantile seizures have been
associated with a missense mutation in the SCN2A gene,
which encodes the alpha-2 subunit of the voltage-gated
sodium channel (2q24).
Early Myoclonic Encephalopathy (EME)
(Neonatal Myoclonic Encephalopathy)
The onset of EME is almost always in the newborn period. First
described by Aicardi and Goutires (342), EME is characterized
by fragmentary or erratic partial myoclonus, massive body
myoclonus, focal motor seizures, and tonic spasms. Erratic par-
tial myoclonus (face, limbs, sometimes just a finger, or the orbic-
ular area) may appear in the very first hours of life (343) and
may persist during sleep (344,345). The myoclonus may shift
from one part of the body to another in an anarchic and asyn-
chronous manner (343). Massive myoclonus may not always be
present. Tonic spasms usually appear later, mostly around the
age of 3 months, and are similar to those found in infantile
spasms (West syndrome). Focal (partial) motor seizures, includ-
ing only eye deviation, may occur (344). Autonomic phenomena
such as apnea or flushing of the face may be present (343).
There are very typical EEG changes (Fig. 26.21) with con-
stantly repetitive generalized bursts of high-voltage slow waves
with spikes, lasting a few seconds and separated from each other
by brief stretches of sudden voltage depression or near flatness.
Although the bursts are synchronous, the spikes themselves
show no bilateral synchrony. The bursts of this burst-suppression
pattern are not associated with motor manifestations. The
506 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 26.21 Early infantile myoclonic encephalopathy in a 3-month-
old patient. Note the burst-suppression-like alternation of mixed slow
activity (with some intermingled slow and spiky discharges) and
stretches of flattening.
burst-suppression pattern varies, in that it may only be seen
during sleep, but not in the waking EEG.
There is pronounced impairment of the neurologic and men-
tal status; marked hypotonia with hyperreflexia is common and
microcephaly may develop. Neuroimaging is usually unremark-
able. The prevalence of this syndrome is quite low. Familial
occurrence has been reported and there is some evidence of an
autosomal-recessive inheritance. No single progressive meta-
bolic or neurodegenerative CNS disease has been identified as
causative, although nonketotic hyperglycinemia must be con-
sidered. We have seen this syndrome with the ARX mutation.
The prognosis is very poor; most children die prior to the age
of 2 years.
Early Infantile Epileptic Encephalopathy
(Ohtahara Syndrome)
Ohtahara et al. (346) and Ohtahara (347) described this
syndrome, introducing the term early infantile epileptic
encephalopathy (EIEE). It is characterized by tonic spasms occur-
ring before the age of 20 days and EEG changes similar to those
in EME (342,348). In contrast to EME, the burst-suppression
pattern occurs in both the awake and sleep states and the inter-
burst interval tends to be relatively constant, often about 3 to
5 seconds. The clinical picture and EEG closely resemble those of
nonketotic hyperglycinemia (343). The prognosis is extremely
poor, with a mortality of 50% before the age of 1 month. Its
prevalence is very low. Ohtahara syndrome has been associated
with a de novo mutation in the gene encoding STXBP1 (also
known as MUNC18-1), a protein essential for synaptic vesicle
release and with the ARX mutation (349,350).
Murakami et al. (351) have emphasized the essentially differ-
ent nature of early myoclonic encephalopathy and Ohtahara
syndrome. Transition into West syndrome and Lennox–Gastaut
syndrome is common in the Ohtahara syndrome and rare in
early myoclonic encephalopathy (351).
INFANCY
Epilepsy of Infancy w ith Migrating Focal Seizures
Malignant migrating partial seizures in infancy is a severe
epileptic disorder described in 1995 (352). It is characterized by
nearly continuous electrographic seizures involving multiple
independent areas of onset, beginning in the first 6 months of
life in normal infants with developmental delay and intractable
seizures. No underlying cause has been found, but a functional
or metabolic dysfunction is suspected. Its prevalence is
unknown; series involving only small numbers of patients have
been published, but the spectrum of clinical manifestations
appears to be large. Most seizures are focal and accompanied by
autonomic manifestations or subtle version; many have motor
manifestations that migrate from one side of the body to the
other. Two of the original 14 patients described by Coppola and
colleagues presented with status epilepticus. Sometimes very
mild clinical features were observed during habitual seizures.
The interictal EEG was normal in three patients at onset, but
was always abnormal during evolution, with multifocal spikes
and sleep abnormalities. The ictal EEG is characterized by
monomorphic rhythmic theta or alpha activity progressively
involving multiple sites, moving from one area to another (thus
the term migrating), and progressively spreading to involve
larger area. Additional seizures starting in other areas in either
hemisphere could start before the end of the first event, or
immediately follow it.
Neuroimaging has not revealed any consistent abnormality
but most patients develop microcephaly during the first year of
life. During short-term follow-up, 3/14 patients described by
Coppola and colleagues died. Conventional treatments were
ineffective, except for the combination of stiripentol and clon-
azepam in two children, in which some neurologic improvement
was also noted. We (personal observations) and others (353) sus-
pect that there is a wider clinical spectrum of this disorder. In
particular we have observed patients with less-than-constant
seizures and Marsh and colleagues report patients with slightly
better outcomes than those reported by Coppola et al.
INFANTILE SPASMS (WEST SYNDROME)
Infantile spasms (IS) are an age-dependent “malignant” epilep-
tic syndrome characterized by the triad: myoclonic or tonic
seizures, hypsarrhythmia, and mental retardation, referred to as
West syndrome. The peak onset is between 4 and 7 months and
always before 12 months of age. Hypsarrhythmia (37) is an EEG
term that denotes the EEG correlate of the condition. Infantile
spasms are considered an epileptic encephalopathy, defined as a
disorder in which the epileptiform activity, in this specific case,
the hypsarrhythmia, causes neurologic dysfunction. This is
referred to as a catastrophic epilepsy, with a very poor prognosis
 Chapter 26
unless the epilepsy is controlled. The hypsarrhythmia pattern
may be a variant form of nonconvulsive status epilepticus (354)
and it is possible that the lack of an organized background and
normal brain rhythms critically interferes with the ability of the
brain to process information in an orderly manner.
The current classification system (2001 glossary, 235) now
uses the term “epileptic spasms” because spasms may continue
or develop after infancy (355). We continue using the term
infantile spasms here.
Infantile spasms are divided into symptomatic, cryptogenic,
and idiopathic forms. Symptomatic infantile spasms are sec-
ondary to a known neurologic disorder, such as tuberous scle-
rosis; cryptogenic refers to a suspected, but not definitely
identified neurologic disorder; idiopathic refers to a case in
which no specific disorder has been identified. The prognosis is
related to the underlying etiology and is worse when the spasms
are either symptomatic or cryptogenic. The spasms themselves
consist of sudden tonic and myoclonic phenomena, including
brief head nods, with quick extension and flexion movements
of the trunk, arms, and legs, occurring in clusters and especially
during transitions from sleeping to awaking. These are also
described as flexor, extensor, or mixed. The actual spasms may
resolve, but other seizure types occur later.
The term hypsarrhythmia (37) is derived from the Greek word
hypselos, which means “high,” thus indicating the high voltage
that generally predominates. No hypsarrhythmic recording can
be appropriately obtained at the standard sensitivity of the EEG
apparatus; lowering the sensitivity is required in order to prevent
“blocking.” Bursts of very high voltage slow waves occur in an
irregular fashion with a varying degree of bilateral synchrony,
which usually increases in sleep. The stages of early non-REM
sleep are particularly conducive to a typical hypsarrhythmic
recording. Long stretches of high-voltage slow and admixed spike
activity may suddenly be interrupted by a generalized elec-
trodecrement, or less commonly in only a few EEG channels or
over one hemisphere; these electrodecrements are practically
limited to sleep tracings (Fig. 26.22). The spike activity shows
Figure 26.22 Markedly depressed stretches in an 8-month-old patient
with infantile spasms-hypsarrhythmia. Record was obtained in sleep.
■ Seizures and Epilepsy in Infants to Adolescents 507
Figure 26.23 Infantile spasms-hypsarrhythmia in an 8-month-old
patient. Note the high-voltage output (see parameter of sensitivity) and
posterior maximum of spikes. Sleep record.
single spikes and sharp waves, as well as very brief sequences of
polyspikes that are usually of smaller amplitude and the spike
activity almost always has a posterior predominance (Fig. 26.23).
The ictal EEG, occurring at the time of the actual infantile
spasms, is quite variable. The most common ictal finding is an
electrodecremental response, a sudden suppression of the EEG
activity that lasts for several seconds (118,356). However, fast
activity and high-voltage spikes (37) or polyspikes and slow
waves may be present (118). Fusco and Vigevano (357) reported
that the characteristic ictal pattern is a high-amplitude vertex
positive wave. A sleep recording is a necessity because the hyp-
sarrhythmia may be confined only to the sleep portion in some
children (356).
The great variability of EEG abnormalities in long-term
recordings has been described by Hrachovy et al. (358):
increased interhemispheric synchronization, asymmetrical
hypsarrhythmia, hypsarrhythmia with a consistent focus of
abnormal discharge, hypsarrhythmia with episodes of attenua-
tion, and hypsarrhythmia comprising primarily high-voltage
slow activity with little sharp wave or spike activity. Kramer et
al. (359) studied these hypsarrhythmia variant patterns and
devised a scoring system. The variant patterns occurred in 69%
of 53 consecutive EEGs. Prognosis was not related to the pres-
ence of any one variant pattern, but was associated with faster
background activity ( 75% delta), a lower total hypsarrhyth-
mia score ( or = 10), and with absence of electrodecremental
discharges on the pre–ACTH EEG.
The evaluation for infantile spasms includes neuroimaging
to exclude anatomic lesions, but other disorders, especially
metabolic and infectious, and now, specific genetic disorders,
need to be excluded. X-linked infantile spasms have been asso-
ciated with a defect in the ARX gene (360) and CDKL5 (cyclin-
dependent kinase like 5) gene. Secondary focal spasms may
occur from focal lesions, such as tumor, stroke, or a focal corti-
cal dysplasia. ACTH has been considered the drug of choice by
many, although benzodiazepines, especially nitrazepam, val-
proic acid, and vigabatrin have been used. A UK study showed
508 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
that ACTH and oral prednisone have a better efficacy than viga-
batrin (361). ACTH and vigabatrin are the only AEDs for which
evidence exists for efficacy and for controlling the hypsarrhyth-
mia (362,363). Vigabatrin may be especially effective when
spasms are caused by tuberous sclerosis (364). However, vigaba-
trin has been associated with retinal dysfunction and visual
field defects and has only just been released in the United States
20 years after its introduction.
The EEG shows an almost immediate improvement under
effective therapy, but a clinical improvement may not necessar-
ily follow. Complete normalization may occur, but this may be
only temporary, and epileptiform activity, mostly over posterior
regions, and other seizure types may develop. In many cases with
a poor therapeutic response and especially in those with preex-
isting brain damage, a transition to the Lennox–Gastaut syn-
drome is common. This occurs in 58.7% according to Ohtahara
et al. (365) and in 50% to 70% according to Hughes (366). In
addition, infantile spasms may persist as epileptic spasms (367),
or infantile spasms in remission may recur and persist as epilep-
tic spasms (368).
If a focal lesion, especially a tumor, is seen on neuroimaging,
then surgical resection may be curative. Even in the absence of
a lesion on neuroimaging, surgery may be done on refractory
cases if an epileptogenic focus is identified, especially using the
PET scan (369). Most of these cases result from focal cortical
dysplasias.
A fine clinical–electrical analysis of infantile spasms was car-
ried out by Fusco and Vigevano (357). In both cryptogenic and
symptomatic cases, the EEG pattern that corresponds to the
spasm consists of a slow wave at the vertex. Two other patterns
include fast activity, called “spindle-like” and the electrodecre-
ment. The vertex positive slow wave was associated with the
actual spasm, the fast activity with clinical staring, and the elec-
trodecrement was postictal. The persistence of the hypsarrhyth-
mic pattern during a cluster of spasms is found in both
symptomatic and the idiopathic and cryptogenic groups and
asymmetric spasms suggest a symptomatic etiology. Persistence
of EEG hypsarrhythmia during a cluster of spasms appears to
indicate a favorable prognosis in the idiopathic group (357).
According to Koo and Hwang (370), occipital lesions are asso-
ciated with the earliest onset of spasms and frontal lesions with
late onset. Our clinical experience has been different and we
have not been able to confirm these observations. The reader is
cautioned to avoid placing too much emphasis on the prognos-
tic value of ictal hypsarrhythmia or to infer much about the
location of the lesion based upon the age of presentation.
Hypsarrhythmia is not always a reliable EEG correlate of
infantile spasms. There are clinically convincing cases with no
hypsarrhythmia, but in these rare exceptions the voltage output
is unusually high and there are almost invariably interictal
epileptiform discharges. A completely normal EEG obtained
awake and asleep virtually excludes the diagnosis of infantile
spasms. It should be stressed here that it does not exclude later
forms of spasms like periodic spasms. In infantile spasms unless
there is a rapid response to treatment, hypsarrhythmia is likely
to appear in the further course of such infants, commonly for
Tabl e 2 6 . 5
Infantile Spasms: Differential Diagnosis
Neonatal sleep myoclonus
Benign myoclonus of infancy
Head banging (jactatio capitis nocturnis)
Spasmus nutans
Sandifer syndrome, abdominal pain (colic), esophageal
spasm
Shuddering attacks
Hyperekplexia
Paroxysmal infantile dystonia, choreoathetosis, torticollis
Benign paroxysmal tonic upward gaze
Infantile self-gratification
Breath-holding episodes with myoclonus/myotonic
posturing
Intoxications
Increased moro reflex with opisthotonus
Myoclonic encephalopathy
From Fejerman N. Differential diagnosis. In: Dulac O, Chugani HT, Dalla
Bernardina B, eds. Infantile Spasms and West Syndrome. London: WB Saunders;
1994:88–98.
weeks. Hypsarrhythmia may evolve later into the multiple inde-
pendent spike syndrome.
On the other hand, if the expected hypsarrhythmia is not
found in a suspected case of infantile spasms, the presumptive
diagnosis may be wrong. The clinical differential diagnosis of
infantile spasms or hypsarrhythmia has been beautifully
demonstrated by Roger et al. (371) and includes a variety of con-
ditions. In spasmus nutans, the EEG is normal. Jactatio capitis
nocturna also has a normal EEG. With salaam tic or “salutatory”
spasms (Moro), there are nonspecific EEG abnormalities, some-
times with spikes in combination with epileptic seizures, but no
hypsarrhythmia. In myoclonic encephalopathy (372), the EEG is
normal (Table 26.5).
Benign Myoclonic Epilepsy in Infancy
Myoclonic seizures are the predominant manifestation in this
epilepsy syndrome. Infants with normal behavioral and cognition
develop myoclonic jerks mostly of the head and proximal arms,
occurring in isolation or in brief runs of recurrent jerks. There is
often a vocalization associated with an expiratory noise (Lin,
1998), considered characteristic of this disorder. Rarely, general-
ized clonic seizures may occur during sleep or upon awakening
that may be prolonged and associated with cyanosis (Vegevano,
1990). The ictal EEG shows a burst of generalized or diffuse spike-
wave activity with the myoclonia, otherwise the interictal EEG is
normal (Auvin, 2005). However, focal discharges, especially in the
rolandic and vertex regions may mimic a focal epilepsy (373).
 Chapter 26
Benign Infantile Epilepsy and Benign
Familial Infantile Epilepsy
Benign infantile epilepsy may be seen in sporadic or familial
forms (374,375; Watanabe, 1993; Fukuyama, 1963). In these
disorders, infants manifest focal seizures that may present with
subtle behavioral arrest or, on the other extreme, bilateral con-
vulsive features. The interictal EEG is usually normal, though
some benign infantile epilepsies may have peculiar low or
medium voltage vertex spikes or sharp waves with a dome-like
morphology (376). Ictal discharges often arise from the tempo-
ral region or posterior quadrant, but this is relatively nonspe-
cific for focal seizures in the immature.
Severe Myoclonic Epilepsy in Infants
(Dravet Syndrome)
First reported by Dravet (377) and subsequently confirmed
by Dalla Bernardina et al. (378), Severe Myoclonic Epilepsy of
Infancy (SMEI) is an uncommon syndrome. There is typically a
family history of epilepsy or febrile seizures (FS), and seizures
begin in the first year of life (typically around 6 months of age)
in a previously normal infant who develops prolonged general-
ized or unilateral febrile clonic seizures, with then frequently
occurring convulsive seizures. In addition, there may be attacks of
pallor, cyanosis, atonic phenomena, and automatisms (379).
Moderate fever may accompany the first attacks. Multiple seizure
types, especially myoclonic seizures then develop between 1 and
4 years of age, and focal seizures may also occur.
The interictal EEG may initially be normal although we have
seen low amplitude occipital polyspikes in some infants in the
first year of life. The theta-pointu pattern may be seen. More
prominent paroxysmal abnormalities tend to develop during
the second year of life with generalized spikes, polyspikes, and
spike-wave-like activity, which may be sleep-activated and pro-
gressive background slowing develops. Photosensitivity is com-
mon and develops in the first phase of the disorder, and focal
abnormalities may be present. Dravet syndrome has been asso-
ciated with mutations in the SCN1A gene (19q13.1) and in the
GABA RA2 (5q34) receptor. These occur in about 70% to 80%
of those with Dravet syndrome (336).
Psychomotor retardation occurs, usually beginning in the
second year of life, with ataxia, interictal myoclonus, and hyper-
reflexia. The seizures are typically treatment resistant.
Myoclonic Encephalopathy in
Nonprogressive Disorders
Myoclonic encephalopathy in nonprogressive disorders is char-
acterized by recurrent myoclonic status in infants and young
children with nonprogressive encephalopathy. In a study
reviewing the findings in 29 patients, 3 main subgroups were
described (380). The first subgroup of 18 patients presented
with myoclonic absences and rhythmic myoclonias. These were
followed by a brief silent period during which there was sub-
continuous delta–theta activity in the central areas, and rhyth-
mic delta waves with superimposed spikes mainly involving the
parietooccipital regions. The latter were often activated by eye
■ Seizures and Epilepsy in Infants to Adolescents 509
closure. All of these children had a genetic cause for their
epilepsy. The second subgroup included five patients with
inhibitory phenomena associated with a dystonic component
and sudden irregular rapid lightning-like jerks. The EEG
showed subcontinuous multifocal slow spike waves, predomi-
nating in frontocentral regions. These patients either had corti-
cal malformations or the etiology was unknown. The third
subgroup included six children who initially suffered from
myoclonic absences. The status was characterized by subcontin-
uous generalized spike-wave-type paroxysms correlating with
the rhythmic myoclonia of face and limbs. After 1 to 3 weeks,
the EEG showed sharp theta waves with very slow pseudorhyth-
mic continuous spikes in the central regions and vertex. The
cause in this subgroup was perinatal anoxic injury.
CHILDHOOD
Febrile Convulsions
Age and Definition
The ICES considers FS a situation-related seizure: the situation
being the illness with fever. A benign febrile seizure is defined as
a short ( 15 minutes) generalized tonic–clonic seizure without
significant postictal depression, in a child with a nonfocal neu-
rologic examination without a family history of epilepsy. A
febrile seizure is otherwise considered complex if it is longer or
predominantly focal. This condition is probably the most com-
mon epileptic seizure disorder; about 3% to 4% of all children
have at least one febrile seizure in infancy or early childhood
(2.2% to 4% according to Knudsen (381)). The seizures tend to
occur between the ages of 6 months and 5 years, especially
between 6 months and 3 years. The onset falls into the range of
6 to 24 months. Beaumanoir (382) thinks that it is unwise to
call fever-induced convulsions after the age of 4 years “simple
febrile convulsions.”
The conventional wisdom regarding the EEG in benign FS
is that the interictal EEG is normal, whereas in a child with
epilepsy, the interictal EEG may be abnormal, with associated
epileptiform features. The recent practice parameter on benign
FS does not require an EEG in the evaluation of a benign febrile
seizure (383). However, a careful literature review shows that
there may not be such a clear EEG distinction for a “benign
febrile seizure.” Alvarez (96) reported hypnagogic paroxysmal
spike-wave activity (minimal epileptiform features, sharp waves
embedded into hyperventilation or hypnagogic hypersyn-
chrony) occurring with a higher incidence in children with FS.
This is similar to the epileptiform activity admixed into vertex
waves and sleep spindles, called dyshormia, which refers to
abnormal epileptiform arousal features during sleep. These fea-
tures may indicate a lower seizure threshold and explains why
they occur in both the normal population and in those with
epilepsy. Japanese child neurologists rely more on EEG and have
introduced the concept of epileptic versus nonepileptic febrile
convulsions (384). However, children with epileptic febrile con-
vulsions treated with AEDs had the same recurrence rate as
those not treated with AEDs.
510 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

In one classic reference, Frantzen et al. concluded that EEG Findings

epileptiform discharges were not predictive of later develop-
ment of a FS (Frantzen, 1968). They did not, however,
separate out the various types of interictal epileptiform
discharges in their analysis. Frantzen and colleagues did
note that focal slowing was seen in the posterior head
region in children with prolonged FS. They noted that foci
of spikes could develop in these children approximately
1.5 years later, but their study did not have sufficient follow-
up to determine if spikes were predictive of later develop-
ment of epilepsy.
Genetics of Febrile Seizures
It has been thought that FS are inherited in an autosomal-dom-
inant manner with incomplete penetrance. In families with FS,
six genetic loci have been mapped: FEB 1 (8q13-q21), FEB 2
(19p13.3), FEB 3 (2q23-q24), FEB 4 (5q14-q15), FEB 5 (6q22-
q24), and FEB 6 (18p11.2) (385). Defects in the interleukin
genes interleukin 1 beta and interleukin 1 receptor antagonist
have also been associated with FS (386) (Table 26.6).
Ictal EEG tracings are hard to obtain in a truly simple febrile
convulsion. Although generalized tonic–clonic seizure EEG
changes might be inferred from the apparent widespread and
symmetric body movements described by many parents, this
sequence is rarely seen in infants (Korrf, 2005). Indeed, EEG
recordings obtained in inpatients with acute FS show lateralized
EEG changes (see samples by Lennox-Buchthal (387)), but an
argument could be made that these are most likely the more dra-
matic and severe cases.
In a pure benign febrile seizure, the interictal recording should
be normal. Gibbs and Gibbs (83) pointed out that short spike-
wave-like bursts in drowsiness and sleep may occur (“pseudo petit
mal discharge”). Des Termes et al. (388) found 31 children with
spike foci among 500 patients with febrile convulsions. The most
common site was the occipital lobe and, in 88%, the spike focus
disappeared within 3 years. Yamatogi et al. (389) have stressed the
excellent prognosis of children with febrile convulsions and per-
fectly normal EEG tracings. Yamatogi and Ohtahara (390) have
placed emphasis on the clinical value of follow-up EEG studies.

Tabl e 2 6 . 6

Differences between Febrile Convulsions and Seizures due to Febrile Brain Disease

Febrile Convulsions Seizures in Febrile Brain Disease

Typical age range
Genetic predisposition
to seizures
Type of seizure
Duration of seizure
Clinical setting in
which seizures occur
Type of underlying
cerebral pathology
Postictal neurologic
deficit (Todd’s paralysis)
EEG
Anticonvulsive
medication
Prognosis
6 months to 3 years (seldom 5 years)
May be strong
Tonic–clonic (modified or
attenuated grand mal)
Mostly 1–3 min seldom prolonged
At the onset of a febrile disease,
mostly upper respiratory illness,
often coinciding with the first sharp
rise in temperature
None
Very uncommon
Rapidly normalizes after convulsion
normal interval tracings in 80–90%
Not necessary (neither for acute
convulsions nor for prevention of
further seizures)
Excellent in the vast majority
(especially those with normal
interval EEG)
Mainly 0–3 years
Mostly minor or insignificant
Tonic–clonic (grand mal-like) or hemiconvulsions
Often prolonged, 10 min to hours, status-like or in
rapid succession
In a variety of CNS infections (encephalitis,
meningoencephalitis), intracranial venous
thrombosis, cerebrovascular accidents of infancy;
also in exanthema subitum and after smallpox
vaccination, but usually less severe
Various types of inflammatory and vascular
changes, in milder cases limited to edema
Common and often mixed with pathology-
determined neurologic defect
Abnormal throughout febrile episode, abnormal
in interval (except for mild encephalitis)
Acutely needed (preferably benzodiazepines,
phenobarbital), long-term treatment required
afterward (except for mild encephalitis)
Guarded; neurologic defects and further seizures
common
 Chapter 26
Abnormal interictal EEG tracings are likely to indicate
underlying cerebral impairment with paroxysmal properties;
these infants might be candidates for a febrile epileptic manifes-
tation (i.e., for a chronic epileptic seizure disorder) in the
future. According to Des Termes et al. (388), the prognosis for
these children is mostly favorable.
Generalized Epilepsy w ith Febrile
Seizures Plus (GEFS )
Generalized epilepsy with FS plus (GEFS ) was described in
1997 (Scheffer, 1997). It is characterized by the association of
generalized FS beyond the age of 6 years and afebrile generalized
convulsions; a positive family history of epilepsy with variable
phenotypes; and a favorable evolution in the majority of cases.
The first description in a large family revealed that the clinical
spectrum of GEFS comprised cases of simple FS; FS beyond
the age of 6 years (FS ); FS and absences; FS and myoclonic
seizures; FS and atonic seizures; and myoclonic–astatic
epilepsy. This spectrum was further extended with other types of
epilepsy, such as FS and temporal lobe epilepsy (Singh, 1999),
Dravet syndrome, and cryptogenic Lennox–Gastaut syndrome
(Singh, 2001). The pattern of transmission in the first family
suggested an autosomal-dominant inheritance. The authors
hypothesized that a single major gene defect, with possible
additive gene effect, was responsible for the phenotypic vari-
ability on the entire spectrum of the generalized epilepsies
(Scheffer, Singh, 1999). Extensive genetic analyses in these
families identified a mutation in the sodium-channel 1
subunit gene (SCN 1) on chromosome region 19q13.1
(Wallace). Additional mutations were later discovered in other
genes, such as the sodium-channel 1 subunit gene on chro-
mosome region 2q24 (Escayg), or the GABAA receptor 2-
subunit gene on chromosome region 5q34 (Baulac). These
findings underline the genetic complexity of this disorder,
which remains to be completely understood. Some have
argued that this is not a specific epilepsy syndrome, but rather
a description of a genetic susceptibility. In either case, this
research has been one of the most exciting and productive
areas in pediatric epilepsy.
The Panayiotopoulos Syndrome
In an excellent overview of benign childhood partial seizures,
Panayiotopoulos (391) described a special type of early-
onset occipital seizure for which he coined the term
Panayiotopoulos syndrome. This syndrome was subsequently
highlighted in a special study (392) with the following initial
characterization of this syndrome: “Panayiotopoulos syndrome
is a childhood-related idiopathic benign susceptibility to
partial, mainly autonomic, seizures that may be genetically
determined. The children have normal physical and neuropsy-
chological development.”
The age of onset shows a range from 1 to 14 years with a
peak at 4 to 5; 76% start between ages 3 and 6 years. The preva-
lence is about 13% among children with a seizure onset
between ages 3 and 6 years and overall, about 0.2% to 0.3% of
the general population of children may be affected. This figure
■ Seizures and Epilepsy in Infants to Adolescents 511
may be much higher if children with atypical and inconspicu-
ous presentation are included.
The aforementioned “mainly autonomic” seizures consist of
feeling sick, looking pale, nausea, retching, and vomiting. There
may also be cyanosis, mydriasis (or miosis), cardiorespiratory
and thermoregulatory alterations, incontinence of urine and/or
feces. Apnea and cardiac asystole may be exceptionally severe.
Two thirds of the seizures occur in sleep. Unilateral eye devia-
tion is common. The seizures commonly end with hemiconvul-
sions or with Jacksonian march or rarely with a generalized
tonic–clonic seizure, which, quite rarely, evolve into status
epilepticus. The duration varies from 1 to 30 minutes, and
autonomic status epilepticus may occur. The final portion of
the seizure may consist of ictal syncope with flaccidity. After
sufficient sleep, the child is perfectly normal.
In a prospective study of 192 children with PS by Caraballo
et al. (393), the following inclusion criteria were used: (i) auto-
nomic manifestations, visual symptoms, and/or simple motor
focal seizures followed or not by impairment of consciousness
with or without secondary generalization; (ii) functional occip-
ital and extra-occipital spikes alone or combined on the inter-
ictal EEG, or a normal EEG; (iii) normal neurologic and mental
status; and (iv) normal brain imaging. Of the 172 treated chil-
dren, 152 became seizure free with treatment, 6% had infre-
quent seizures, and 6% had frequent seizures (380).
The EEG is the most useful test, once neuroimaging is nor-
mal. Ninety percent of EEGs show multifocal spikes and spike
waves with posterior accentuation. The spikes were occipital in
75%, temporal in 26%, and frontal in 10%; with the occipital
spikes, 60% were bilateral and synchronous, with voltage asym-
metries, in the other 40%, spikes were unilateral and greater on
the right side (70%). The spikes were sleep-activated in 90% and
three had generalized spikes. The ictal EEG demonstrates rhyth-
mic delta activity intermixed with usually small spikes. Onset is
unilateral, often posterior, but may also be anterior and not
strictly localized to one electrode. In four EEGs with seizures
recorded from the onset, the onset was occiptotemporal in two,
and one each from the frontal or frontotemporal regions (394).
The EEG in PS may show shifting or multiple epileptic foci,
some with an occipital predominance (Ferrie, 2006). In follow-
up EEGs, 17% had a shift of epileptiform activity from the
occipital to centrotemporal areas and 3% to frontal areas (394).
In PS, frontal spikes may be activated by occipital foci (395) and
the “synchronous occipital and frontopolar spike phenomenon”
(an occipital and frontopolar spike) has been reported (396), in
which negative occipital spikes precede the frontal spikes, sug-
gesting propagation; this did not occur in all children.
The family history is usually negative and there is no relation-
ship to migraine in the family. The alternative diagnoses made in
these children include encephalitis, syncope, migraine, sleep dis-
order, and gastroenteritis, with SE occurring in a large percent-
age of cases. A recent consensus statement defined autonomic SE
as epileptic activity manifesting as autonomic dysfunction at
seizure onset lasting greater than 30 minutes and emphasized
that it also occurs in childhood symptomatic epilepsies, rarely in
adults, and is not limited to the Panayiotopoulos syndrome.
512 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Myoclonic–Astatic Epilepsy of Childhood
This syndrome has been described by Doose (397) (“akinetic
petit mal”). This condition is often confused with Lennox–
Gastaut syndrome, but is clearly different. Doose (398) feels that
this epileptic condition is taxonomically closer to primary gen-
eralized epilepsy rather than to the Lennox–Gastaut syndrome.
Deonna et al. (1986) reported a combination of myoclonic–
astatic epilepsy and “benign focal epilepsy of childhood.”
There are now indications that the Lennox–Gastaut syndrome
and the syndrome of myoclonic–astatic epilepsy are indeed two
separate entities with different neurophysiologic mechanisms
generating the myoclonus. In the Lennox–Gastaut syndrome,
myoclonus originates from a stable generator in the frontal cor-
tex. Spreading to contra- and ipsilateral cortical areas, whereas in
myoclonic–astatic epilepsy, myoclonus appears to be a primary
generalized epileptic phenomenon (399).
Onset is typically between 7 months and 6 years of age. The
seizures consist primarily of generalized myoclonic, astatic, or
myoclonic–astatic seizures, short absences, and mostly GTC
seizures typically without tonic seizure or drop attacks.
However, tonic seizures may occur in cases with a poor outcome.
At the onset, the EEG background may be normal and char-
acterized by a 4 to 7 Hz rhythm with a parietal midline accen-
tuation. There are generalized 2 to 3 Hz spike and wave
discharges with admixed slow waves. There is a high incidence
of seizures or EEG abnormalities in relatives. Prior to the onset
of the seizures, which may start as FS, the development is nor-
mal. Valproic acid and ethosuximide have been considered the
drugs of choice, and the ketogenic diet has also been effective
(400). The prognosis is variable, but in our experience many
can do well with prompt initiation of effective treatment.
The myoclonic–astatic seizure is now referred to as a
myoclonic–atonic seizure (235).
Benign Rolandic Epilepsy
In children with spikes over the central region and/or adjacent
midtemporal and parietal areas, a benign and readily control-
lable type of epileptic seizure disorder with focal motor seizures
and/or generalized tonic–clonic seizures are the rule. This is a
special form of childhood epilepsy. Earlier observations of
Gastaut (401), Hess (402), Nayrac and Beaussart (403), Isler
and Hess (404), and Gibbs and Gibbs (83) have set the stage for
more recent work in this field (46,405–415). This work has out-
lined the characteristics of benign rolandic epilepsy, as this con-
dition has been termed over the past few years.
Beaumanoir (382) has listed this form of childhood epilepsy
among the primary generalized epilepsies despite its prominent
focal features in the ictal and EEG semiology. Such a classifica-
tion certainly appears to be provocative. There are indeed cer-
tain relationships between benign rolandic epilepsy and
primary generalized epilepsies, and conversion from one form
to the other may occur (416–419). There is certainly good rea-
son to separate benign rolandic epilepsies from the bulk of focal
(partial) epileptic seizure disorders, which will be presented
somewhat later. In this section, we also discuss benign seizure
disorders in children with spikes outside the rolandic region.
Age and Prevalence
Benign rolandic epilepsy occurs at age 3 to 12 years, with the
majority 6 to 10 years old. Disappearance of the seizures during
adolescence or even prior to puberty is the rule. The seizures may
occasionally recur much later in life, probably due to seizure-
facilitating factors such as severe illness or toxic-metabolic factors.
The gender distribution shows that boys are more often
affected. The prevalence is not quite clear and might be some-
where between 5% and 10% in a population of epileptics below
age 15. Beaumanoir (382) goes even further; she thinks that the
prevalence of benign rolandic epilepsy exceeds that of petit mal
absences and reaches 14% to 15% of all childhood epilepsy
cases (Fig. 26.24).
Ictal Manifestations
The actual seizures are hardly ever seen by the physician, even by
the epileptologist who sees sizable numbers of these children
(413,420,421). One therefore depends heavily on descriptions by
the patient or the patient’s family; nocturnal videotape or long-
term monitoring recordings from the Epilepsy Monitoring Unit
are quite helpful. Whether the seizures are focal or generalized,
they tend to occur during nocturnal sleep (413), mostly during
the last hour of sleep or in the first 2 hours. About 80% of the
attacks occur in sleep; of the remaining 20%, about 10% take
place shortly after awakening. Nocturnal seizures may awaken
the child afterward. Preservation of consciousness and hence the
ability to describe the experienced seizure, indicating focality,
was found in 58% and generalized tonic–clonic seizures were
noted in 26% (413).
Focal seizures often involve the face. The midtemporal spike
localization has been thought to be related to paroxysmal activ-
ity in the very closely located lower portion of the motor strip
(faciolaryngopharyngeal muscles); this has been stressed by
Lombroso (412), who uses the term sylvian seizures. Hemifacial
twitching is definitely more common than clonic motions in
the contralateral arm; least common is clonic activity in the leg.
In some cases, the entire half of the body participates, but a typ-
ical Jacksonian march does not seem to occur. Speech arrest is
quite common (39% of the seizures, 413). This is apparently an
ictal anarthria with preserved internal speech.
Oropharyngeal involvement is very often reported (53%,
413), with sounds described as “guttural,” “gargling,” “throaty,”
“wheezing,” or “as if going to vomit.” The feeling of suffocation
is reported as coming from the mouth but not from the chest.
These patients also have focal seizures that are not rolandic,
with blindness, vertigo, and torsion of the body as ictal signs.
This underscores the complexity of the underlying neurophys-
iologic mechanisms.
EEG
Spiking over the central–midtemporal area in children is of lim-
ited epileptogenicity. It is reasonable to presume that 50% to
70% of these children have seizures and the remaining 30% to
50% are seizure free. These latter patients are referred to the EEG
laboratory because of a variety of symptoms such as behavior
disorder, headaches, and other complaints or deviations.
 Chapter 26
The spatial distribution of the spike activity requires an
appropriate number of electrodes; the International 10-20
Electrode System is particularly suitable. Otherwise, the rolandic
cortex may lie between a frontal and a parietal electrode, and
strictly local spiking may escape detection, especially when the
midtemporal region is not explored ideally. In our experience, a
central maximum of spike discharge is slightly more often noted
than a midtemporal maximum. As above, the EEG has distinc-
tive, high amplitude, diphasic spikes or sharp waves with a
prominent slow wave, in the midtemporal (T3,T4) and central
(C3,C4) regions. However, when additional scalp electrodes are
placed, maximum negativity was in the central region, either in
a “high” (C3,C4) or “low” (C5,C6) location (Legarda, 1994).
The spikes themselves are large and may be either spikes in
the strict sense or sharp waves (Fig. 26.25).
Central spiking with dipole formation has been reported by
Gregory and Wong (19,422). In fact, Gregory and Wong (422)
have stressed the clinical relevance of presence or absence of
dipole fields in children with epileptic seizure disorder and
rolandic spikes. According to their data, the presence of dipole
discharges (usually located over frontal or temporal regions) is
associated with a benign functional rolandic focus with little or
no clinical problems, whereas absence of a dipole discharge is
more likely to be found in children with neurologic and behav-
ioral abnormalities. In our opinion, however, the differential
diagnosis of benign rolandic epilepsy should not hinge on cen-
tral spikes of dipole or nondipole type.
In the course of the 1990s, presence and orientation of
dipoles of rolandic spikes have become a widely debated issue.
According to Wong (423), the dipole-dependent topography of
rolandic spikes “contains information on the behavior or func-
tional state of brain tissue even distant to the brain region
responsible for the spike generation.” Van der Meij et al. (424)
feel that sequential mapping can differentiate between epileptic
and nonepileptic rolandic spikes. Classically, rolandic spikes
■ Seizures and Epilepsy in Infants to Adolescents 513
Figure 26.24 Presumed physiopathogenetic mechanisms
involved in primary generalized epilepsy. The dyshormic
and the photosensitive subgroups show considerable over-
lap. Note the greater significance of the age factor in the
dyshormic group; also note differences of spike-wave fre-
quencies and types of seizures. CGE, common generalized
epilepsy.
considered pathognomonic for benign rolandic epilepsy show
dipole configurations with the negative pole over the central
(centrotemporal) region and a positive pole over the frontal
region (303,425). Typical spikes indicating benign rolandic
seizure disorder start with a spike of centroparietal negativity
and superior frontal negativity, changing within 12 msec to sole
central negativity and after further 16 msec to a dipole of central
negativity and frontomedian positivity (424). Similar findings
were observed with visual analysis during ictal manifestations of
benign rolandic epilepsy (426). According to Legarda and
Jayakar (427), the dipolar distribution of rolandic spikes was not
found to be a reliable indicator of potential epileptogenicity.
Yoshinaga et al. (428) have given strong support to the
importance of dipole tracing in benign rolandic and other types
of childhood epilepsies. Confusing elements in the assessment
of rolandic spikes and their dipoles may arise from the assump-
tion of a single dipole—a concept presently widely supplanted
by the concept of multiple sources and extended source geom-
etry (303,422). It has to be added that every investigator of the
cerebral dipole problem has to take into account the stringent
limitations of dipole theory (429).
Spiking is usually enhanced in light non-REM sleep, during
which the discharges may become extremely abundant. Their
random character may give way to quasirhythmic or periodic
spiking at intervals of less than 1 second; previously unilateral
spikes become bilateral synchronous or asynchronous (382). In
many children, rolandic spikes are found in the sleep portion
only. According to Beaumanoir (382), REM sleep restores the
unilateral character of the spiking. Beaumanoir (382) also
described the occurrence of bilateral parietooccipital 4 Hz
spike-wave-like discharges of moderate voltage seen in the wak-
ing patient (Fig. 26.26).
In some patients, interictal rolandic spikes can be blocked
with contralateral finger or hand movements (Niedermeyer,
1970), either actively or passively (in sleep) performed. The
514 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Figure 26.25 A: Left, a few suggestions of right cen-

tral spikes in a 9-year-old boy. Center, some activity
over right central area during flicker (same patient).
Right, marked spikes and sharp waves over central
regions, especially on the left, during non-REM sleep
(same patient). (From Niedermeyer E, Koshino Y. My-
rhythmus: vorkommen und klinische bedeutung. EEG-
EMG. 1975;6:69–78.) B: Sleep tracing in a 7-year-old
boy with grand mal attacks since age 3. Postictally, the
patient was temporarily aphasic and hemiparetic on
the right side. Note the left midtemporal spikes and a
burst of generalized spikes mixed with slow activity.

Figure 26.26 A: A 5-year-old patient with recent onset of petit mal absences. The recorded generalized spike-wave burst
was subclinical. B: The same patient at age 8 years. There are numerous independent spikes over left central and right cen-
troparietal regions. This case is a typical example of crossover between primary generalized epilepsy and benign rolandic
epilepsy.
 Chapter 26 ■ Seizures and Epilepsy in Infants to Adolescents 515

Figure 26.27 A 14-year-old patient with severe personality disorder

but no history of epileptic seizures. The lower portion was recorded
with identical montage, sensitivity, and paper speed. Note abundance
of right central spikes and excellent blocking response to contralateral
movement (clenching left fist).

reactivity of rolandic spikes was studied by Fonseca et al. (430),

who noted a strong spike blocking response tongue movements
inside or outside the mouth. On the other hand, Nadkarni et al.
(431) reported a child with centro-midtemporal spikes that
could be triggered by blinking.
The rest of the tracing is usually normal in these patients,
and the frequency spectrum corresponds to age. A central mu
rhythm is sometimes present (382,415), and there is reason to
presume that central spikes of childhood may be gradually Figure 26.28 A: An 8-year-old boy with epilepsia partialis continua or
replaced by mu rhythm, at least in some of the patients (Fig. focal motor status (clonic motions, left arm). The EEG shows consider-
26.27). There is indubitable evidence that, in a limited number able slowing (note lowered gain). There is evidence of constant muscle
of patients, the central spike activity can be blocked by con- activity, but no authentic cerebral spikes are demonstrable. B: Same
tralateral fist clenching or, even better, by alternate clenching patient. Electrocorticography. In general anesthesia, suppression of
and opening of the fist (45,415). This provides further evidence focal motor attacks but prominent spiking, especially in leads 4 (pre-
for the functional character of the spikes (Fig. 26.28). central gyrus) and 7 (postcentral gyrus).
Mitsudome et al. (432) have pointed out that, in benign
rolandic epilepsy, central spikes may be associated with local
slowing. Both rolandic spikes and local slowing were found to that omission of midline leads would result in missing the abnor-
respond to clonazepam. mality (436). Some of these patients show focal motor or sensory
In a small number of cases, the spike activity shows a consis- ictal activity of leg predominance; more often, the ictal symptoms
tent maximum over the vertex (433–435) or over the centropari- do not correspond with the spike localization. Even the parietal
etal midline region, which may be the sole region of spiking, so midline may be strongly involved in the spike activity (435).
516 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Clinical Signs of Nonictal Character
One must agree with Beaumanoir (382) when she points out
that neurologic deficits are not compatible with benign
rolandic epilepsy. This condition is based on dysfunction rather
than structural pathology. It is worthwhile, however, to search
carefully for a true intracranial lesion. Arteriovenous malfor-
mation, tumor, or cortical dysplasia occasionally may be the
cause of discharges and associated seizure.
Central, midtemporal, or parietal spikes or spikes over the
midline may also occur in children with evidence of cerebral
palsy, in mostly diplegic, quadriplegic, or choreoathetoid forms.
In these children, rolandic spikes have a different connotation
and do not herald a good prognosis for the seizure disorder.
The reader will find more extensive discussion in the section on
cerebral palsy.
Behavior disorders are very common in children with true
rolandic spikes; they may range from hyperkinetic behavior and
signs of minimal cerebral dysfunction to severe anxiety neuro-
sis. Various types of headaches may occur; see the symptom
profile of Gibbs and Gibbs (83). Migraine headaches may be
associated with the centro-temporal spike (Anderman, 341).
The intelligence is normal in true benign rolandic epilepsy.
Much research has recently been devoted to the neuropsy-
chological deficits that may accompany benign rolandic
epilepsy. There have been learning disabilities, attention deficit
disorder, attention deficit disorder with hyperactivity, impaired
motor and visual coordination, and specific language disorders.
Weglage et al. (437) correlated IQ deficits with greater than 6
spikes per minute in the waking EEG; children with 5 spikes or
less had no deficits. Staden et al. (438) found language dysfunc-
tion in 13/20 children and correlated this with greater than 10
spikes per minute. Croona et al. (439) demonstrated improve-
ment in the follow-up studies, after the spikes had resolved.
Massa (2001) correlated the following EEG findings with
cognitive and behavioral problems: intermittent slow wave
focus, multiple asysnchronous spike-wave foci, long spike-wave
clusters, generalized 3 Hz discharges, atonia, myoclonia or brief
alteration of awareness with the discharges, and abundance of
interictal epileptiform discharges. Nicolai et al. reported that
dysfunction was related to the spike-wave index.
Course
The seizures are easily controlled with routine anticonvulsive
treatment; oxcarbazepine and levetiracetam are preferable.
Treatment may even be withheld unless seizures repeat them-
selves (382). A population study in Canada demonstrated that
treatment is not required (440). The seizures typically remit
before 15 to 16 years of age. According to Lerman, the seizure
frequency is low, 13% may have just one seizure, 66% have
infrequent seizures, and 21% have frequent seizures (441).
Freedom from seizures in adolescence is the rule. Lerman
and Kivity (442) reported EEG normalization by adulthood in
all patients. The return of a single major convulsion may occa-
sionally occur under the influence of infections, stress, or toxic
substances. These cases show no resurgence of the central spike
focus, which renders the EEG diagnosis very difficult. The pres-
ence of central mu rhythm may serve as a hint that the patient
has had central spikes in the past, but such conclusions can be
made only with reservations.
A younger age of onset is associated with a tendency for more
frequent seizures (443). Massa et al. (444) identified the follow-
ing EEG characteristics as predictive of a “complicated” course:
intermittent slow wave focus; multiple, asynchronous spike-
wave foci; spike-wave clusters; generalized 3 Hz spike-wave dis-
charges; discharges with positive or negative myoclonia; or
abundance of interictal abnormalities during waking or sleep.
Etiologic Considerations
This form of epilepsy is due to dysfunction rather than to pathol-
ogy. A genetic basis is the most logical thought. The work of Bray
and Wiser (445) on familial occurrence of midtemporal spikes
has pioneered in this area, and other studies will certainly follow
suit. The work of Doose et al. (446) has further substantiated the
significance of genetic factors. A positive family history of
epilepsy was reported in 11.3% of 80 children with benign
rolandic epilepsy (447). Gelisse et al. (448) have pointed out that
benign rolandic epilepsy can also affect children with static brain
lesions, which, however, have no influences on the benign course.
These authors feel that neuroimaging procedures are usually
unnecessary in children with benign rolandic epilepsy.
The centro-temporal sharp waves seen in “benign rolandic
epilepsy” have an autosomal-dominant inheritance (449) and
are associated with a defect in the elongator protein complex 4
(ELP4), involved in transcription and tRNA modification (450).
Problems of Differential Diagnosis
Children with benign rolandic epilepsy must be differentiated
from the following groups:
1. Children with rolandic spikes and who have no seizures what-
soever (these children should not be considered as epileptics;
about 30% to 50% of the children with rolandic spikes have
no overt clinical seizures (451);
2. Children with rolandic spikes and a history of antecedent
brain damage or cerebral palsy;
3. Children who have typical psychomotor seizures and evidence
of temporal lobe epilepsy that may gradually progress in
severity; these children may have atypical spike localization
(central, midtemporal), whereas the classical anterotemporal
sharp wave focus does not materialize before adolescence;
4. Children with midtemporal spikes, marked tendency to gen-
eralization of spike-wave formation, clinically with aphasia
(Landau–Kleffner syndrome);
5. Children with frequent focal motor seizures that become
progressively worse: “malignant” rolandic epilepsy of child-
hood; and
6. Children with centroparietal spikes elicited by tactile stimu-
lation of corresponding cutaneous areas of the body.
The differentiation of these conditions rests on a careful com-
bined clinical–electroencephalographic assessment of each case.
Spike Foci Outside the Rolandic Region in Children
Occipital spike foci are usually found between the ages of 2 and
5 years. These children show no neurologic or ophthalmologic
 Chapter 26
deficit; about 40% of them have clinical seizures, mostly gener-
alized tonic–clonic, with good prognosis.
According to Gibbs and Gibbs (83), frontal spike foci in chil-
dren are associated with epileptogenicity, with about 80% hav-
ing overt seizures, and a guarded prognosis. Multiple spike foci
(two or more areas of independent spiking) are also highly
epileptogenic; the prognosis is guarded and probably fairly
good if rolandic spikes predominate.
A theory of spike migration of childhood from occipital to
midtemporal and then to conversion into 14 and 6 Hz positive
bursts, with good prognosis, or anterior temporal spike activity,
with poor prognosis, was advanced by Gibbs (452) but has found
little acceptance (see Ref. 404). “Focus-migration” must be
understood strictly in terms of dysfunction-induced epileptoge-
nesis, which may undergo changes of spatial origin and distribu-
tion. A structurally determined epileptogenic focus is unlikely to
display any wanderlust. Andermann and Oguni (453) and Blume
(454) have discussed the pros and cons of the focus-migration
theory; both authors have cautioned against the facile use of this
term. In a recent study, the migration of spike foci was correlated
with a better outcome in pediatric epilepsy: seizure frequency
and the number of prescribed AEDs were significantly greater in
those with a “fixed focus” versus a “migrating focus” (455).
Considerations of Basic Mechanisms
True benign rolandic epilepsy is likely to be based on temporary
paroxysmal hyperirritability of the motor cortex, which natu-
rally has a lower threshold of epileptic excitability. This is merely
a working hypothesis in need of further substantiating evidence.
Rasmussen Encephalitis and Other Causes of
“Malignant” Rolandic Epilepsy
Cases of progressively worsening focal motor seizures and pro-
longed episodes of epilepsia partialis continua (Kozhevnikov
■ Seizures and Epilepsy in Infants to Adolescents 517
syndrome) are quite rare but do constitute a special epileptologic
entity (456; Niedermeyer et al., 1977c; 457). Motor deficits and
mental decline are associated with the seizure disorder (Figs.
26.29 and 26.30). The etiology is poorly defined; chronic local-
ized encephalitis may be one of the causes (458–460). Chronic
encephalitis, or Rasmussen syndrome, is characterized by refrac-
tory focal seizures, frequently epilepsia partialis continua, and
progressive hemiatrophy with progressive contralateral hemi-
paresis (461,462). Although various autoimmune treatments,
such as IVIG, corticosteroids, or plasmapheresis have been used
(462,463), hemispherectomy seems to be the only effective cure,
and limited cortical excisions or lobectomies are ineffective (464;
Muto, 2010) (Fig. 26.30).
The EEG shows endless sequences of ictal spike discharges
during focal motor attacks but may not demonstrate any elec-
trographic seizure activity during epilepsia partialis continua,
which probably originate from deep structures or possibly from
lamina V of the motor cortex without participation of the
superficial layers (Elger and Speckman 1979, 1980, 1983).
A recent European conference outlined the following two-
tiered clinical, EEG, and MRI criteria (A and B) for diagnosis.
Tier A includes: clinical: focal seizures, with or without epilep-
sia partialis continua and unilateral cortical deficits; EEG: uni-
hemispheric slowing with or without epileptiform activity and
unilateral seizure onset; MRI: unihemispheric cortical atrophy
and at least one of either gray or white matter T2/FLAIR hyper-
intensity or hyperintense signal or atrophy of the caudate. If all
three of A are present, this confirms the diagnosis, or if not, 2
out of 3 of B: epilepsia partialis continua or progressive unilat-
eral cortical deficits, on MRI, progressive unihemispheric corti-
cal atrophy, or histopatholgy showing encephalitis with
activated microglial cells and reactive astrocytosis (461).
A malignant rolandic-sylvian seizure syndrome has been
described in children consisting of refractory sensorimotor
Figure 26.29 Ictal spiking in
right frontal leads (depth and
scalp) associated with left facial
twitching in an 8-year-old boy with
chronic encephalitis, not relieved
by right frontal lobectomy. Much
improved after right hemispherec-
tomy at age 15.
518 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Genetic screening for the nicotinic acetylcholine receptor gene

mutation (CHRNA4 and CHRNB2) is commercially available.

Benign Occipital Lobe Epilepsy

This epileptologic entity has been individualized by Gastaut
(468) and updated by Taylor et al. (469,470). It has been divided
into early onset, or Panayiotopoulos syndrome, or later onset,
Gastaut syndrome. However, Panayiotopoulos syndrome is now
considered an autonomic, rather than an occipital epilepsy (471).
Benign occipital epilepsy occurs in children and adolescents;
there is no persistence into adult life. Seizures are mostly initiated
by visual symptoms (amaurosis, phosphenes, or figurative hallu-
cinations), which may be followed by other seizure manifesta-
tions (mostly hemiclonic attacks, but also psychomotor seizures
and grand mal). Migrainous or pseudomigrainous symptoms
with headache, nausea, and vomiting often follow the seizure.
The EEG shows very characteristic changes. In the interval,
there are frequently recurrent stretches of rhythmical and well-
formed spike-wave discharges of high voltage, at 1.5 to 3 Hz,
located over occipital regions and the immediate vicinity.
Moderate lateralization is common. Ictal EEG discharges consist
of continuous spike-wave activity over one occipital region even
though the interictal spike-wave pattern is bilateral (468). Figure
26.31 shows an example of the occipital spike-wave discharge.

Figure 26.30 The patient discussed in Figure 26.29 after right hemi-
spherectomy. Note the almost complete loss of activities on the right.

seizures, typically presenting as “tics,” frontotemporal EEG

spikes, rolandic-sylvian spike sources on dipole analysis or
MEG, and cognitive regression (465). However, even with very
frequent seizures, remission may occur (466).

Autosomal-Dominant Nocturnal
Frontal Lobe Epilepsy
Patients with autosomal-dominant nocturnal frontal lobe
epilepsy have clinical manifestations that may include sudden
awakenings with dystonic movements (tonic posturing) or vio-
lent movements and the hypermotor seizures or tonic postur-
ing may often be brief and nocturnal as the name suggests. The
EEG findings may be relatively bland or may show focal fea-
tures such as slowing and spikes. Ictal recordings are character-
ized by rhythmic frontal discharges with a diffuse frontal
predominance. The ictal discharges may lateralize or even local-
ize to a region suggesting a focal structural lesion. However,
interictal and ictal epileptiform activity may be lacking and the
seizures mistaken for sleep disorders, nocturnal paroxysmal
dystonia, or nonepileptic seizures. In a study of 40 patients,
interictal and ictal EEGs were normal in 26% (Oldani, 1998) Figure 26.31 A 13-year-old boy with benign occipital lobe epilepsy.
and in a study of children, the interictal EEG was normal in 3/8 There are widespread 1.5 to 2 Hz spike-wave complexes with occipital-
but ictal EEG was abnormal in all showing bifrontal beta activ- posterotemporal maximum, occurring in long bursts during wakeful-
ity with or without rhythmic slowing (467). ness. The spike waves are lateralized to the left (but the lateralization
A careful family history is warranted before considering focal was not consistent). At other times, the patient had a well-developed
respective surgery in people with nocturnal frontal lobe seizures. posterior alpha rhythm.
 Chapter 26
When one considers the indubitably existing relationship to
migraine, the question arises as to whether there is a continuum
of more or less epileptogenic conditions between benign occip-
ital lobe epilepsy and forms of basilar artery migraine with
paroxysmal posterior EEG changes (472–474). Lugaresi et al.
(475) reported cases of “scotosensitive seizures” elicited by dark-
ness and eye closure; these authors feel that this seizure type is
closely related to benign occipital lobe epilepsy.
Benign occipital lobe epilepsy is unassociated with photosen-
sitivity; there are no paroxysmal responses to flickering light.
Epilepsy w ith Myoclonic Absences
(Myoclonic Absence Epilepsy)
This was first described by Tassinari (476,477). Epilepsy with
myoclonic absences consists of myoclonic absence seizures with
variable alteration of awareness associated with bilateral rhyth-
mic myoclonic jerks that may have a tonic component (478). The
myoclonias typically involve the shoulders, arms, and legs with a
typical progressive elevation of the arms. Eyelid involvement is
rare. The myoclonic absences have an abrupt onset and cessation,
similar to CHILDHOOD ABSENCE EPILEPSY (CAE). Simple
absence seizures, generalized tonic–clonic seizures, and drop
attacks may occur. This is now classified as an absence seizure
with special features, the myoclonic absence (235).
The interictal EEG has a normal background. Generalized
spike waves may occur, and rarely, these discharges may have
focality. The ictal discharge is a 3 Hz spike and wave discharge.
In one study of 14 children (479), chromosomal abnormali-
ties were present in 7, including Angelman syndrome, 12p tri-
somy, and an inversion/duplication of chromosome 15.
Lennox–Gastaut Syndrome
This is a severe, intractable, mixed seizure disorder with tonic,
atonic, and atypical absence as the “core” seizures, although gen-
eralized tonic–clonic seizures, myoclonic seizures, atypical
absence, and focal seizures do occur. Tonic and atonic seizures
are referred to as “drop attacks.” Mental retardation is frequent,
and a slow spike and wave pattern is seen on the EEG. This is a
typical example of an “epileptic encephalopathy.” Gibbs et al.
(77) first noticed the severity and poor prognosis of the seizure
disorder in patients with slow spike-wave complexes (called the
“petit mal variant,” in contrast to the classical 3 to 4 Hz spike-
wave complexes in patients with petit mal absences and the
prognostically more favorable “primary generalized epilepsy”).
A detailed account of the clinical and ictal symptomatology of
these epileptics was given by Lennox (78), and further important
work in this field was done by Gastaut et al. (40). The clinical
features were codified by Dravet in her doctoral thesis. The term
Lennox–Gastaut syndrome was introduced by Niedermeyer
(44,480). The work of Oller-Daurella (481–483) deserves special
mention in this context.
Tonic seizures are a major seizure type in LGS and strict cri-
teria for LGS require tonic seizures for the diagnosis (484).
Tonic seizures are typically sleep activated and repetitive,
although long-term monitoring may be needed to discover
these, as the tonic component may be subtle. LGS is typically
resistant to therapy with standard AEDs and alternative thera-
■ Seizures and Epilepsy in Infants to Adolescents 519
pies such as corticosteroid therapy, the ketogenic diet, or vagus
nerve stimulation (VNS) may be helpful.
The EEG hallmark of LGS is the slow spike and wave dis-
charge (1.5 to 2.5 Hz) superimposed on a slow background.
Bursts of fast rhythms (10 Hz) are also a prerequisite; these may
be frequent, sleep-activated, and typically occur during the
tonic seizure. Multifocal spikes may also be seen and in some
cases may be the predominant epileptiform activity.
A multitude of etiologies may cause this condition, and in
many cases the cause remains unknown. According to Gastaut et
al. (485), the prevalence of the Lennox–Gastaut syndrome in a
major epilepsy center (5.1% with 10.2% of patients below age 15
and 0.6% of patients above this age) is greater than infantile
spasms. LGS usually starts between the ages of 1 and 10 years;
onset in the second decade of life is much less common (16% after
Komai (486)) and adult onset is rare (487). Onset at age 6 to 12
months has been observed and requires solid EEG documentation
for differentiation from hypsarrhythmia-infantile spasms. We have
observed a form of epilepsy intermediate between the two that we
describe as late infantile epileptogenic encephalopathy (244).
About 10% to 20% of the children with LGS have passed through
a period of infantile spasms-hypsarrhythmia before the LGS
becomes evident (44–46). The transition from infantile spasms to
LGS has been investigated by Olmos-Garcia de Alba et al. (488).
Even in LGS, EEG focality may suggest a hidden focal process.
Ictal Manifestations
The types of seizure occurring in the Lennox–Gastaut syn-
drome are best divided into the following groups (Table 26.7).
Tonic seizures show a variety of manifestations; see Gastaut
and Broughton (238) and their subdivision into axial, axorhi-
zomelic, and global tonic seizures. The attacks are short and last
from about 5 to 20 seconds. There is extension of the axial mus-
culature with some opisthotonus; moderate flexion of the arms is
noted, and this may be followed by extension. Tracheobronchial
hypersecretion occurs with repeated attacks, and a fairly danger-
ous status may evolve (489). Tonic seizures are quite common in
non-REM sleep and often are observed in a routine sleep tracing.
Bilateral synchronous fast or moderately fast spike activity of
about 10 to 25 Hz of medium to high voltage and frontal accen-
tuation is the EEG concomitant of these attacks (“runs of rapid
spikes”). Electrodecrements may also occur. A diffuse slow ictal
pattern has been mentioned by Gastaut and Broughton (238),
who consider it extremely rare.
EEG
In the atypical absence seizure, the clinical onset and termina-
tion are less abrupt than in classical absences and may be asso-
ciated with marked changes in posture and tone. Atypical
absence status epilepticus may occur.
EEG findings are crucial in the diagnosis of Lennox–Gastaut
syndrome. The outstanding feature is the slow spike-wave com-
plex ranging from 1 to 2.5 Hz. It is more often an interictal rather
than an ictal discharge, is most often seen in a generalized
synchronous pattern, although lateralization is also fairly com-
mon. Focal slow spike-wave activity is quite rare. A maximum
over the frontal midline is the rule (Fig. 26.32). This discharge is
520 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Tabl e 2 6 . 7

Types of Seizure Occurring in Lennox–Gastaut

Syndrome

Incidence in
Lennox–Gastaut
Syndrome
Seizures occurring in other seizure
disorders
Grand mal (generalized tonic–clonic)
Psychomotor (complex partial)
Myoclonus
Focal motor and other focal (elementary
partial)
Adversive
Seizures occurring almost exclusively in
the Lennox–Gastaut syndrome
Tonic
Atonic drop (with tonic and myoclonic
elements)
Clonic (rapidly repetitive clonic
movements)
Hemiclonic

enhanced in non-REM sleep and may become almost continu-

ous. Such an abundance of generalized slow spike-wave activity
in non-REM sleep, however, must be carefully distinguished
from the condition known as electrical status epilepticus during
sleep in children (ESES) syndrome (490,491).
Another important pattern is runs of rapid spikes (Figs. 26.33
and 26.34), which are seen in non-REM sleep only. This pattern
is more common in older children, adolescents, or adults. The
EEG as such (“background EEG”) is often disorganized and
excessively slow, some patients show an unremarkable posterior
alpha rhythm, the basic rhythm. The degree of general slowing
and disorganization usually underscores the severity or
advanced stage of the case.
These children are often overmedicated with anticonvulsants, Figure 26.32 A: Tracing of an 8-year-old boy with Lennox–Gastaut syn-
resulting adverse effects including somnolence and incoordina- drome (mentally retarded, uncontrolled psychomotor automatisms, and
tion. Interestingly, high and toxic levels of phenobarbital may be occasion grand mal). Note the irregular slow (1.5 to 2.5 Hz) spike-wave
completely unassociated with fast EEG activity; this absence of complexes. B: Pronounced slow (about 1.5 Hz) spike-wave complexes in
barbiturate-induced fast frequencies is quite characteristic in generalized synchrony with superior frontal maximum in an 11-year-old
advanced cases with considerable cerebral impairment (492). patient with Lennox–Gastaut syndrome. The patient was awake at the
The adult patient may gradually lose the characteristics of the time of this recording; there were no clinical ictal manifestations.
Lennox–Gastaut syndrome, as the slow spike-wave complex may
disappear after age 20 or 15, replaced by only runs of rapid spikes
in sleep. This pattern, too, may gradually vanish and spikes or (495) secondary temporalization is a very common development
sharp waves of temporal and especially anterotemporal localiza- (78% of their patients).
tion may then become predominant. This change may be associ- Clinical entities need revisiting from time to time. Refreshing
ated with the development of complex partial seizures of a new thoughts have been presented by Ohtahara et al. (496) who
temporal origin. This process has been described as “secondary have reemphasized the wide borderland of this syndrome, the
temporalization” (45,493,494). According to Hughes and Patil role of focal cortical elements and secondary bilateral synchrony.
 Chapter 26
Figure 26.33 A run of rapid spikes in a 14-year-old patient.
These authors therefore advocate a more refined subclassifica-
tion, especially in view of the suitability of neurosurgical treat-
ment (callosotomy or vagus nerve stimulator).
The highly critical approach of Hirt (497) attempts to cast
doubt on practically all of the electroclinical diagnostic pillars
of the diagnosis of Lennox–Gastaut syndrome. Such negativis-
tic analyses could be done in most clinical diagnostic entities;
with sufficient destructive nihilism, the “gestalt” of diseases and
syndromes could be turned into an empty and worthless con-
struct. Criticism is always welcome but has to be coupled with
a constructive attitude.
Electrical Status Epilepticus during Sleep
(ESES Syndrome)
This syndrome is characterized by continuous slow spike-wave
activity during non-REM sleep. The spike-wave complexes are
not well formed, and the spike component is usually more pro-
nounced than the slow wave component. The maximum is vari-
able; a maximum over the vertex but also over the occipital
region may be present, according to personal observations.
This electroclinical condition has been described by Patry et al.
(490) and Tassinari et al. (491,498). ESES was originally defined as
having spike and wave discharges present during 85% of slow
wave sleep. This disorder occurs in children (mostly around age
8 years) with chiefly nocturnal but also diurnal seizures and
■ Seizures and Epilepsy in Infants to Adolescents 521
Figure 26.34 A tonic seizure in light sleep in a 20-year-old patient
with Lennox–Gastaut syndrome. A generalized run of rapid spikes is
preceded by a few large slow waves mixed with spikes.
mild mental retardation. The seizures consist of generalized
tonic–clonic and absence seizures. The unique EEG manifesta-
tions during non-REM sleep are coupled with rather bland EEG
findings in the waking state and REM sleep; occasional focal or
bilaterally synchronous spikes may be present while awake. The
dramatic changes in non-REM sleep disappear during adoles-
cence. From the viewpoint of the sleep researcher, it is interesting
to note that these children awaken from such a severely disturbed
nocturnal sleep pattern in a normally refreshed state. Although
also referred to as continuous spike-waves of slow sleep, this is
not absolutely correct, in that the discharges are activated in
non-REM sleep, and typically at the onset of drowsiness.
This disorder is rare, occurring in 0.2% of pediatric epilepsy
syndromes (499) and is divided into symptomatic and crypto-
genic groups, determined by whether normal neurologic or psy-
chomotor development was present before the onset of the
CSWS. These children commonly have seizures, but the hallmark
is regression. Tassinari reported 29 children with CSWS. All except
one child had seizures; one had a single seizure and one had only
three seizures. Eighteen had normal, and 11 had abnormal psy-
chomotor development prior to onset. In the 18 with normal
development, all had severe loss of IQ and behavioral distur-
bances, defined as decreased attention span, hyperactivity, aggres-
sion, difficulties with interaction and inhibition, and 2 patients
developed a psychotic state. In the 11 with abnormal psychomo-
tor development, mental deterioration occurred in all, 3 devel-
oped a marked hyperactivity, and 1 showed “massive regression”
including language and a loss of interest in all activities.
The regression occurs in a more generalized manner, predom-
inantly with cognitive and behavioral manifestations, rather than
predominantly the language regression seen in Landau–Kleffner
522 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
syndrome (LKS). Guilhoto and Morrell (500) reported that a
more focal EEG pattern is associated with language regression,
whereas if the EEG pattern is more generalized, the regression is
more global, and not predominantly in language skills.
There has been a recent upsurge of interest in ESES and (LKS).
New waves of research in this domain have resulted in a book
dealing with both conditions that was edited by Beaumanoir et al.
(1995). There have been advocates of a unified syndrome com-
prising both conditions (501) and also opponents of this idea.
Beaumanoir (1996) as well as Tassinari (1996) have expressed a
cautious view aiming at preservation of the status quo.
LANDAU–KLEFFNER SYNDROME
(ACQUIRED EPILEPTIC APHASIA)
Landau–Kleffner syndrome (LKS) (502) is also a rare epileptic
syndrome, occurring in 0.2% of pediatric epilepsies (499) and
is characterized by language regression and an abnormal EEG.
LKS usually develops in a previously normal child older than 4
years and may first manifest as an apparent word deafness, a
“verbal auditory agnosia.” Seizures and behavior disturbances,
particularly hyperactivity, each occur in approximately 2/3 of
the children. The majority are classified as idiopathic, although
LKS may be caused by any pathologic process affecting auditory
cortex. We have seen “symptomatic” cases, one with a left tem-
poral lobe tumor and another with a left sided middle cranial
fossa arachnoid cyst. Therefore, neuroimaging is needed.
The term LKS variant refers to children without the classic
features. These include children with involvement of more ante-
rior language areas with dysfunction characterized by an expres-
sive disorder with oral-motor apraxia, sialorrhea, seizures, and
an abnormal EEG (centro-temporal spikes similar to those seen
with benign focal epilepsy), children with PDD (autism) with
language regression and abnormal EEGs, and even children with
congenital aphasias, also called developmental language disor-
ders, with epileptiform EEGs. In our experience, children with
PDD comprise the largest group of the “variants.”
The EEG in LKS shows bilateral, multifocal spikes, and spike
and wave discharges, occurring usually in the posterior regions,
especially the temporal region, with a marked activation during
sleep. However, discharges occur in many locations, and may
even be generalized. Some centers require ESES to diagnose
LKS, but the strict ESES definition would exclude many cases.
However, sleep-activated epileptiform discharges are required
for the diagnosis. The EEG may improve over time, either spon-
taneously or with treatment and the EEG abnormalities may be
an epiphenomenon.
The evaluation of ESES or LKS includes a baseline history,
physical examination, and sleep-deprived EEG, a formal neu-
ropsychologic evaluation, and neuroimaging, with MRI pre-
ferred. Although long-term video-EEG monitoring (LTM) has
been required, the presence of strict ESES in an overnight EEG
may be predicted by the abundance of sleep-activated epilep-
tiform features with drowsiness. In certain cases, spike map-
ping or functional neuroimaging with either SPECT or PET
scan, and magnetoencephalography (MEG) might be helpful,
especially in the child with medically refractory LKS. The
frequency-modulated auditory evoked response (FM-AER)
may differentiate an expressive from a receptive aphasia, and
localize the superior temporal gyrus (503), but is not widely
available. All children should have intensive speech therapy
and in our experience, even if the EEG activity is controlled,
there is a persistent language disorder.
CHILDHOOD ABSENCE EPILEPSY (CAE)
Absence seizures were discussed previously under seizure types.
This section shall refer more to the syndrome of CAE rather than
absence seizures. For syndrome classification, historically there are
two different forms of absence seizures: (i) simple absence
seizures, starting at age 4 or shortly thereafter, with a large num-
ber of absences per day (sometimes exceeding 100/day); and
(ii) juvenile absence seizures with an onset at age 9 to 15 years,
more prolonged or mixed with more motor activity. The absences
of the first type have been called “pycnoleptic” (237,315,504),
using an old term introduced by Sauer (504), which characterizes
the high frequency of the attacks. The absences with onset around
age 10 to 15 have been called “spanioleptic” (“spanios,” meaning
seldom, because of their smaller number of daily seizures) (397).
A cyclic change of the frequency has also been noted (“cycloleptic
attacks”) (237,315). This distinction (pycnoleptic vs. spanioleptic
absences) has been doubted by Loiseau et al. (506). In a study of
various absence syndromes, there were no syndrome-specific fea-
tures to absence seizures (507). The absence seizures were shorter
in Juvenile Myoclonic Epilepsy (JME) and there was greater
awareness.
Children with childhood absences often start having gener-
alized tonic–clonic seizures in early adolescence, with the inci-
dence ranging from 31% (508) to 54% (509), to Matthes (315),
who reports that about two thirds of the children with petit mal
absences will have generalized tonic–clonic seizures. In most of
these cases, the generalized tonic–clonic seizures do not pose a
major problem and are readily brought under control. Janz
(510,511) has stressed the role of the stage of awakening as a
strongly facilitating factor in these major convulsions. This view
has shed much light on the pathophysiologic basis of these
attacks and deserves full support.
In a recent study of childhood absence seizures, using strict
classification criteria, only 5/47 (11%) would be classified as
CAE (269).
Regarding the EEG findings, the organization and regularity
of spike-wave discharges are greater (less fragmented) in CAE,
than in other absence syndromes (253,254). There is a tendency
for the spike-wave frequency and morphology to vary with
absence syndromes: 3 Hz with CAE, faster spikes with Juvenile
Absence Epilepsy (JAE), and polyspikes with JME (512).
CAE has been associated with GABARA1 and GABARG2
(5q34), chloride channel 2 (3q26), and the voltage-dependent
T-type alpha1H subunit of the calcium channel (16p13).
EYELID MYOCLONIA WITH ABSENCES
(JEAVONS SYNDROME)
This seizure type is now classified as an absence seizure with
special features, eyelid myoclonia (235). The eyelid myoclonia
and absence seizures may occur independently. Eyelid myoclo-
 Chapter 26
nia typically occurs immediately after eye closure, is associated
with brief bilateral spike and wave activity, is more frequent in
bright light and does not occur in the dark. Eyelid myoclonia
looks like rapid blinking, with upward deviation. However,
eyelid myoclonia occurs in other idiopathic generalized epilep-
sies, especially with photosensitive generalized epilepsy, and
have occurred in other family members without seizures,
referred to as paroxysmal eye movements (513). The frequency
may help to differentiate these: the eyelid flutter with an absence
seizure has a frequency of 3 Hz, eyelid myoclonia has a fre-
quency of 4 to 6 Hz, and paroxysmal eyelid movements have a
frequency of 10 Hz (513).
ADOLESCENCE
Juvenile Absence Epilepsy (JAE)
The onset of JAE onset is usually around puberty, from 9 to 17
years of age, but in contrast to the very frequent seizures in
CAE, the absence seizures in JAE may be less frequent but more
prolonged and have more associated motor activity. GTC and
myoclonic seizures may also occur. There are few outcome
studies of JAE. In a larger study of 64 patients, only 24/64 (37%)
were seizure free after a mean follow-up of 22 years; in a smaller
study, only 8/17 (44%) were seizure free after a mean follow-up
of 6 years (range 2 to 12 years) (514). Both studies found that
those with GTC seizures had less chance of remission.
The EEG background is normal in JAE. The spike-wave dis-
charge is usually faster than 3 Hz (3 to 4 Hz), and similar to
CAE, may be faster, 4 to 5 Hz, at the onset. The spike-wave dis-
charge is less organized (more fragmented) than in CAE, but
has greater organization (less fragmented) than JME (253,254).
Focal features have been reported (515). JAE has been associ-
ated with a defect in the chloride channel 2 (3q26) and the EF-
hand domain (C-terminal)-containing 1 (6p12-11).
Juvenile Myoclonic Epilepsy (JME)
JME of Janz is characterized by mild myoclonic, generalized
tonic–clonic, sometimes clonic–tonic–clonic seizures and
absence seizures and has been referred to as impulsive petit mal.
Onset is typically around puberty. The myoclonic seizures are
usually mild-to-moderate in intensity, may involve the entire
extremity, rather than isolated muscles, and generally are bilat-
eral. These myoclonic seizures typically occur after awakening
and are aggravated by fatigue, sleep deprivation, or alcohol. The
patient may drop objects or fall. Myoclonus may also be associ-
ated with brief petit mal absences and myoclonic status epilep-
ticus has occurred. GTC seizures are frequent.
Lennox (279) conceived of this combination (“myoclonic petit
mal”) as a part of his “petit mal triad.” Janz (237) has extensively
investigated this type of attack and uses the term impulsive petit
mal; Matthes (315) prefers the term Janz syndrome. An extensive
study on myoclonic petit mal in 399 patients with genetic evalua-
tion has been carried out by Tsuboi (516). Delgado-Escueta and
Enrile-Bascal (517) first stressed the good response to sodium val-
proate and hence the ultimately good prognosis, which contrasts
with the experience prior to valproate. This is similar to the find-
ings of Asconape and Penry (518). Further information is found
in the work of Schmitz and Sander (519).
■ Seizures and Epilepsy in Infants to Adolescents 523
Delgado-Escueta et al. (1984) reported that GTC seizures
occurred in 41/43, and Asconape and Penry (518) reported GTC
seizures in 83%. Myoclonic jerks usually precede the GTC and
may evolve into a GTC seizure. GTC seizures usually occur shortly
after awakening. Absence seizures occurred in 40% in the series of
Delgado-Escueta et al. (1984). These usually occur in association
with GTC seizures, and most commonly occur shortly after awak-
ening. Postmyoclonic inhibition may occur (520).
The interictal EEG in JME is characterized by a fast spike
and wave discharge: 3.5 to 6 Hz spike and wave and polyspike
and wave complexes. However, patients may have spike-wave
discharges less than 3.5 Hz (521). Ten to sixteen hertz rapid
spikes with slow waves occur with myoclonic seizures, photo-
sensitivity is common, and abnormalities may occur in sleep.
Although classified as “generalized epilepsy,” focal EEG abnor-
malities have been reported (522,523); in one study, focal
spikes, focal sharp waves or slow waves occurred in 37% of
EEGs (524), and in another 12/85 had focal EEG findings,
12/85 had focal clinical findings, and only 2/85 had both clini-
cal and EEG focality (Asconape, 1984). The interictal EEG has
been normal in 6% (521) or 10% and sleep deprivation may
activate epileptiform activity and the discharges may have a
longer duration (525).
JME, similar to JAE, may require lifelong treatment. Twelve
out of thirteen patients who discontinued valproate after seizure
freedom had recurrence (517). In 24 patients followed for 25
years, only 175 were seizure free and 13% had myoclonus after
stopping AEDs (526). In addition, children with absence epilepsy
without remission (527) and patients with JAE (277) may evolve
into JME.
The calcium channel, voltage-dependent beta-4 subunit
(2q24), chloride channel 2 (3q26), GABA1RA (5q34), GABAR
delta (1p36), and the EF-hand domain (C-terminal)-containing
1 (6p12-11) have been associated with JME.
Epilepsy w ith GTC Seizures upon Awakening
In this disorder, GTC seizures occur exclusively or predomi-
nantly shortly after awakening. The onset is usually in the sec-
ond decade. Absence or myoclonic seizures may occur. The
EEG shows the fast spike and wave burst, similar to JME. These
bursts are relatively short and dominated by 4 Hz or 4 to 5 Hz
spike-wave complexes that are, contrary to the classical 3 Hz or
3 to 4 Hz spike waves, not readily activated by hyperventila-
tion. Many of these have photosensitivity (photoconvulsive
response), and positive family histories of epilepsy are often
obtained in this group. This syndrome may represent JME
without myoclonus.
CHILDREN WITH MULTIFOCAL SPIKES
(MULTIFOCAL INDEPENDENT SPIKE
SYNDROME, MISS)
Multiple independent spikes are defined as three or more inde-
pendent spike foci, from noncontiguous electrodes, with at least
one focus originating in each hemisphere and the spikes may
vary from different locations over time. First described by
Noriega-Sanchez and Markand (1977) in 108 patients, prior
EEGs could show a hypsarrhythmia or slow spike and wave
524 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
pattern, and the most common age range was 4 to 7 years (40%).
Blume reported multiple independent spikes, from noncontigu-
ous electrodes, in 63/1500 (4.2%) of EEGs (528). This disorder
was termed severe epilepsy with multiple independent spike foci
by Ohtahara (529), with seizures consisting predominantly of
tonic spasms. Kotagal (530) supported this as a distinct electro-
clinical syndrome. The MISS may occur following infantile
spasms or LGS. Background slowing and sleep activated dis-
charges are common, and similar to infantile spasms, increased
synchrony may occur during sleep.
Some children with rather dramatic EEG findings show sur-
prisingly normal neurologic and mental findings. In the latter
group, the widespread generation of spike activity is likely to be
due to temporary paroxysmal hyperexcitability. Genetic predis-
position has been suggested by Hauser et al. (531). Diffuse EEG
background slowing differentiates these benign epilepsies from
the pure MISS.
PRAGMATIC USE OF THE EEG IN DIAGNOSIS
Clinical evaluation of a presumed epileptic individual without
the use of the EEG is almost unthinkable. As a function-
oriented test, the EEG is ideally suited for the demonstration of
epileptiform abnormalities.
Epileptic seizures are necessarily accompanied by an electro-
graphic seizure, which is a transient hypersynchronous neu-
ronal discharge consisting of repetitive epileptiform discharges
that have an evolution in frequency, amplitude, and morphol-
ogy in addition to a temporal and spatial evolution. Single
spikes or sharp waves do not constitute an electrographic
seizure. However, these may be undetectable on the scalp
because the generator is buried in a cortical sulcus or in deep
limbic or subcortical regions. A limitation of the ictal firing to
lamina V may be responsible for a lack of spikes on the cortical
surface (532). Without concomitant activation of larger cortical
areas, deep discharges may not make it to the surface, where the
recording electrodes are routinely placed. Invasive EEG record-
ings must be restricted to only those who are candidates for
seizure surgery, and even certain local epileptic events may not
be detected. Invasive monitoring is myopic: the recording is
from a very limited area.
Alternatively, epileptiform discharges may also occur in a
substantial number of patients without epilepsy (221,533).
Hence, epileptic discharges may be either too concealed or too
readily available, enticing the electroencephalographer to
imprudent interpretations. It must be kept in mind, however,
that epileptiform discharges in nonepileptic patients are false-
positive findings in a very limited sense only; they are not
meaningless. Spikes in nonepileptics clearly indicate a certain
predisposition to an epileptic seizure or may foreshadow a
future seizure disorder. The diagnosis of an epileptic seizure
disorder must never be made based solely on EEG spikes alone.
Useful practical guides for routine studies in epilepsy have been
presented by Engel (534) and Dinner and Lüders (535).
In the course of the development of epilepsy monitoring
units, the clinical value of interictal spikes has been belittled by
most investigators, who, in these special units, are hunting for
seizure documentation, that is, ictal rather than interictal pat-
terns. The interictal spike discharges, however, must not be
underrated as an important indicator of paroxysmal activity
(21,536). In the epilepsy surgery evaluation, interictal spikes
define the irritative zone.
USE OF ACTIVATION METHODS
Various activation methods have been used to activate otherwise
undetectable epileptiform activity, such as hyperventilation and
intermittent photic stimulation. In most major North American
EEG laboratories, sleep, either natural or with sedation, is a stan-
dard part of the recording. Sleep deprivation further facilitates
epileptiform features, is a true activation, and avoids concerns
regarding conscious sedation. A whole list of procedures can be
put together; all of them aim at the demonstrability of otherwise
hidden epileptic discharges (Table 26.8). The timing of the
recording may be important. Focal as well as generalized spiking
may activate shortly after a seizure (537–539), but postictal sup-
pression or attenuation of a focus may also occur. A very useful
review of the practical value of EEG studies in epileptic patients
has been presented by Engel (534).
Eye opening and closing will activate certain forms of visu-
ally induced epilepsies, especially eye closure. Hyperventilation
Tabl e 2 6 . 8
EEG Activation Methods for Enhancement of
Paroxysmal Discharges
A. First-line methods
Eye opening and closure
Hyperventilation
Intermittent photic stimulation (flicker)
Sleep (short diagnostic non-REM sleep)
Cautious reduction of anticonvulsive medication (timing
depending on type of medication and its metabolism)
Sleep deprivation
Timing of EEG recording to menstrual cycle
(premenstrual hydration)
B. Second-line methods (mostly historical)* indicates historical
Convulsants*
Pentylenetetrazol (Metrazol) i.v.
Bemegride (Megimide) i.v.
Combined flicker and pentylenetetrazol i.v.*
Methohexital (Brevital) i.v.
Use of special trigger mechanisms (cutaneous stimuli,
viewing geometrical structures, visual exploration,
startling stimuli)
Induced hypoglycemia (insulin, fasting)*
Hypoxia (breathing nitrogen)*
Alpha-chloralose*
C. Use of special technology for greater length of recording
Long-term monitoring
 Chapter 26 ■ Seizures and Epilepsy in Infants to Adolescents 525

is the classical activation for childhood absences and the 3 Hz
spike-wave discharge; it is much less effective in other epileptic
seizure disorders but is capable of inducing practically every
type of seizure or EEG discharge. Even in some cases of tempo-
ral lobe epilepsy, hyperventilation may be remarkably informa-
tive. Children below the age of 3.5 years usually do not
cooperate with this test; some older children enjoy it. The yield
of intermittent photic stimulation is enormous, although lim-
ited to a rather small domain of epileptology, but it may induce
epileptiform responses in nonepileptics.
Sleep records made in periods of deep drowsiness, stage 1 and
non-REM sleep and stage 2, usually 10 to 30 minutes, are an
important part of the EEG evaluation. In generalized as well as
in focal (partial) epilepsies, the epileptiform discharges may be
limited to light sleep. The informative value of short sleep trac-
ings with or without sedation has been emphasized by Gastaut
et al. (540,541) and Niedermeyer and Hšller (542). Long-term
EEG monitoring has demonstrated the increase of spike activity
during nocturnal sleep (Martins da Silva et al., 1984).
Sleep deprivation works differently; the important element
in facilitation of epileptiform discharges is not the sleep itself,
but the fatigue; it is important that the patient not be allowed
to falling asleep in the laboratory. An impressive bulk of literature
on sleep deprivation has accumulated during the past 15 years.
More information is found in the work of Binnie et al. (543),
Degen and Degen (544), and Ellingson et al. (545), to name just
a few of the numerous investigators in this field.
Bilateral synchronous spikes, polyspikes, and spike-wave
complexes are usually enhanced in light non-REM sleep
(mainly stage 2) and may be found in conjunction with K com-
plexes. The role of arousal has been pointed out in detail by
Niedermeyer (45,72,414,546–548), who introduced the term
dyshormia, which means faulty or deviant arousal. It is thought
that arousing stimuli generate K complexes contaminated with
spikes. These K complexes are maximal over the frontal midline
(Fz electrode) and thus differ from the vast majority of K com-
plexes that have a vertex (C2) maximum (Fig. 26.35). A
schematic view is shown in Figure 26.36. These data and the
resulting concept have been confirmed by the work of Halász
and his coworkers (61,549–553). Nocturnal (all night) sleep
recordings also demonstrate an increase of generalized paroxys-
mal discharges in light non-REM sleep (554,555). Massive

Figure 26.35 A: Spike-wave discharges with frequently intermixed polyspikes in a 30-year-old patient with unusually late
onset of petit mal absences at age 18. Stage 2, non-rapid eye movement (non-REM) sleep. Marked frontal maximum.
B: Same patient, stage 2, non-REM sleep. Transversal montages clearly demonstrate a voltage peak (based on phase rever-
sal) of spike waves and polyspikes over frontal and central midline, but when Fz and Cz are run against each other (chan-
nels 15 to 16), it becomes clear that the frontal midline is the maximally involved area.
526 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 26.35 (Continued) C: Spikes and polyspikes riding on K com-
plexes in stage 2, non-REM sleep in an 18-year-old patient. Well-defined
maximum of spiking over frontal midline (Fz ). (From Niedermeyer E.
Generalisierte epilepsien. EEG Labor. 1979;1:119–131.)
arousal leading to immediate awakening usually does not pro-
duce epileptic K complexes; more subtle arousing stimuli are
more effective. Age may be a factor: the sharp component of the
K complex is particularly prominent after the age of 4 years and
starts to decline during the second decade. Such naturally sharp
K complexes apparently facilitate paroxysmal responses.
There may even be an association of spike bursts and spin-
dles in patients with generalized epilepsies (556). This may be
similar to the aforementioned dyshormia concept. The rela-
tionship of sleep spindles and spike-wave bursts is bolstered by
the experimental work of Gloor and his coworkers (557–559).
The reduction of long-standing anticonvulsive medication
must be handled with great caution and we typically limit this to
inpatients in the Epilepsy Monitoring Unit, since status epilepti-
cus may be precipitated. Interestingly, interictal focal spike activ-
ity is not significantly enhanced during the withdrawal phase
despite the enormously heightened risk of grand mal seizures
(560). No effect on spike localization and seizure symptomatol-
ogy was noted with acute withdrawal of antiepileptics (561).
There are certain mysterious day-to-day changes in the
degree of a patient’s epileptic activity. It may require a detec-
tive’s acumen to obtain some insight into certain cycles of
epileptogenicity in a given patient. In epileptic patients with
improvement of seizure frequency after initiation of antiepilep-
tic medication, the interpretation of repeat tracings requires
particular caution and restraint because the congruence of clin-
ical improvement and the amount of paroxysmal EEG activity
is quite limited (562).
Repetition of EEG recordings in a patient who stubbornly
refuses to show the expected paroxysmal discharges will gradu-
ally lead to success, but this philosophy is neither elegant nor
economical. The EEG obtained in the laboratory is a small win-
Figure 26.36 Schematic view of the paroxysmal arousal response over
frontal midline and the physiologic arousal responses in sleep over ver-
tex (K complex, vertex waves). Individual predisposition for paroxys-
mal responses might be based on a local biochemical defect. (From
Niedermeyer E. Generalisierte Epilepsien. EEG Labor. 1979;1:119–131.)
dow in the patient’s daily EEG activity, but this limited record-
ing period can be exploited with the crafty use of numerous
channels for better localization and an arsenal of activating
methods. This window can be enlarged by prolonged recording
in the laboratory over, say, several hours.
The pragmatist may ask for a numerically precise answer to
the following questions: “How useful is the EEG in epileptic
seizure disorder? How many false-positive or false-negative
EEG findings occur in the EEG evaluation for presumed
seizures?” Unfortunately, there are no simple answers and this is
controversial. A practice parameter for afebrile seizures in chil-
dren recommends an EEG for a single seizure. This has been
criticized because we do not usually treat a single seizure but it
certainly may give information about the seizure type or
epilepsy syndrome that is helpful in patient management. Table
26.9 demonstrates the degree of usefulness of EEG findings in
various forms of epileptic seizure disorder.
The value of EEG is determined by intangibles such as the
technical quality of the laboratory technicians and the caliber of
the electroencephalographer; on his or her experience and
effort hinge the thoroughness and trustfulness of the EEG
interpretation. For such obvious reasons, the reader will look in
vain for easy answers; there are none. There are no shortcuts on
the thorny path to expertise in EEG.
 Chapter 26 ■ Seizures and Epilepsy in Infants to Adolescents 527

Tabl e 2 6 . 9

Usefulness of EEG in Various Epileptic Conditions

Overall
Usefulness False Negatives False Positives Comments
Neonatal convulsions Normal in benign Most valuable in differentiation
forms of severe and benign forms
West syndrome May occur (normal Hypsarrhythmic pattern practically
(infantile spasms) but very high voltage) diagnostic
Febrile convulsions Normal EEG the Abnormal EGG suggestive Value of EEG lies in predominantly
rule! of febrile seizures normal findings
Lennox–Gastaut Almost negligible (Difficult differential A diagnosis grossly based upon
diagnosis) EEG; sleep record helpful
Primary generalized Almost negligible (Seizure-free relatives, Often sleep record needed for
epilepsy (in general) other conditions) EEG evidence, photic stimulation
helpful
Pure petit mal No petit mal absence Extremely reliable correlation
absences without spike waves
Juvenile myoclonic ? Sleep record, photic simulation
epilepsy helpful
Benign rolandic Almost negligible (Seizure free despite Often sleep record needed for
epilepsy predisposition) EEG evidence
Benign occipital lobe
epilepsy
Temporal lobe (Repeat records Sleep record a necessity, EEG
epilepsy often needed) diagnosis may require patience
Frontal lobe epilepsy Difficult EEG diagnosis
Epilepsy of motor 0– May be ? A weak spot in the EEG diagnosis
cortex
Other focal epilepsies 0– May or may not be reliable
Epilepsies due to ? Value of EEG depends on
unusual triggering imaginative use of activation
factors
Alcohol-withdrawal Normal EEG the Abnormal record suggests Value of EEG lies in predominantly
epilepsy rule! (mostly of different genesis of normal findings
low voltage) epilepsy
Status epilepticus / Negligible May indicate severity of status—
convulsive helps differential grand
mal/tonic/myoclonic
Status epilepticus Negligible Excellent in detection of absence
nonconvulsive status, good in detection of
psychomotor status
Psychogenic seizures Normal EEG Abnormal records not EEG useful in spite of its
(pseudoseizures) expected to be the uncommon in patients limitations (Video EEG
rule but there are with psychogenic seizures monitoring may be needed)
many exceptions
528 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
REFERENCES
1. Juul-Jensen P, Foldspang A. Natural history of epileptic seizures.
Epilepsy (New York). 1983;24:297–312.
2. Gumnit RJ. The Epilepsy Handbook. New York: Raven Press; 1984.
3. Annegers JF. The epidemiology of epilepsy. In: Wyllie E, ed. The
Treatment of Epilepsy. Philadelphia, PA: Lea and Febiger; 1993:
157–164.
4. Greulich W, Gerber U. Epidemiologie der Epilepsien. In: Froescher
W, Vassella F, eds. Die Epilepsien. Berlin: De Gruyter; 1994:43–55.
5. Hauser WA. Incidence and prevalence. In: Engel J Jr, Pedley T, eds.
Epilepsy. A Comprehensive Textbook (3 vol.). Vol 1. Philadelphia,
PA: Lippincott-Raven; 1998:47–57.
6. IFSECN. Preliminary proposal for an EEG terminology by the termi-
nology committee of the International Federation for Electroen-
cephalography and Clinical Neurophysiology. Electroencephalogr
Clin Neurophysiol. 1961;13:646–650.
7. IFSECN. Proposal for an EEG terminology committee of the
International Federation for Electroencephalography and Clinical
Neurophysiology. Electroencephalogr Clin Neurophysiol. 1966;30:
306–310.
8. IFSECN. A glossary of terms commonly used by clinical electroen-
cephalographers. Electroencephalogr Clin Neurophysiol. 1974;37:
538–548.
9. Noachtar S, Binnie C, Ebersole J, et al. A glossary of terms most
commonly used by clinical electroencephalographers and pro-
posal for the report form for the EEG findings. Recommendations
for the Practice of Clinical Neurophysiology: guidelines of the
International Federation of Clinical Physiology. Electroencephalogr
Clin Neurophysiol Suppl. 1999;52:21–41.
10. Dutertre F. Catalogue of the main EEG patterns. In: Remond A,
ed-in-chief. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 11A. Amsterdam: Elsevier; 1977:40–79.
11. Kiloh LG, McComas AJ, Osselton JW. Clinical Electroencephalo-
graphy. 3rd ed. London: Butterworth; 1972.
12. Roger H, Cornil L, Paillas JE. Les Épilepsies. Paris: Flammarion;
1949.
13. Gotman J. Could interictal epileptic spikes actually be postictal
epileptic spikes? Electroencephalogr Clin Neurophysiol. 1984;58:9P.
Abstract.
14. Gotman J, Marciani MG. Electroencephalographic spiking activ-
ity, drug levels and seizure occurrence in epileptic patients. Ann
Neurol. 1985;17:597–603.
15. Pollen DA, Reid KH, Perol P. Micro-electrode studies of experimen-
tal wave and spike in the cat. Electroencephalogr Clin Neurophysiol.
1964;17:57–67.
16. Matsuo F, Knott JR. Focal positive spikes in electroencephalogra-
phy. Electroencephalogr Clin Neurophysiol. 1977;42:15–25.
17. Maulsby RL. Some guidelines for assessment of spikes and sharp
waves in EEG tracings. Am J EEG Technol. 1971;11:3–16.
18. Binnie CD, Polkey CE, Spencer S. Positive spikes over a cortical lacer-
ation. Electroencephalogr Clin Neurophysiol. 1989;72:106P. Abstract.
19. Gregory DL, Wong PK. Topographical analysis of the centrotempo-
ral discharges in benign rolandic epilepsy of childhood. Epilepsia
(New York). 1984;25:705–711.
20. Wong PKH, Gregory D. Dipole fields in rolandic discharges. Am J
EEG Technol. 1988;28:243–250.
21. Niedermeyer E. Clinical relevance of EEG signals in epilepsies. In:
Wada JA, Ellingson RJ, eds. Handbook of Electroencephalography
and Clinical Neurophysiology. Revised series, Vol 4. Amsterdam:
Elsevier; 1990:237–261.
22. Shewmon DA, Erwin RJ. The effect of focal interictal spikes on
perception and reaction time. I. General considerations.
Electroencephalogr Clin Neurophysiol. 1988;69:319–337.
23. Shewmon DA, Erwin RJ. The effect of focal interictal spikes on
perception and reaction time. II. Neuroanatomic specificity.
Electroencephalogr Clin Neurophysiol. 1988;69:338–352.
24. Shewmon DA, Erwin RJ. Focal spike-induced cerebral dysfunction
is related to the after-coming slow wave. Ann Neurol. 1988;23:
131–137.
25. Gambardella A, Aguglia U, Oliveri RI, et al. Negative myoclonic
status due to antiepileptic drug tapering: report of three cases.
Epilepsia. 1997;38:819–823.
26. Werhahn KJ, Noachtar S. Epileptic negative myoclonus. In: Lueders
HO, Noachtar S, eds. Epileptic Seizures: Pathophysiology and Clinical
Semiology. New York: Churchill Livingstone; 2000:473–483.
27. Fisch BJ. Interictal epileptiform activity: diagnostic and behavioral
implications. J Clin Neurophysiol. 2003;20:155–162.
28. Warach S, Ives JR, Schlaug G, et al. EEG-triggered echo-planar func-
tional MRI in epilepsy. Neurology. 1996;47:89–93.
29. Aarts JHP, Binnie CD, Smit AM, et al. Selective cognitive impair-
ment during focal and generalized epileptiform EEG activity. Brain.
1984;107:293–308.
30. Hughes JR. The significance of interictal spike discharge: a review.
Clin Electroencephalogr. 1997;28:60. Abstract.
31. Jasper HH. Epilepsy. In: Penfield W, Erickson TC, eds. Epilepsy and
Cerebral Localization. Springfield, IL: Charles C Thomas; 1941:
380–454.
32. Cukier F, André N, Monod N, et al. EEG contribution to diagnosing
intraventricular haemorrhage in premature infants. Electroen-
cephalogr Clin Neurophysiol. 1974;36:840. Abstract.
33. Werner SS, Stockard JE, Bickford RG. Atlas of Neonatal
Electroencephalography. New York: Raven Press; 1977.
34. Zobniw AM, Yaworth S, Niedermeyer E. Depth electroencephalog-
raphy. J Electrophysiol Technol. 1975;2:215–240.
35. Ribeiro M, Niedermyer E, Hertz S. Slow sharp waves. Clin
Electroencephalogr. 1999;30:114–117.
36. Chatrian GE, Shaw CM, Leffman H. The significance of periodic
lateralized epileptiform discharges in EEG: an electrographic, clin-
ical and pathological study. Electroencephalogr Clin Neurophysiol.
1964;17:177–193.
37. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 2. 2nd ed.
Cambridge, MA: Addison-Wesley; 1952.
38. Gibbs FA, Davis H, Lennox WG. The electroencephalogram in
epilepsy and in conditions of impaired consciousness. Arch Neurol
Psychiatry (Chicago). 1935;34:1133–1148.
39. Jasper HH, Kershman J. Classification of the EEG in epilepsy.
Electroencephalogr Clin Neurophysiol. 1949;2(suppl 2):123–131.
40. Gastaut H, Roger J, Soulayrol R, et al. Childhood epileptic
encephalopathy with diffuse slow spike-waves (otherwise known as
“petit mal variant”) or Lennox syndrome. Epilepsia (Amsterdam).
1966;7:139–179.
41. Brenner RP, Atkinson R. Generalized paroxysmal fast activity: elec-
troencephalographic and clinical features. Ann Neurol. 1982;11:
386–390.
42. Chayasirisobhon S, Cullis P, Sack R, et al. Grand mal discharge.
Clin Electroencephalogr. 1984;15:155–158.
43. Rodin E, Smid N, Mason K. The grand mal pattern of Gibbs,
Gibbs and Lennox. Electroencephalogr Clin Neurophysiol. 1976;40:
401–406.
44. Niedermeyer E. The Lennox–Gastaut syndrome: a severe type of
childhood epilepsy. Dtsch Z Nervenheilk. 1969;195:263–282.
 Chapter 26
45. Niedermeyer E. The Generalized Epilepsies. Springfield, IL: Charles
C Thomas; 1972.
46. Niedermeyer E. Compendium of the Epilepsies. Springfield, IL:
Charles C Thomas; 1974.
47. Niedermeyer E, Walker AE, Burton C. The slow spike-wave com-
plex as a correlate of frontal and fronto-temporal posttraumatic
epilepsy. Eur Neurol (Basel). 1970;3:330–346.
48. Niedermeyer E. Widersprüchlich beurteilte EEG-Tätigkeiten und
ihre Bedeutung. EEG-Labor. 1988a;10:3–7.
49. Niedermeyer E. The electroencephalogram in the differential diag-
nosis of the Lennox–Gastaut syndrome. In: Niedermeyer E, Degen
R, eds. The Lennox–Gastaut Syndrome. New York: Alan R. Liss; 1988:
177–220.
50. Arroyo S, Lesser RP, Fisher RS, et al. Clinical and electroencephalo-
graphic evidence for sites of origin of seizures with diffuse elec-
trodecremental pattern. Epilepsia. 1994;35:974–987.
51. RamachandraNair R, Ochi A, Imai K, et al. Epileptic spasms in
older pediatric patients: MEG and ictal high frequency oscillations
suggest focal onset seizures in a subset of epileptic spasms. Epilepsy
Research. 2008;78:216–224.
52. Silverman D. Clinical correlates of the spike-wave complexes.
Electroencephalogr Clin Neurophysiol. 1954;6:663–669.
53. Clark EC, Knott JR. Paroxysmal wave and spike activity and diag-
nostic subclassification. Electroencephalogr Clin Neurophysiol. 1955;
7:161–164.
54. Lundervold A, Henriksen GF, Fegersten L. The spike wave com-
plex. A clinical correlation. Electroencephalogr Clin Neurophysiol.
1959;11:13–22.
55. Gastaut H, Hunter J. An experimental study of the mechanism of
photic activation in idiopathic epilepsy. Electroencephalogr Clin
Neurophysiol. 1950;2:263–287.
56. Cohn R. Spike-dome complex in the human electroencephalo-
gram. Arch Neurol Psychiatry (Chicago). 1954;71:699–706.
57. Weir B. The morphology of the spike-wave complex. Electroen-
cephalogr Clin Neurophysiol. 1965;19:284–290.
58. Jung R, Toennies F. Über Entstehung and Erhaltung von
Krampfentladungen. Die Vorgänge im Reizort und die Krampf-
fähigkeit des Gehirns. Arch Psychiatr Nervenkr. 1950;1985:701–735.
59. Pollen DA. Intracellular studies of cortical neurons during thala-
mic induced wave and spike. Electroencephalogr Clin Neurophysiol.
1964;17:398–406.
60. Weir B, Sie PG. Extracellular unit activity in cat during the spike-
wave complex. Epilepsia (Amsterdam). 1966;7:30–43.
61. Halász P. The generalized epileptic spike-wave mechanism and the
sleep-wakefulness system. Acta Physiol Acad Scient Hung. 1972;
42:293–314.
62. Niedermeyer E. Immediate transition from a petit mal absence
into a grand mal seizure. Eur Neurol (Basel). 1976;14:11–16.
63. Dondey M. Transverse topographical analysis of petit mal dis-
charges: diagnostical and pathogenic implications. Electroen-
cephalogr Clin Neurophysiol. 1983;55:361–371.
64. Rodin E, Ancheta O. Cerebral electrical fields during petit mal
absences. Electroencephalogr Clin Neurophysiol. 1987;66:457–466.
65. Gotman J. Interhemispheric relations during bilateral spike-and-
wave activity. Epilepsia (New York). 1981;22:453–466.
66. Lemieux JF, Blume WT. Topographical evolution of the spike-wave
complexes. Electroencephalogr Clin Neurophysiol. 1983;56:30P.
Abstract.
67. Lüders H, Daube J, Johnson R, et al. Computer analysis of gener-
alized spike-and-wave complexes. Epilepsia (New York). 1980;21:
183. Abstract.
■ Seizures and Epilepsy in Infants to Adolescents 529
68. Baddeley AD. Working Memory. Oxford: Oxford University Press;
1986.
69. Fuster JM. Memory and planning. In: Jasper HH, Riggio S, Patricia
Goldman-Rakic PS, eds. Epilepsy and the Functional Anatomy of
the Frontal Lobe. New York: Raven Press; 1995:9–19.
70. Fuster JM. Memory in the Cerebral Cortex. Cambridge, MA: MIT
Press; 1996.
71. Niedermeyer E. Frontal lobe functions and dysfunctions.
Presented at Symposium; 1997; Lima (Peru).
72. Niedermeyer E. Primary (idiopathic) generalized epilepsy. Clin
Electroencephalogr. 1996;27:1–21.
73. Walker A, Morello G. Experimental petit mal. Trans Am Neurol
Assoc. 1967;97:57–61.
74. Hunter J, Jasper HH. Effects of thalamic stimulation in unanes-
thetized animals. Electroencephalogr Clin Neurophysiol. 1949;1:
303–324.
75. Sperling MR, Skolnick BE. Cerebral blood flow during spikewave
discharges. Epilepsia. 1995;36:156–163.
76. Gianelli M, Cantello R, Civardi C, et al. Idiopathic generalized
epilepsy: magnetic stimulation of motor cortex time-locked and
unlocked to 3-Hz spike-and-wave discharges. Epilepsia. 1994;35:
53–60.
77. Gibbs FA, Gibbs EL, Lennox WG. The influence of the blood sugar
level on the wave and spike formation in petit mal epilepsy. Arch
Neurol Psychiatry (Chicago). 1939;47:1111–1116.
78. Lennox WG. Epilepsy and Related Disorders. Boston, MA: Little,
Brown; 1960.
79. Gastaut H. Clinical and electroencephalographic correlates of gen-
eralized spike and wave bursts occurring spontaneously in man.
Epilepsia (Amsterdam). 1968;9:179–184.
80. Silverman D. Phantom spike-wave and the fourteen and six per
second positive spike pattern: a consideration of their relationship.
Electroencephalogr Clin Neurophysiol. 1967;23:203–217.
81. Marshall C. Some clinical correlates of the wave and spike phan-
tom. Electroencephalogr Clin Neurophysiol. 1955;7:633–636.
82. Thomas JE. A rare electroencephalographic pattern: the six-per
second spike and wave discharge. Neurology (Minneapolis). 1957;7:
438–442.
83. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 3.
Reading, MA: Addison-Wesley; 1964.
84. Hughes JR, Schlagenhauff RE, Magoss M. Electro-clinical correla-
tions in the six per second spike and wave complex. Electroen-
cephalogr Clin Neurophysiol. 1965;18:71–77.
85. Thomas JE, Klass DW. Six-per-second spike and wave pattern in
the electroencephalogram: a reappraisal of clinical significance.
Neurology (Minneapolis). 1968;18:587–593.
86. Small JG. Small sharp spikes in a psychiatric population. Arch Gen
Psychiatry. 1970;22:277–284.
87. Hecker A, Kocher R, Ladewig D, et al. Das Minatur-Spike-Wave-
Muster. Das EEG Labor. 1979;1:51–56.
88. Hughes JR. Two forms of the 6/sec spike and wave complex.
Electroencephalogr Clin Neurophysiol. 1980;48:535–550.
89. Hughes JR. Changes in the annual incidence of the 6/sec spike and
wave pattern. Clin Electroencephalogr. 1991;22:71–74.
90. Kocher R, Scollo-Lavizzari G, Ladewig D. Miniature-spikewave-
Muster: Elektroencephalographisches Korrelat in der Abstinenzphase
bei Medikamentenabhängigkeit. Z EEG-EMG. 1975;6:78–82.
91. Tharp BR. The 6-per-second spike and wave complex. Arch Neurol
(Chicago). 1966;15:533–537.
92. Westmoreland BF. Benign EEG variants and patterns of uncertain
clinical significance. In: Daly DD, Pedley TA, eds. Current Practice
530 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
of Clinical Electroencephalography. 2nd ed. New York: Raven
Press; 1990:243–252.
93. Klass DW, Westmoreland BF. Nonepileptogenic epileptiform
electroencephalographic activity. Ann Neurol. 1985;18:627–635.
94. Niedermeyer E. Controversial EEG patterns and their signifi-
cance. Editorial. Am J EEG Technol. 1987;27:129–132.
95. Santoshkumar B, Chong JJR, Blume WT, et al. Prevalence of benign
epileptiform variants. Clin Neurophysiol. 2009;120:856–861.
96. Alvarez N, Lombroso CT, Medina C, et al. Paroxysmal spike and
wave activity in drowsiness in young children: its relationship to
febrile convulsions. Electroencephalogr Clin Neurophysiol. 1983;
56:406–413.
97. Koshino Y, Niedermeyer E. The clinical significance of small
sharp spikes in the electroencephalogram. Clin Electroencephalogr.
1975;6:131–140.
98. Westmoreland BF, Reiher J, Klass DW. Recording small sharp
spikes with depth electroencephalography. Epilepsia (New York).
1979;20:599–612.
99. Small JG, Small IF, Milstein V, et al. Familial associations with
EEG variants in manic depressive disease. Arch Gen Psychiatry.
1975;32:43–48.
100. Reiher J, Klass DW. Two common EEG patterns of doubtful clin-
ical significance. Med Clin North Am. 1968;52:933–940.
101. Lebel M, Reiher J, Klass D. Small sharp spikes (SSS). Reassessment
of electroencephalographic and clinical significance. Electroen-
cephalogr Clin Neurophysiol. 1977;43:463. Abstract.
102. White JC, Langston JW, Pedley TA. Benign epileptiform tran-
sients of sleep: clarification of the small sharp spike controversy.
Neurology. 1977;27:1061–1068.
103. Gutrecht JA. Clinical implications of benign epileptiform transients
of sleep. Electroencephalogr Clin Neurophysiol. 1989;72:486–490.
104. Low MD, Henrikson K, Lamont D, et al. The relationship of small
sharp spikes to seizures. Electroencephalogr Clin Neurophysiol. 1984;
58:11P. Abstract.
105. Hughes JR, Gruener GT. Small sharp spikes revisited: further
data on this controversial pattern. Electroencephalogr Clin Neuro-
physiol. 1984;15:208–213.
106. Saito F, Fukushima Y, Kubota S. Small sharp spikes: possible rela-
tionship to epilepsy. Clin Electroencephalogr. 1987;18:144–119.
107. Molaie M, Santana HB, Cavanaugh WA. Benign epileptiform
transients of sleep (BETS) revisited. Electroencephalogr Clin Neuro-
physiol. 1989;72:19P. Abstract.
108. Gibbs CJ Jr, Gajdusek DC, Asher DM, et al. Creutzfeldt–Jakob
disease (spongiform encephalopathy): transmission to the chim-
panzee. Science. 1968;161:388–389.
109. Lairy GC, Harrison A, Leger EM. Foyers EEG biocciptaux asyn-
chrones de pointes chez l’enfant mal voyant et aveugle d’âge sco-
laire. Rev Neurol (Paris). 1964;111:351–353.
110. Cannon WB, Rosenblueth A. The Supersensitivity of Denervated
Structures: A Law of Denervation. New York: Macmillan; 1949.
111. Lombroso CT, Schwarts IH, Clark DM, et al. Ctenoids in healthy
youth. Controlled study of 14 and 6-per-second positive spiking.
Neurology (Minneapolis). 1966;16:1152–1158.
112. Hughes JR. The 14 and 7 per second positive spikes—a reappraisal
following a frequency count. Electroencephalogr Clin Neurophysiol.
1960;12:495–496.
113. Hughes JR. A review of the positive spike phenomenon. In: Wilson
WP, ed. Application of Electroencephalography in Psychiatry.
Durham: Duke University Press; 1965:54–101.
114. Gibbs EL, Gibbs FA. Electroencephalographic evidence of thala-
mic and hypothalamic epilepsy. Neurology (Minneapolis). 1951;1:
136–144.
115. Gibbs FA, Gibbs EL. Fourteen and six per second positive spikes.
Electroencephalogr Clin Neurophysiol. 1963;15:353–358.
116. Henry EE. Positive spike discharges in the EEG and behavior
abnormality. In: Glaser GE, ed. EEG and Behavior. New York:
Basic Books; 1963:315–344.
117. Hughes JR, Gianturco D, Stein W. Electro-clinical correlations in
the positive spike phenomenon. Electroencephalogr Clin Neuro-
physiol. 1961;13:599.
118. Kellaway P. Neurologic status of patients with hypsarrhythmia.
In: Gibbs FA, ed. Molecules and Health. Philadelphia, PA: JB
Lippincott; 1959:134–149.
119. Millen FJ, White B. Fourteen and six per second positive spike
activity in children. Neurology (Minneapolis). 1954;4:541–549.
120. Niedermeyer E, Knott JR. Über die Bedeutung der 14 and 6/sec-
positiven Spitzen im EEG. Arch Psychiatr Nervenkr. 1961;202:
266–280.
121. Poser CM, Ziegler DK. Clinical significance of 14 and 6 per sec-
ond positive spike complexes. Neurology (Minneapolis). 1958;8:
903–912.
122. Refsum S, Presthus J, Skulstad AA, et al. Clinical correlates
of the 14 and 6 per second positive spikes. An electroen-
cephalographic and clinical study. Acta Psychiatr Scand. 1960;35:
330–344.
123. Schwade ED, Geiger SG. Matricide with electroencephalographic
evidence of thalamic or hypothalamic disorder. Dis Nerv Syst.
1953;14:18–20.
124. Shimoda Y. The clinical and electroencephalographic study of the
primary diencephalic epilepsy or epilepsy of brain stem. Acta
Neurovegetat (Vienna). 1961;23:181–191.
125. Stephenson WA. Intracranial neoplasm associated with fourteen
and six per second positive spikes. Neurology (Minneapolis). 1951;
1:372–376.
126. Walter RD, Colbert EG, Koegler RR, et al. A controlled study of the
fourteen-and-six-per-second EEG pattern. Arch Gen Psychiatry.
1960;2:559–566.
127. Long MT, Johnson LC. Fourteen and six positive spikes in
nonpatient population. Neurology (Minneapolis). 1967;17:316.
Abstract.
128. Little SC. A general analysis of the fourteen and six per second
dysrhythmia. Proceedings of the Sixth International Congress of
Electroencephalography and Clinical Neurophysiology. Vienna:
Wiener Med. Akad; 1965:313–315.
129. Beydoun A, Drury I. Unilateral 14 and 6 Hz positive bursts.
Electroencephalogr Clin Neurophysiol. 1992;82:310–312.
130. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 4. Menlo
Park, CA: Addison-Wesley; 1978.
131. Okada S, Kato T, Miyashita K, et al. 14 and 6 Hz positive spikes:
relationship with a negative spiky phase in the frontal area.
Electroencephalogr Clin Neurophysiol. 1990;75:S108. Abstract.
132. Niedermeyer E, Ray CD, Walker AE. Depth studies in a patient
with fourteen and six per second positive spikes. Electroen-
cephalogr Clin Neurophysiol. 1967;22:86–89.
133. Gibbs FA. An epidemiological note? Clin Electroencephalogr. 1990;
21:112–118.
134. Silverman D. Fourteen and six per second positive spike pattern
in a patient with hepatic coma. Electroencephalogr Clin Neuro-
physiol. 1964;16:395–398.
135. Ford RG, Freeman AM. Positive spike bursts in a comatose adult.
Electroencephalogr Clin Neurophysiol. 1982;53:29P. Abstract.
136. Drury I. 14- and 6-Hz positive bursts in diverse encephalopathies
of childhood. Electroencephalogr Clin Neurophysiol. 1989;72:12P.
Abstract.
 Chapter 26
137. Lipman IL, Hughes JR. Rhythmic mid-temporal discharges. An
electroclinical study. Electroencephalogr Clin Neurophysiol. 1969;
27:43–47. 160.
138. Egli M, Hess R, Kuritzke G. Die Bedeutung der “rhythmic mid-
temporal discharges.” Z EEG-EMG. 1978;9:74–85. 161.
139. Hughes JR. The continuous rhythmic mid-temporal discharge.
Clin Electroencephalogr. 2001;32:10–13.
140. Garvin JS. Psychomotor variant pattern. Dis Nerv Syst. 1968;29: 162.
59–76.
141. Boutros NN, Hughes JR, Weiler M. Psychiatric correlates of the
rhythmic mid-temporal discharges and the 6/sec spike and wave 163.
complexes. Biol Psychiatry. 1986;21:94–99.
142. Naquet R, Louard C, Rhodes J, et al. À propos de certaines
décharges paroxystiques du carrefour temporo-pariéto-occipital: 164.
leur activation par l’hypoxie. Rev Neurol (Paris). 1961;105:203–207.
143. Herranz F, Lopez S. Subclinical paroxysmal rhythmic activity.
Electroencephalogr Clin Neurophysiol. 1984;4:419–442. 165.
144. Westmoreland BF, Klass DW. A distinctive rhythmic EEG discharge
of adults. Electroencephalogr Clin Neurophysiol. 1981;51:186–191.
145. Westmoreland BF, Klass DW. Unusual variants of subclinical 166.
rhythmic electrographic discharge of adults (SREDA). Electroen-
cephalogr Clin Neurophysiol. 1997;102:1–4. 167.
146. Miller CR, Westmoreland BF, Klass DW. Subclinical rhythmic
EEG discharge of adults (SREDA): further observations. Am J
EEG Technol. 1985;25:217–224. 168.
147. O’Brien TJ, Sharbrough FW, Westmoreland BF, et al. Subclinical
rhythmic electrographic discharges of adults (SREDA) revisited:
a study using digital EEG analysis. J Clin Neurophysiol. 1998l;15: 169.
493–501.
148. Gaches J. Les activitiés périodiques en EEG. Rev EEG Neuro- 170.
physiol. 1971;1:9–33.
149. Bauer G, Pieber R. Über periodische Komplexe im EEG. Z EEG- 171.
EMG. 1974;5:75–86.
150. Spehlmann R. EEG Primer. Amsterdam: Elsevier; 1981. 172.
151. Rabending G, Radermecker FJ. Subacute sclerosing panen-
cephalitis (SSPE). In: Remond A, ed-in-chief. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol. 15A. 173.
Amsterdam: Elsevier; 1977:28–35.
152. Celesia GG. Pathophysiology of periodic EEG complexes in sub- 174.
acute sclerosing panencephalitis (SSPE). Electroencephalogr Clin
Neurophysiol. 1973;35:293–300.
153. Broughton R, Nelson R, Gloor P, et al. Petit mal epilepsy evolving 175.
to subacute sclerosing panencephalitis. In: Lugaresi E, Pazzaglia P,
Tassinari CA, eds. Evolution and Prognosis of Epilepsies. Bologna:
Gaggi; 1973:63–72. 176.
154. Gökcil Z, Odabasi Z, Aksu A, et al. Acute fulminant SSPE: clini-
cal and EEG features. Clin Electroencephalogr. 1998;29:43–48.
155. Giovanardi Rossi P, Gobbi G, Moschen R, et al. Subacute scleros- 177.
ing panencephalitis-like EEG pattern and periodic myoclonic
seizures: two case reports. Electroencephalogr Clin Neurophysiol.
1982;54:3P. Abstract.
156. Prier S, Benoit C, Cambier J. Bilateral periodic stereo-typed EEG 178.
complexes in multiple sclerosis. Clin Neurophysiol. 1992;22(suppl
1–1992):107s. Abstract.
157. Radermecker FJ, Poser CM. The significance of repetitive 179.
paroxysmal electroencephalographic patterns. Their specificity
in subacute sclerosing leukoencephalitis. World Neurol. 1960;1:
422–433. 180.
158. Cobb WA. The periodic events of subacute sclerosing leucoen-
cephalitis. Electroencephalogr Clin Neurophysiol. 1966;21:278–294.
159. Cobb WA. Depth recording in subacute sclerosing leucoencephali-
tis. In: Kellaway P, Petersén I, eds. Clinical Electroencephalography of
■ Seizures and Epilepsy in Infants to Adolescents 531
Children. New York: Grune & Stratton (Stockholm: Almquist and
Wiksell); 1968:275–286.
Lombroso CT. Remarks on the EEG and movement disorder in
SSPE. Neurology (Minneapolis). 1968;18:69–75.
Bogacz J, Castells C, San Julian J, et al. Nonepidemic progressive
subacute encephalitis (type Van Bogaert). II. Serial EEG abnormal-
ities and deep electrography. Acta Neurol Lat Am. 1959;5:158–183.
Lesse S, Hoefer PFA, Austin JH. The electroencephalogram in dif-
fuse encephalopathies. Arch Neurol Psychiatry (Chicago).
1958;79:359–375.
Storm van Leeuwen W. Electroencephalographical and neuro-
physiological aspects of subacute sclerosing leuco-encephalitis.
Psychiatr Neurol Neurochir (Amsterdam). 1964;67:312–322.
Martinovic Z. Periodic generalized bursts of fast waves in suba-
cute sclerosing panencephalitis. Electroencephalogr Clin Neuro-
physiol. 1986;63:236–238.
Radermecker FJ. Systématique et Électroencéphalographie des
Encéphalites et Encéphalopathies. Amsterdam: Elsevier; 1956
(Electroencephalogr Clin Neurophysiol Suppl 5).
Millar JHD, Coey A. The EEG in necrotizing encephalitis.
Electroencephalogr Clin Neurophysiol. 1959;11:582–585.
Perier O, Parmentier B, Brihaye J, et al. A case of inclusion body
necrotizing encephalitis. In: Van Bogaert L, et al., eds.
Encephalitides. Amsterdam: Elsevier; 1961:235–242.
Carmon A, Behar A, Beller A. Acute necrotizing encephalitis
presenting clinically as a space-occupying lesion. A clinico-
pathological study of six cases. J Neurol Sci. 1965;2:328–343.
Rawls WE, Dyck PJ, Klass DW, et al. Encephalitis associated with
herpes simplex virus. Ann Intern Med. 1966;64:104–115.
Adams JH, Jennett WB. Acute necrotizing encephalitis: a problem
in diagnosis. J Neurol Neurosurg Psychiatry. 1967;30:248–260.
Upton A, Gumpert J. Electroencephalography in diagnosis of
herpes simplex encephalitis. Lancet. 1970;1:650–652.
Gaches J, Arfel G. Certitude et suspicions d’encéphalites herpé-
tiques. Aspects électroencéphalographiques. Encéphale. 1972;61:
510–549.
Elian M. Herpes simplex virus encephalitis and EEG. J
Electrophysiol Technol. 1975;1:161–170.
Kugler J, Marltin JJ, Radermecker RJ, et al. Periodische Komplexe
im EEG bei nekrotisierender Herpes-Enzephalitis. Z EEG-EMG.
1976;7:63–71.
Cobb WA. Evidence on the periodic mechanism in herpes sim-
plex encephalitis. Electroencephalogr Clin Neurophysiol. 1979;46:
345–350.
Gupta PC, Seth P. Periodic complexes in herpes simplex
encephalitis. A clinical and experimental study. Electroencephalogr
Clin Neurophysiol. 1973;35:67–74.
Fineyre R. Periodic discharge of unusually slow character in
neonatal status epilepticus caused by necrotizing encephalitis
(herpes simplex). Electroencephalogr Clin Neurophysiol. 1979;46:80.
Abstract.
Bergey GK, Coyle PK, Krumholz A, et al. Herpes simplex
encephalitis with occipital localization. Arch Neurol (Chicago).
1982;39:312–313.
Creutzfeldt HG. Über eine eigenartige herdförmige Erkrankung
des Zentralnervensystems. Z Gesamte Neurol Psychiatr. 1920;57:1.
(Quoted after Kirschbaum, W.R., 1968.)
Jakob A. Über eigenartige Erkrankungen des Zentralnervensys-
tems mit bemerkenswerten anatomischen Befunden (spastische
Pseudosklerose, Encephalomyelopathie mit disseminierten
Degenerationsbeschwerden). Dtsch Z Nervenheilkd. 1921;70:132.
(Quoted after Kirschbaum, W.R., 1968.)
532 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
181. Kirschbaum WR. Zwei eigenartige Erkrankungen des Zentralner -
vensystems nach Art der spastischen Pseudosklerose (Jakob). Z
Gesamte Neurol Psychiatr. 1924;92:157. (Quoted after Kirschbaum,
W.R., 1968.)
182. Kirschbaum WR. Jakob–Creutzfeldt’s Disease. New York: American
Elsevier; 1968.
183. Jones DP, Nevin S. Rapidly progressive cerebral degeneration
(subacute vascular encephalopathy) with mental disorder, focal
disturbances and myoclonic epilepsy. J Neurol Neurosurg
Psychiatry. 1954;17:148–159.
184. Heidenhain A. Klinische und anatomische Untersuchungen einer
eignartiger organische Erkrankung des Zentralnervensystems im
Praesenium. Z Gesamte Neurol Psychiatr. 1929;118:49–111.
185. Gibbs CJ Jr, Gajdusek DC. Infection as the etiology of spongi-
form encephalopathy (Creutzfeldt–Jakob disease). Science. 1969;
165:1023–1025.
186. Niedermeyer E, Gibbs CB Jr, Marsh R, et al. EEG studies in sub-
acute and degenerative neurological diseases experimentally pro-
duced in ferrets and chimpanzees. Electroencephalogr Clin
Neurophysiol. 1972;33:351–352. Abstract.
187. Drury I, Beydoun A. Evolution of periodic complexes in
Creutzfeldt–Jakob disease. Am J EEG Technol. 1996;36:230–234.
188. Tariska P, Gereby G, Majtenyi K. The clinical significance of peri-
odic lateralized epileptiform discharges (PLEDs). Electroen -
cephalogr Clin Neurophysiol. 1985;61:11P–12P. Abstract.
189. Chiappa K, Young R. The EEG as a definitive diagnostic tool early
in the course of Creutzfeldt–Jakob disease. Electroencephalogr
Clin Neurophysiol. 1978;45:26. Abstract.
190. Chiofalo N, Fuentes A, Galvez S. Serial EEG findings in 27 cases of
Creutzfeldt–Jakob disease. Arch Neurol (Chicago). 1980;37:143–145.
191. Poole EW. Jakob–Creutzfeldt disease. Patients who might have
had this. Electroencephalogr Clin Neurophysiol. 1980;49:89P.
192. Niedermeyer E, Fineyre R, Riley T, et al. Myoclonus and the elec-
troencephalogram. A review. Clin Electroencephalogr. 1979;10:
75–95.
193. Terzano MG, Mancia D, Calzetti S, et al. Diagnostic value of EEG
periodic discharges and cyclic changes in Creutzfeldt–Jakob dis-
ease. Electroencephalogr Clin Neurophysiol. 1981;52:52P. Abstract.
194. Bortone E, Bettoni L, Giorgi C, et al. Reliability of EEG in the
diagnosis of Creutzfeldt–Jakob disease. Electroencephalogr Clin
Neurophysiol. 1994;90:323–330.
195. Aminoff MJ, Greenberg DA, Wiekle DJ. Periodic EEG complexes
in infectious mononucleosis encephalitis. Electroencephalogr Clin
Neurophysiol. 1982;53:28P. Abstract.
196. Markand ON, Daly DD. Pseudoperiodic lateralized paroxysmal
discharges in the electroencephalogram. Neurology (Minneapolis).
1971;21:975–981.
197. Dauben RD, Adams AH. Periodic lateralized epileptiform dis-
charges in EEG. A review with special attention to etiology and
recurrence. Clin Electroencephalogr. 1974;8:116–124.
198. Bauer G, Aichner F, Hengl W. Der diagnostische Wert periodisch
lateralisierter Komplexe im EEG. Z EEG-EMG. 1981;12:135–141.
199. Naquet R, Vigouroux RA. The EEG in relation to pathology in
simple cerebral ischaemia. In: Remond A, ed-in-chief. Handbook
of Electroencephalography and Clinical Neurophysiology. Vol 14A.
Amsterdam: Elsevier; 1972:38–44.
200. De la Paz D, Brenner RP. Bilateral independent periodic lateralized
epileptiform discharges—clinical significance. Electroencephalogr
Clin Neurophysiol. 1982;53:27P. Abstract.
201. Schear HE. Periodic EEG activity. Clin Electroencephalogr. 1984;
15:32–39.
202. Filley CM, Kelly MA, Hwang PALS, et al. An analysis of typical
and “indeterminate” PLEDs. Electroencephalogr Clin Neurophysiol.
1984;58:32P. Abstract.
203. Aldrich MS, Pugh JE. Cerebral bigeminy: alternating periodic
epileptiform discharges. Electroencephalogr Clin Neurophysiol.
1984;58:35P. Abstract.
204. Hughes JR, Taber J, Uppal H. TRI-PLEDs: a case report. Clin
Electroencephalogr. 1998;29:106–108.
205. PeBenito R, Cracco JB. Periodic lateralized epileptiform dis-
charges in children. Ann Neurol. 1979;6:47–50.
206. Andriola M. PLEDS and bi-PLEDS in children. Electroen -
cephalogr Clin Neurophysiol. 1982;53:88P. Abstract.
207. Ritaccio AL, March G. The significance of BIPLEDs in complex
partial status epilepticus. Am J EEG Technol. 1993;33:27–34.
208. Silbert PL, Radhakrishnan K, Sharbrough FW, et al. Ipsilateral
independent periodic lateralized epileptiform discharges.
Electroencephalogr Clin Neurophysiol. 1996;98:223–226.
209. Chabolla DR, Moore JL, Westmoreland BF. Periodic lateralized
epileptiform discharges in multiple sclerosis. Electroencephalogr
Clin Neurophysiol. 1996;98:5–8.
210. Gandelman-Marton R, Rabey JM, Flechter S. Periodic lateralized
epileptiform discharges in multiple sclerosis: a case report. J Clin
Neurophysiol. 2003;20:117–121.
211. Pohlmann-Eden B, Hoch DB, Cochius JI, et al. Periodic lateralized
epileptiform discharges—a critical review. J Clin Neurophysiol.
1996;13:519–530.
212. Westmoreland BF, Frere RC, Klass DW. Periodic epileptiform dis-
charges in the midline. J Clin Neurophysiol. 1997;14:495–498.
213. Butenuth J, Kubicki St. Über die prognostische Bedeutung bes-
timmter Formen der Myoklonien und korrespondierender EEG-
Muster nach Hypoxien. Z EEG-EMG. 1971;2:78–83.
214. Madison DS, Niedermeyer E. Epileptic seizures resulting from acute
cerebral anoxia. J Neurol Neurosurg Psychiatry. 1970;33:381–386.
215. Takahashi M, Kubota F, Nishi Y, et al. Persistent synchronous
periodic discharges caused by anoxic encephalopathy due to car-
diopulmonary arrest. Clin Electroencephalogr. 1993;24:166–172.
216. Thorner MW. Procurement of electroencephalograph tracings in
1000 flying cadets for evaluating the Gibbs technique in relation
to flying ability. USAD School of Aviation Medical Research
Report, No. 7-1; 1942.
217. Gibbs FA, Gibbs EL, Lennox WG. Electroencephalographic clas-
sification of epileptic patients and control subjects. Arch Neurol
Psychiatry. 1943;50:111–128.
218. Harty JE, Gibbs EG, Gibbs FA. Electroencephalographic study of
two hundred and seventy-five candidates for military service.
War Med (Chicago). 1942;2:923–930.
219. Williams D. The nature of transient outbursts in the electroen-
cephalogram of epileptics. Brain. 1944;67:10–37.
220. Buchthal F, Lennox M. The EEG effect of Metrazol and photic
stimulation in 682 normal subjects. Electroencephalogr Clin
Neurophysiol. 1953;5:545–558.
221. Zivin L, Ajmone Marsan C. Incidence and prognostic signifi-
cance of “epileptiform” activity in the EEG of nonepileptic sub-
jects. Brain. 1968;91:751–777.
222. Chatrian GE. Paroxysmal pattern in “normal” subjects. In: Remond
A, ed-in-chief. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 6A. Amsterdam: Elsevier; 1976:114–123.
223. Nekhorocheff MI. La stimulation lumineuse intermittante chez
l’enfant normal. Rev Neurol (Paris). 1950;83:601–602.
224. Kellaway P, Fox BJ. Electroencephalographic diagnosis of cerebral
pathology in infants during sleep. I. Rationale, technique and the
 Chapter 26
characteristics of normal sleep in infants. J Pediatr. 1952;41:
262–287.
225. Brandt S, Brandt H. The electroencephalographic patterns in
young healthy children from 0 to five years of life. Acta Psychiatr
Scand. 1955;30:77–89.
226. Corbin HPF, Bickford RG. Studies of the electroencephalogram of
normal children: comparison of visual and automatic frequency
analysis. Electroencephalogr Clin Neurophysiol. 1955;7:15–28.
227. Trojaborg W. Focal spike discharges in children. Acta Paediatr
Scand Suppl. 1966;168:1–13.
228. Eeg-Olofsson O, Petersén I, Selldén U. The development of the
electroencephalogram in normal children from the age of 1
through 15 years. Paroxysmal activity. Neuropaediatrie. 1971;2:
375–404.
229. Cavazzuti GB, Cappella L, Nalin A. Longitudinal study of epilep-
tiform EEG patterns in normal children. Epilepsia (New York).
1980;21:43–55.
230. Wolf P. The classification of seizures and the epilepsies. In: Porter
RJ, Morselli PL, eds. The Epilepsies. London: Butterworth; 1985:
107–124.
231. Baudewig J, Bittermann HJ, Paulus W, et al. Simultaneous EEG
and functional MRI of epileptic activity: a case report. Clin
Neurophysiol. 2001;112:1196–1200.
232. Commission on Classification and Terminology of the
International League Against Epilepsy. Proposal for revised clinical
and electroencephalographic classification of epileptic seizures.
Epilepsia (New York). 1981;22:489–501.
233. Dreifuss FE. The international classification of seizures and
epilepsies: advantages. In: Wolf P, ed. Epileptic Seizures and
Syndromes. London: Libbey; 1994:9–14.
234. Nordli DR Jr, Bazil CW, Scheuer ML, et al. Recognition and clas-
sification of seizures in infants. Epilepsia. 1997;38:553–560.
235. Berg AT, Berkovic SF, Brodie MJ, et al. Revised terminology and
concepts for organization of seizures and epilepsies: report of the
ILAE Commission on Classification and Terminology, 2005–2009.
Epilepsia. 2010;51:676–685.
236. Gowers WR. Epilepsy and Other Chronic Convulsive Diseases: Their
Causes, Symptoms and Treatment. New York: Dover; 1881
(reprinted 1964) (American Academy of Neurology reprint series).
237. Janz D. Die Epilepsien. Stuttgart: Thieme; 1969.
238. Gastaut H, Broughton R. Epileptic Seizures. Springfield, IL:
Charles C Thomas; 1972.
239. Karbowski K. Epileptische AnfŠlle. Berlin: Springer; 1985.
240. Tabbador K, Balagura S. Acute epidural hematoma following
epileptic seizures. Arch Neurol (Chicago). 1981;38:198–199.
241. Gotman J, Ives JR, Gloor P. Frequency content of EEG and EMG
at seizure onset: possibility of removal of EMG artifact by digital
filtering. Electroencephalogr Clin Neurophysiol. 1981;52:626–639.
242. Gastaut H, Fischer-Williams M. The physiopathology of epilep-
tic seizures. In: Field J, Magoun HW, Hall VE, eds. Handbook of
Physiology. Vol 1. Baltimore, MD: Williams & Wilkins; 1959:
329–364.
243. Zifkin B, Dravet C. Generalized convulsive seizures. In: Engel J Jr,
Pedley T, eds. Epilepsy. A Comprehensive Textbook (3 vol.). Vol 1.
Philadelphia, PA: Lippincott-Raven; 1998:567–577.
244. Korff C, Nordli DR Jr. Do generalized tonic–clonic seizures in
infancy exist? Neurology. 2005;65:1750–1753.
245. Jung R. Hirnelekrische Untersuchungen über den Elektrokrampf.
Die Erregungsabläufe in corticalen und subcorticalen Hirnregio-
nen bei Katze und Hund. Arch Psychiatr Nervenkr. 1949;183:
206–244.
■ Seizures and Epilepsy in Infants to Adolescents 533
246. Rodin EA, Gonzales S, Caldwell D, et al. Photic evoked responses
during induced epileptic seizures. Epilepsia (Amsterdam). 1966;
7:202–214.
247. Petsche H. Epileptischer Anfall und kortikale Neuronenpopula-
tion. Wien Z Nervenheilkd. 1968;26:45–55.
248. Gastaut H, Naquet R, Fischer-Williams M. The pathophysiology
of grand mal seizures generalized from the start. J Nerv Ment Dis.
1958;127:21–33.
249. Temkin O. The Falling Sickness. 2nd ed. Baltimore, MD: Johns
Hopkins Press; 1971.
250. Esquirol. Quoted in Temkin, O. 1971; 1815.
251. Calmei L-F. De l’épilepsie, étudiée sous la rapport de son siège et de
son influence sur la production de l’aliénation mentale. Paris:
Thése de Paris; 1824.
252. Bamberger P, Matthes A. Anfalle im Kindesalter. Basel: S. Karger;
1959.
253. Sadleir LG, Scheffer IE, Smith S, et al. Automatisms in absence
seizures in children with idiopathic generalized epilepsy. Arch
Neurol. 2009;66:729–734.
254. Sadleir LG, Scheffer IE, Smith S, et al. EEG features of absence
seizures in idiopathic generalized epilepsy: impact of syndrome,
age, and state. Epilepsia. 2009;50:1572–1578.
255. Niedermeyer E. Frontal lobe epilepsy: the next frontier. Clin
Electroencephalogr. 1998;29:163–169.
256. Niedermeyer E. Electrophysiology of the frontal lobe. Clin
Electroencephalogr. 2003;34:5–12.
257. Pavone A, Niedermeyer E. Absence seizures and the frontal lobe.
Clin Electroencephalogr. 2000;31:153–156.
258. Janz D. Die klinische Stellung der Pyknolepsie. Dtsch Med
Wochenschr. 1955;80:1392–1400.
259. Jung R. Uber vegetative Reaktionen und Hemmungswirkungen
von Sinnesreizen im kleinen epileptischen Anfall. Nervenarzt.
1939;12:169–185.
260. Mirsky AF, Van Buren JM. On the nature of the “absence” in cen-
trencephalic epilepsy. A study of some behavioral, electroen-
cephalographic and autonomic factors. Electroencephalogr Clin
Neurophysiol. 1965;18:334–348.
261. Cornil L, Gastaut H, Corriol J. Appréciation du degré de con-
science au cours des paroxysmes épileptiques “Petit Mal.” Rev
Neurol (Paris). 1951;84:149–151.
262. Fischgold H. La conscience et ses modifications. Systemes de
references en E.E.G. clinique. In: Proceedings of the First
International Congress of Neurological Sciences. Reports and
discussions. Vol 2. Brussels: Les editions Acta Med Belg; 1957:
181–213.
263. Lehmann HJ. Praparoxysmale Weckreaktionen bei pyknoleptis-
chen Absenzen. Arch Psychiatr Nervenkr. 1963;204:417–426.
264. Schwab RS. Methods of measuring consciousness in attacks
of petit mal. Arch Neurol Psychiatry (Chicago). 1939;41:215–217.
265. Shimazono Y, Hirai T, Okuma T, et al. Disturbance of conscious-
ness in petit mal epilepsy. Epilepsia (Boston). 1953;2:49–55.
266. Oller-Daurella L, Oller-Ferrer-Vidal L. Atlas de Crisis Epilepticas.
Spain, Barcelona: Geigy Division Farmaceut; 1977.
267. Mirsky AF, Tecce JJ. The analysis of visual potentials during spike
and wave activity. Epilepsia (Amsterdam). 1968;9:211–220.
268. Penry JK, Porter RJ, Dreifuss RE. Simultaneous recording of
absence seizures with video tape and electroencephalography. A
study of 374 seizures in 48 patients. Brain. 1975;98:427–440.
269. Sadleir LG, Farrell K, Smith S, et al. Electroclinical features of
absence seizures in childhood absence epilepsy. Neurology. 2006;
67:413–418.
534 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
270. Chaix Y, Daquin G, Monteiro F, et al. Absence epilepsy with onset
before age three years: a heterogeneous and often severe condi-
tion. Epilepsia. 2003;44:944–949.
271. Leary LD, Wang D, Nordli DR Jr, et al. Seizure characterization
and electroencephalographic features in Glut-1 deficiency.
Epilepsia. 2003;44:701–707.
272. Suls A, Mullen SA, Weber YG, et al. Early-onset absence epilepsy
caused by mutations in the glucose transporter GLUT1. Ann
Neurol. 2009;66:415–419.
273. Von Moers A, Brockmann K, Wang D, et al. EEG features of Glut-
1 deficiency syndrome. Epilepsia. 2002;43:941–945.
274. Rütti W. Absenzen-Epilepsie im Erwachsenenalter. Schweiz Med
Wochenschr. 1982;112:434–441.
275. Rütti W, Karbowski K. Absence epilepsy in adults. Electroen-
cephalogr Clin Neurophysiol. 1983;55:18P. Abstract.
276. Janz D, Beck-Mannagetta G, Spröder B, et al. Childhood absence
epilepsy (pyknolepsy) and juvenile absence epilepsy: one or two
syndromes? In: Wolf P, ed. Epileptic Seizures and Syndromes.
London: Libbey; 1994:115–126.
277. Trinka E, Baumgartner S, Unterberger I, et al. Long-term prog-
nosis for childhood and juvenile absence epilepsy. J Neurol. 2004;
251:1235–1241.
278. DeMarco P. Reflex petit mal absence? Clin Electroencephalogr.
1990;21:74–76.
279. Lennox WG. The petit mal epilepsies; their treatment with trid-
ione. JAMA. 1945;129:1069–1074.
280. Watson CW, Denny-Brown D. Studies of the mechanism of
stimulus-sensitive myoclonus in man. Electroencephalogr Clin
Neurophysiol. 1955;7:341–356.
281. Halliday AM. The electrophysiological study of myoclonus.
Brain. 1967;90:241–284.
282. Halliday AM. Cerebral evoked potentials in familial progressive
myoclonic epilepsy. J R Coll Physicians Lond. 1967;1:123–134.
283. Halliday AM. The clinical incidence of myoclonus. In: Williams
D, ed. Modern Trends in Neurology. London: Butterworth; 1967:
69–105.
284. Halliday AM. The neurophysiology of myoclonus—a reappraisal.
In: Charlton MH, ed. Myoclonic Seizures. Amsterdam: Excerpta
Medica; 1975:1–29.
285. Shibasaki H, Kuroiwa Y. Electroencephalographic correlates of
myoclonus. Electroencephalogr Clin Neurophysiol. 1975;39:455–463.
286. Chadwick D, Mallett M, Harris R, et al. Clinical biochemical and
physiological features distinguishing myoclonus response to 5-
hydroxytryptophan, tryptophan with a monoamino oxidase
inhibitor and clonazepam. Brain. 1977;100:455–487.
287. Hallett M, Chadwick D, Marsden CD. Ballistic movement overflow
myoclonus: a form of essential myoclonus. Brain. 1977;100:299–312.
288. Kelly JJ Jr, Sharbrough FW, Daube JR. A clinical and electrophys-
iological evaluation of myoclonus. Neurology (New York). 1981;
31:581–589.
289. Hallett M. Myoclonus: relation to epilepsy. Epilepsia (New York).
1985;26(suppl 1):567–577.
290. Gastaut H. Sémiologie des myoclonies et nosologie analytique
des syndromes myocloniques. Rev Neurol (Paris). 1968;119:1–30.
291. Oller-Daurella L, Oller-Ferrer-Fidal L. Altas de Crisis Epilepticas.
2nd ed. Basel: Geigy; 1981.
292. Weingarten K. Die Myoklonischen Syndrome. Vienna: Maudrich;
1957.
293. Aigner BR, Mulder DW. Myoclonus: clinical significance
and approach to classifications. Arch Neurol (Chicago). 1960;2:
600–615.
294. Bauer G. Myoklonien: Erscheinungsbild, Pathophysiologie und
klinische Bedeutung. Wien Klin Wochenschr. 1974;124:577–581.
295. Kinsbourne M, Rosenfield DB. Nonprogressive myoclonus. In:
Charlton MH, ed. Myoclonic Seizures. Amsterdam: Excerpta
Medica; 1975:30–59.
296. Sulibhavi DC, Schneck L. Myoclonus epilepsy in progressive dis-
ease. In: Charlton MH, ed. Myoclonic Seizures. Amsterdam:
Excerpta Medica; 1975:60–76.
297. Farrell DF, Swanson PD. Infectious diseases associated with
myoclonus. In: Charlton MH, ed. Myoclonic Seizures. Amsterdam:
Excerpta Medica; 1975:77–110.
298. Charlton MH. Infantile spasms. In: Charlton MH, ed. Myoclonic
Seizures. Amsterdam: Excerpta Medica; 1975:111–120.
299. Niedermeyer E, Fineyre F, Riley T, et al. Myoclonus and the elec-
troencephalogram. Clin Electroencephalogr. 1979;10:75–95.
300. Aicardi J. Myoclonus as a manifestation of degenerative disorders
of the central nervous system in childhood. Electroencephalogr
Clin Neurophysiol. 1983;56:2P.
301. Tassinari CA. New perspectives in epileptology. In: Japanese
Epilepsy Association, ed. Trends in Modern Epileptology. Tokyo:
Proceedings of the International Public Seminar of Epileptology;
1981:42–59.
302. Aguglia U, Gambardella A, Zappia M, et al. Negative myoclonus
during valproate-related stupor. Neurophysiological evidence of a
cortical non-epileptic origin. Electroencephalogr Clin Neurophysiol.
1995;94:103–108.
303. Baumgartner C, Podreka I, Olbrich A, et al. Epileptic negative
myoclonus. An EEG-single-photon-emission CT study indicating
involvement of premotor cortex. Neurology. 1996;46:753–758.
304. Rubboli G, Parmeggiani L, Tassinari CA. Frontal inhibitory spike
component associated with epileptic negative myoclonus.
Electroencephalogr Clin Neurophysiol. 1995;95:201–205.
305. Bravais LF. Recherches sur les Symptomes et le Traitement de
l’épilepsie Hémiplégique. Paris: Thèse de Paris No. 118; 1827.
306. Fritsch G, Hitzig E. Uber die elekrische Erregbarkeit des
Grosshirns. Arch Anat Physiol. 1870;37:300–332.
307. Ferrier D. Experimental researches in cerebral physiology and
pathology. West Riding Lunatic Asylum Med Rep. 1873;3:30–96.
308. Luciani L. Sulla patogenesi dell’epilessia. Riv Sper Freniatr Med
Leg. 1878;4:617–646.
309. Neufeld MY, Blumen S, Nisipeanu P. Lingual seizures.
Electroencephalogr Clin Neurophysiol. 1988;69:59P. Abstract.
310. Ahuja GK, Tharakan J. Ipsilateral seizures. Epilepsia (New York).
1981;23:555–561.
311. Elger CE, Speckmann E-J. Focal interictal epileptiform discharges
(FIED) in the epicortical EEG and their relations to spinal field
potentials in the rat. Electroencephalogr Clin Neurophysiol. 1980;
48:447–460.
312. Elger CE, Speckmann E-J. Vertical inhibition in motor cortical
epileptic foci and its consequences for descending neuronal activity
to the spinal cord. In: Speckmann E-J, Elger CE, eds. Epilepsy and
Motor System. Munich: Urban & Schwarzenberg; 1983:152–160.
313. Classen J, Witte OW, Schlaug G, et al. Epileptic seizures triggered
directly by focal transcranial stimulation. Electroencephalogr Clin
Neurophysiol. 1995;94:19–25.
314. Aird RB, Woodbury DM. The Management of Epilepsy.
Springfield, IL: Charles C Thomas; 1974.
315. Matthes A. Epilepsie. 3rd ed. Stuttgart: Thieme; 1977.
316. Penfield W, Jasper HH. Epilepsy and the Functional Anatomy of
the Human Brain. Boston, MA: Little, Brown; 1954.
317. Penfield W, Kristiansen K. Epileptic Seizure Patterns. Springfield,
IL: Charles C Thomas; 1951.
318. Fossas P, Sanchez ME, Oller L. Occipital origin of partial oculo-
clonic versive epileptic seizures in a case of post-traumatic
epilepsy. Electroencephalogr Clin Neurophysiol. 1985;60:58P.
 Chapter 26
319. Robillard A, Saint-Hilaire J-M, Mercier M, et al. The lateralizing
and localizing value of adversion in seizures. Electroencephalogr
Clin Neurophysiol. 1982;54:41P. Abstract.
320. Ochs RF, Gloor P, Ives JR. The diagnostic value of head turning in
the localization of seizures. Electroencephalogr Clin Neurophysiol.
1981;51:21P–22P. Abstract.
321. Quesney LF, Constain M, Rasmussen T. Seizures from the dorso-
lateral frontal lobe. In: Chauvel P, Delgado-Escueta AV, Halgren E,
et al., eds. Frontal Lobe Seizures and Epilepsies. New York: Raven
Press; 1992:233–243.
322. Pachatz C, Fusco L, Vigevano F. Epileptic spasms and partial
seizures as a single ictal event. Epilepsia. 2003;44:693–700.
323. Kallen C, Wyllie E, Luders HO, et al. Hypomotor seizures in
infants and children. Epilepsia. 2002;43:882–888.
324. Meencke H-J, Janz D. Neuropathological findings in primary
generalized epilepsy: a study of eight cases. Epilepsia (New York).
1984;25:8–21.
325. Meencke HJ, Janz D. The significance of microdysgenesia
in primary generalized epilepsy: an answer to the considera-
tions to Lyon and Gastaut. Epilepsia (New York). 1985;26:
368–371.
326. Lyon G, Gastaut H. Considerations on the significance attributed
to unusual histological findings recently described in eight
patients with primary generalized epilepsy. Epilepsia (New York).
1985;26:365–367.
327. Capovilla G, Berg AT, Cross JH, et al. Conceptual dichotomies in
classifying epilepsies: partial versus generalized and idiopathic
versus symptomatic. Epilepsia. 2009;50:1645–1649.
328. Lombroso CT. Consistent EEG focalities detected in subjects
with primary generalized epilepsies monitored for two decades.
Epilepsia. 1997;38:797–812.
329. Chahine LM, Mikati MA. Benign pediatric localization-related
epilepsies. Part II. Syndromes in childhood. Epileptic Disord.
2006;8:243–258.
330. Lagae L. What’s new in: “genetics in childhood epilepsy.” Eur J
Pediatr. 2008;167:715–722.
331. Lu Y, Wang X. Genes associated with idiopathic epilepsies: a
current review. Neurol Res. 2009;31:135–143.
332. Lucarini N, Verrotti A, Napolioni V, et al. Genetic polymorphisms
and idiopathic generalized epilepsies. Pediatr Neurol. 2007;37:
157–164.
333. Helbig I, Scheffer IE, Mulley JC, et al. Navigating the channels and
beyond: unraveling the genetics of the epilepsies. Lancet Neurol.
2008;7:231–245.
334. Hirose S, Mitsudome A, Okada M, et al., on behalf of The
Epilepsy Study Group, Japan. Genetics of idiopathic epilepsies.
Epilepsia. 2005;46(suppl 1):38–43.
335. Ottman R, Hirose S, Jain S, et al. Genetic testing in the epilepsies—
Report of the ILAE genetics commission. Epilepsia. 2010;51:
655–670.
336. Scheffer IE, Zhang UH, Gecz J, et al. Genetics of the epilepsies:
genetic twists in the channels and other tales. Epilepsia. 2010;
51(suppl 1):33–26.
337. Urdinguio RG, Sanchez-Mut JV, Esteller M. Epigenetic mecha-
nisms in neurological diseases: genes, syndromes, and therapies.
Lancet Neurol. 2009;8:1056–1072.
338. Berg AT, Shinnar S, Levy SR, et al. How well can epilepsy syn-
dromes be identified at diagnosis? A reassessment 2 years after
initial diagnosis. Epilepsia. 2000;41:1269–1275.
339. Hirsch E, Velez A, Sellal F, et al. Electroclinical signs of benign
familial convulsions. Ann Neurol. 1993;34:835–841.
340. Aso K, Watanabe K. Benign familial neonatal convulsions: gener-
alized epilepsy? Pediatr Neurol. 1992;8:226–228.
■ Seizures and Epilepsy in Infants to Adolescents 535
341. Bye AM. Neonate with benign familial neonatal convulsions:
recorded generalized and focal seizures. Pediatr Neurol. 1994;10:
164–165.
342. Aicardi J, Goutieres F. Encephalopathie myoclonique neonatale.
Rev EEG Neurophysiol. 1978;8:99–101.
343. Aicardi J. Early myoclonic encephalopathy. In: Roger J, Dravet C,
Bureau M, et al., eds. Epileptic Syndromes in Infancy, Childhood,
and Adolescence. London: Libbey; 1985:12–21.
344. Dalla Bernardina B, Capovilla G, Gattoni MB, et al. Épilepsie
myoclonique grave de la prémière année. Rev. EEG Neurophysiol.
1982;12:21–25.
345. Dalla Bernardina B, Dulac O, Fejerman N, et al. Early myoclonic
epileptic encephalopathy (E.M.E.E.). Eur J Pediatr. 1983;140:
248–252.
346. Ohtahara S, Ishida T, Oka E, et al. On the specific age-dependent
epileptic syndrome. The early infantile epileptic encephalopathy
with suppression-burst [in Japanese]. No-to-Hattatsu (Tokyo).
1976;8:270–280.
347. Ohtahara S. Clinico-electrical delineation of epileptic
encephalopathies in childhood. Asian Med J. 1978;21:499–509.
348. Talada Serra M. Early epileptic encephalopathy with a pattern of
periodic bursts (Ohtahara syndrome). Electroencephalogr Clin
Neurophysiol. 1985;60:55P. Abstract.
349. Saitsu H, Kato M, Mizuguchi T, et al. De novo mutations in the
gene encoding STXBP1 (MUNC18-1) cause early infantile
encephalopathy. Nat Genet. 2008;40:782–788.
350. Fullston T, Brueton L, Willis T, et al. Ohtahara syndrome in a
family with an ARX protein truncation. Eur J Hum Genet. 2010;18:
157–162.
351. Murakami N, Ohtsuka Y, Ohtahara S. Early infantile epileptic
syndromes with suppression-bursts. Jpn J Psychiatr Neurol. 1993;47:
197–200.
352. Coppola G, Plouin P, Chiron C, et al. Migrating partial seizures
in infancy: a malignant disorder with developmental arrest.
Epilepsia. 1995;36:1017–1024.
353. Marsh E, Melamed SE, Barron T, et al. Migrating partial seizures
in infancy: expanding the phenotype of a rare seizure syndrome.
Epilepsia. 2005;46:568–72.
354. Cross JH. When is epileptic encephalopathy nonconvulsive status
epilepticus? Epilepsia. 2007;48(suppl 8):42–43.
355. Goldstein J, Slomski J. Epileptic spasms: a variety of etiologies
and associated syndromes. J Child Neurol. 2008;23:407–414.
356. Jeavons PM, Bower BD. Infantile Spasms. A Review of the
Literature and a Study of 112 Cases. London: Heinemann; 1964.
357. Fusco L, Vigevano F. Ictal clinical electroencephalographic
findings of spasms in West syndrome. Epilepsia. 1993;34:
671–673.
358. Hrachovy RA, Frost JD Jr, Kellaway P. Hypsarrhythmia: varia-
tions on a theme. Epilepsia (New York). 1984;25:317–325.
359. Kramer U, Sue WC, Mikati MA. Hypsarrhythmia: frequency of
variant patterns and correlation with etiology and outcome.
Neurology. 1997;48:197–203.
360. Kato M, Das S, Petras K, et al. Polyalanine expansion of ARX
associated with cryptogenic West syndrome. Neurology. 2003;
61:267–268.
361. Lux AL, Edwards SW, Hancock E, et al. The United Kingdom
Infantile spasms study comparing vigabatrin with prednisolone
or tetracosatide at 14 days: a multicenter randomized controlled
trial. Lancet. 2004;364:1773–1778.
362. Mackay MT, Weiss SK, Adams-Webber T, et al. Practice parameter:
medical treatment of infantile spasms: report of the American
Academy of Neurology and the Child Neurology Society. Neurology.
2004;62:1668–1681.
536 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
363. Pellock JM. Infantile spasms: US consensus report. Epilepsia. 2010,
DOI: 10.1111/j.1528–1167.2010.02657.x.
364. Elterman RD, Shields WD, Mansfield KA, et al. Randomized trial
of vigabatrin in patients with infantile spasms. Neurology. 2001;
57:1416–1421.
365. Ohtahara S, Yamatogi Y, Ohtsuka Y, et al. Prognosis of the West
syndrome with special reference to Lennox syndrome: a develop-
mental study. In: Wada JA, Penry JK, eds. Advances in Epileptology.
Xth Epilepsy International Symposium. New York: Raven Press;
1980:149–154.
366. Hughes JR. Natural history of hypsarrhythmia. Clin Electroen-
cephalogr. 1985;16:128–130.
367. Talwar D, Baldwin MA, Hutzler R, et al. Epileptic spasms in older
children: persistence beyond infancy. Epilepsia. 1995;36:151–155.
368. Camfield P, Camfield C, Lortie A, et al. Infantile spasms in remis-
sion may reemerge as intractable epileptic spasms. Epilepsia. 2003;
44:1592–1593.
369. Chugani HT, Shields WD, Shewmon DA, et al. Infantile spasms:
I. PET identifies focal cortical dysgenesis in cryptogenic cases for
surgical treatment. Ann Neurol. 1990;27:406–413.
370. Koo B, Hwang P. Localization of focal cortical lesions influences
age of onset of infantile spasms. Epilepsia. 1996;37:1068–1071.
371. Roger J, Soulayrol R, Pinsard N. Diagnostic différential. In:
Gastaut H, Soulayrol R, Roger J, et al., eds. L’Encéphalopathie
Myoclonique Infantile avec Hypsarrhythmie (Syndrome de West).
Paris: Masson; 1964:143–149.
372. Kinsbourne M. Myoclonic encephalopathy of infants. J Neurol
Neurosurg Psychiatry. 1962;25:271–276.
373. Darra F, Fiorini E, Zoccante L, et al. Benign myoclonic epilepsy of
infancy (BMEI): a longitudinal electroclinical study of 22 cases.
Epilepsia. 2006;47(suppl 5):31–35.
374. Vigevano F. Benign familial infantile seizures. Brain Dev. 2005;
27:172–177.
375. Specchio N, Vigevano F. The spectrum of benign infantile
seizures. Epilepsy Res. 2006;70(suppl 1):156–167.
376. Capovilla G, Beccaria F. Benign partial epilepsy in infancy and
early childhood with vertex spikes and waves during sleep: a new
epileptic form. Brain Dev. 2000;22:93–98.
377. Dravet C. Les Épilepsies graves de l’enfant. Vie Med. 1978;8:
543–548.
378. Dalla Bernardina B, Dulac O, Bureau M, et al. Encephalopathie
myoclonique precoce avec epilepsie. Rev EEG Neurophysiol. 1982;
12:8–14.
379. Dravet C, Bureau M, Roger J. Severe myoclonic epilepsy in
infants. In: Roger J, Dravet C, Bureau M, et al., eds. Epileptic
Syndromes in Infancy, Childhood, and Adolescence. London: Libbey;
1985:58–67.
380. Caraballo RH, Cersosimo RO, Espeche A, et al. Myoclonic status
in nonprogressive encephalopathies: study of 29 cases. Epilepsia.
2007;48:107–113.
381. Knudsen FU. Febrile convulsions. In: Dam M, Gram L, eds.
Comprehensive Epileptology. New York: Raven Press; 1991:133–143.
382. Beaumanoir A. Les Epilepsies infantiles. Problèmes de Diagnostic et
de Traitement. Basel: Editiones Roche; 1976.
383. Provisional Committee on Quality Improvement, Subcommittee
on Febrile Seizures. Practice Parameter: the neurodiagnostic eval-
uation of the child with a first simple febrile seizure. Pediatrics.
1996;97:769–775.
384. Okumura A, Ishiguro Y, Sofue A, et al. Treatment and outcome in
patients with in patients with febrile convulsion associated with
epileptiform discharges on electroencephalography. Brain Dev.
2004;26:241–244.
385. Nakayama J, Arinami T. Molecular genetics of febrile seizures.
Epilepsy Res. 2006;70S:S190–S198.
386. Heida JG, Moshe SL, Pittman QJ. The role of interleukin 1-B in
febrile seizures. Brain Dev. 2009;31:388–393.
387. Lennox-Buchthal MA. Fieberkrémpfe. Stuttgart: Hippokrates; 1977.
388. Des Termes H, Mises J, Plouin P, et al. The “spike focus” during the
evolution of febrile convulsions: an electrophysiological and clin-
ical study of 35 patients. Electroencephalogr Clin Neurophysiol.
1978;45:370. Abstract.
389. Yamatogi Y, Ishida S, Terasaki T, et al. An electroencephalo-
graphic study of febrile convulsions. Electroencephalogr Clin
Neurophysiol. 1982;54:27P–28P. Abstract.
390. Yamatogi Y, Ohtahara S. EEG in febrile convulsions. Am J EEG
Technol. 1990;30:267–280.
391. Panayiotopoulos CP. Benign Childhood Partial Seizures and
Related Epileptic Syndromes. London: Libbey; 1999.
392. Panayiotopoulos CP. The Panyiotopoulos Syndrome. London:
Libbey; 2002.
393. Caraballo R, Cersosimo R, Fejerman N. Panayiotopoulos syn-
drome: a prospective study of 192 patients. Epilepsia. 2007;48:
1054–1061.
394. Specchio N, Trivisano M, Di Ciommo V, et al. Panayiotopoulos
syndrome: a clinical, EEG, and neuropsychological study of 93
consecutive patients. Epilepsia. 2010, DOI: 10.1111/j.1528–1167.
2010.02639.x.
395. Leal AJR, Ferreira JC, Dias AI, et al. Origin of frontal lobe spikes
in the early onset benign occipital lobe epilepsy (Panayiotopoulos
syndrome). Clin Neurophysiol. 2008;119:1985–1991.
396. Yoshinaga H, Kobayashi K, Ohtsuka Y. Characteristics of the
synchronous occipital and frontal spike phenomenon in
Panayiotopoulos syndrome. Brain Dev. 2010;32(8):603–608.
397. Doose H. Das Akinetische Petit mal. Arch Psychiatr Nervenkr.
1964;205:625–654.
398. Doose H. Myoclonic astatic epilepsy of early childhood. In: Roger
J, Dravet C, Bureau M, et al., eds. Epileptic Syndromes in Infancy,
Childhood and Adolescence. London: Libbey; 1985:78–88.
399. Bonanni P, Parmeggiani L, Guerrini R. Different neurophysiologi-
cal patterns of myoclonus characterize Lennox–Gastaut syndrome
and myoclonic-astatic epilepsy. Epilepsia. 2002;43:609–615.
400. Kilaru S, Bergqvist C. Current treatment of myoclonic astatic
epilepsy: clinical experience at the Children’s Hospital of
Philadelphia, PA. Epilepsia. 2007;48:1703–1707.
401. Gastaut Y. Un élément Džroutant De La Séméiologie élec-
troencžphalographique: les pointes rolandiques sans signfication
focale. Rev Neurol (Paris). 1952;87:448–450.
402. Hess R. Verlaufsuntersuchungen über Anfälle und EEG bei
kindlichen Epilepsien. Arch Psychiatr Nervenkr. 1958;197:568–593.
403. Nayrac P, Beaussart M. Les pointes-ondes prérolandiques.
Expression EEG trés particulière: étude électroclinique de 21 cas.
Rev Neurol (Paris). 1958;99:201–206.
404. Isler W, Hess R. Verlaufsuntersuchungen Ÿber AnfŠlle und EEG
bei fokalen Epilepsien im Kindesalter. Arch Psychiatr Nervenkr.
1960;200:257–266.
405. Beaumanoir A, Ballis T, Varfis G, et al. Benign epilepsy of childhood
with rolandic spike. A clinical electroencephalographic and telen-
cephalographic study. Epilepsia (Amsterdam). 1974;15:301–315.
406. Beaussart M. Benign epilepsy of children with rolandic (centro-
temporal) paroxysmal foci. A clinical entity. Study of 221 cases.
Epilepsia (Amsterdam). 1972;13:795–811.
407. Blom S, Brorson LO. Central spikes or sharp waves (rolandic
spikes) in children’s EEG and their clinical significance. Acta
Pediatr Scand. 1966;55:385–393.
 Chapter 26
408. Blom S, Heijbel J. Benign epilepsy of children with centrotempo-
ral EEG foci: a follow-up study in adulthood of patients initially
studied as children. Epilepsia (New York). 1982;23:629–632.
409. Dalla Bernardina B, Beghini G. Rolandic spikes in children with
and without epilepsy (20 subjects photographically studied dur-
ing sleep). Epilepsia (New York). 1976;17:161–168.
410. Hejbel J, Blom S, Bergfors PG. Benign epilepsy of children with
centrotemporal EEG foci. A study of incidence rate in outpatient
care. Epilepsia (New York). 1975;16:657–664.
411. Lairy GC, Harrison A. Functional aspects of EEG foci in children.
In: Kellaway P, Petersen I, eds. Clinical Electroencephalography in
Children. New York: Grune & Stratton (Stockholm: Almqvist &
Wiksell); 1968:197–212.
412. Lombroso CT. Sylvian seizures and midtemporal spike foci in
children. Arch Neurol (Chicago). 1967;17:52–59.
413. Loiseau P, Beaussart M. The seizures of benign childhood epilepsy
with Rolandic paroxysmal discharges. Epilepsia (Amsterdam).
1973;14:381–389.
414. Niedermeyer E. Spitzen über der Zentralregion und mu
Rhythmus: gedanken zum Problem der “funktionellen” Spitzen.
Z EEG-EMG. 1970;1:133–141.
415. Niedermeyer E. Focal and generalized seizure discharges in the
electroencephalogram and their responses to intravenous
diazepam. Int Med Dig. 1972;7:49–61.
416. Gastaut H. Individualization of so-called benign and functional
epilepsy at different times of life. Electroencephalogr Clin Neuro-
physiol. 1982;53:79P. Abstract.
417. Loiseau P, Pestre M, Dartigues F, et al. Long-term prognosis in
two forms of childhood epilepsy: typical absence seizures and
epilepsy with rolandic (centrotemporal) EEG foci. Ann Neurol.
1983;13:642–648.
418. Niedermeyer E. Complexities of primary generalized epilepsy.
Clin Electroencephalogr. 1981;12:177–191.
419. Nogueira de Melo A, Niedermeyer E. Crossover phenomena in
epileptic syndromes in childhood epilepsies. Clin Electroen-
cephalogr. 1991;22:75–82.
420. Aicardi J, Chevrie JJ, Rousselie F. Le syndrome spasmes en
flexion, agénesie calleuse, anomalies choriorétiniennes. Archive
Francaise de Pediatric. 1969;26:1103–1120.
421. Blom S, Heijbel J, Bergfors PG. Benign epilepsy of children with
centrotemporal EEG foci. Prevalence and follow-up study of 40
patients. Epilepsia (Amsterdam). 1972;13:609–613.
422. Gregory DL, Wong PKH. Clinical relevance of a dipole field in
rolandic spikes. Epilepsia (New York). 1992;33:36–44.
423. Wong PK. The importance of source behavior in distinguishing
populations of epileptic foci. J Clin Neurophysiol. 1993;10:314–322.
424. Van der Meij W, Van Huffelen AC, Wieneke GH, et al. Sequential
EEG mapping may differentiate “epileptic” from “nonepileptic”
rolandic spikes. Electroencephalogr Clin Neurophysiol. 1992;82:
408–414.
425. Luders H, Lesser RP, Dinner DS, et al. Benign focal epilepsy of
childhood. In: Luders H, Lesser RP, eds. Epilepsy Electroclinical
Syndromes. Berlin: Springer; 1987:303–346.
426. Gutierrez AR, Brick JF, Bodensteiner J. Dipole reversal: an ictal
feature of benign partial epilepsy with centrotemporal spikes.
Epilepsia. 1990;31:544–548.
427. Legarda S, Jayakar P. Electroclinical significance of rolandic
spikes and dipoles in neurodevelopmentally normal children.
Electroencephalogr Clin Neurophysiol. 1995;95:257–259.
428. Yoshinaga H, Amano R, Oka E, et al. Dipole tracing in childhood
epilepsy with special reference to rolandic. Brain Topogr. 1992;
4:193–199.
■ Seizures and Epilepsy in Infants to Adolescents 537
429. Niedermeyer E. Dipole theory and electroencephalography. Clin
Electroencephalogr. 1996;27:121–131.
430. Fonseca LC, Tedrus GM, Bastos A, et al. Reactivity to rolandic
spikes. Clin Electroencephalogr. 1996;27:116–120.
431. Nadkarni MA, Postolache V, Gold A, et al. Central mid-temporal
spikes triggered by blinking. Electroencephalogr Clin Neurophysiol.
1994;90:36–39.
432. Mitsudome A, Ohu M, Yasumoto S, et al. Rhythmic slow activity
in benign childhood epilepsy with centrotemporal spikes. Clin
Electroencephalogr. 1997;28:44–48.
433. Nelson KR, Brenner RP, De La Paz D. Midline spikes—electroen-
cephalographic and clinical features. Electroencephalogr Clin
Neurophysiol. 1984;57:42P–43P. Abstract.
434. Pedley TA, Tharp BR, Herman KR. Electroencephalographic and
clinical correlates of vertex spike foci. Electroencephalogr Clin
Neurophysiol. 1980;50:153P. Abstract.
435. Pourmand RA, Markand ON, Thomas C. Midline spike dis-
charges: clinical and EEG correlates. Clin Electroencephalogr. 1984;
15:232–236.
436. Ehle AL. Midline spike foci in children. Electroencephalogr Clin
Neurophysiol. 1980;50:153P. Abstract.
437. Weglage J, Demsky A, Pietsky M, et al. Neuropsycholgical, intel-
lectual, and behaviorla findings in patients with centrotemporal
spikes with and without seizures. Dev Med Child Neurol. 1997;
39:646–651.
438. Staden U, Isaacs E, Boyd SG, et al. Language dysfunction in chil-
dren with rolandic epilepsy. Neuropediatrics. 1998;29:242–248.
439. Croona C, Kihlgren M, Lundberg S, et al. Neuropsychological
findings in children with benign childhood epilepsy with cen-
trotemporal spikes. Dev Med Child Neurol. 1999;41:813–818.
440. Peters JM, Camfield CS, Camfield PR. Population study of benign
Rolandic epilepsy: is treatment needed? Neurology. 2001;57:
537–539.
441. Lerman P. Benign partial epilepsy with centro-temporal spikes.
In: Roger J, Bureau M, Dravet C, et al., eds. Epileptic Syndromes in
Infancy, Childhood and Adolescence. 2nd ed. London: John Libbey
and Company; 1992:189–200.
442. Lerman P, Kivity S. Benign focal epilepsy of childhood. A follow-
up study of 100 recovered patients. Arch Neurol. 1975;23:
261–264.
443. Kramer U, Zelnik N, Lerman-Sagie T, et al. Benign childhood
epilepsy with centrotemporal spikes: clinical characteristics and
identification of patients at risk for multiple seizures. J Child
Neurol. 2002;17:17–19.
444. Massa R, de Saint Martin A, Carcangiu R, et al. EEG criteria pre-
dictive of a complicated evolution in idiopathic rolandic epilepsy.
Neurology. 2001;57:1071–1079.
445. Bray PF, Wiser WC. The relation of focal to diffuse epileptiform
EEG discharges in genetic epilepsy. Arch Neurol (Chicago). 1965;13:
223–237.
446. Doose H, Gerken H, Kiefer R, et al. Genetic factors in childhood
epilepsy with focal sharp waves. II. EEG findings in patients and
siblings. Neuropadiatrie. 1977;8:10–20.
447. Gereby G. Benign centro-temporal epilepsy of childhood. EEG
and clinical follow-up of 80 cases. Electroencephalogr Clin
Neurophysiol. 1985;61:7P. Abstract.
448. Gelisse P, Corda D, Raybaud C, et al. Abnormal neuroimaging in
patients with benign epilepsy with centrotemporal spikes.
Epilepsia. 2003;44:372–378.
449. Bali B, Kull LL, Strug LJ, et al. Autosomal dominant inheritance
of centrotemporal sharp waves in Rolandic epilepsy families.
Epilepsia. 2007;48:2266–2272.
538 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
450. Strug LJ, Clarke T, Chiang T, et al. Centrotemporal sharp wave
EEG trait in rolandic epilepsy maps to Elongator Protein
Complex 4 (ELP4). Eur J Hum Genet. 2009;17:1171–1181.
451. Lerman P, Kivity-Ephraim S. Focal epileptic EEG discharges in chil-
dren not suffering from clinical epilepsy: etiology, clinical signifi-
cance and management. Epilepsia (New York). 1981;22:551–558.
452. Gibbs FA. Differentiation of mid-temporal, anterior temporal
and diencephalic epilepsy. In: Baldwin M, Bailey P, eds. Temporal
Lobe Epilepsy. Springfield, IL: Charles C Thomas; 1958:109–117.
453. Andermann F, Oguni H. Do epileptic foci migrate? The pros.
Electroencephalogr Clin Neurophysiol. 1991;76:96–99.
454. Blume WT. Do epileptic foci migrate? The cons. Electroen-
cephalogr Clin Neurophysiol. 1991;76:100–105.
455. Lee IC, Chen YJ, Lee HS. Migration of epileptic spike foci in
encephalograms may correlate with a better outcome in pediatric
epilepsy. Brain Dev. 2010; Jan 8.
456. Niedermeyer E, Rocca U. Scalp, cortical and depth EEG contribu-
tion to focal motor epilepsy and epilepsia partialis continua.
Electroencephalogr Clin Neurophysiol. 1980;50:160P. Abstract.
457. Rocca U, Niedermeyer E. Severe forms of focal motor seizure dis-
orders in childhood. In: Proc. Congr. Latin. Amer. (Buenos Aires),
1982:277–291.
458. Andrews PI, McNamara JO, Lewis DV. Clinical and electroen-
cephalographic correlates in Rasmussen’s encephalitis. Epilepsia.
1997;38:189–194.
459. Aguilar MJ, Rasmussen T. Role of encephalitis in pathogenesis of
epilepsy. Arch Neurol (Chicago). 1960;2:663–676.
460. Rasmussen T, Olszewski J, Lloyd-Smith D. Focal seizures due to
chronic localized encephalitis. Neurology (Minneapolis). 1958;8:
435–445.
461. Bien CG, Granata T, Antozzi C, et al. Pathogenesis, diagnosis, and
treatment of Rasmussen encephalitis. A European consensus
statement. Brain. 2005;128:454–471.
462. Bien CG, Schramm J. Treatment of Rasmussen encephalitis half
a century after its initial description: promising prospects and a
dilemma. Epilepsy Res. 2009;86:101–112.
463. Granata T, Fusco L, Gobbi G, et al. Experience with immunomod-
ulatory treatments in Rasmussen’s encephalitis. Neurology. 2003;
61:1807–1810.
464. Daniel RT, Villemure J-G, Granata T, et al. Experience with
immunomodulatory treatments of Rasmussen encephalitis.
Neurology. 2004;63:1761–1762.
465. Otsubo H, Chitoku S, Ochi A, et al. Malignant rolandic-sylvian
epilepsy in children. Diagnosis, treatment and outcomes. Neuro-
logy. 2001;57:590–596.
466. Ong HT, Wyllie E. Benign childhood epilepsy with centrotempo-
ral spikes: is it always benign? Neurology. 2000;54:1182–1185.
467. Raju P, Sarco DP, Poduri A, et al. Oxcarbazepine in children with
nocturnal frontal lobe epilepsy. Pediatr Neurol. 2007;37:345–349.
468. Gastaut H. A new type of epilepsy: benign partial epilepsy of
childhood with occipital spike-waves. Clin Electroencephalogr.
1982;13:13–22.
469. Taylor I, Scheffer IE, Berkovic SF. Occipital epilepsies: identifica-
tion of specific and newly recognized syndromes. Brain. 2003;
126:753–769.
470. Taylor I, Berkovic SF, Kivity S, et al. Benign occipital epilepsies of
childhood. Brain. 2008;131:2287–2294.
471. Panayiotopoulos CP, Michael M, Sanders S, et al. Benign child-
hood focal epilepsies: assessment of established and newly recog-
nized syndromes. Brain. 2008;131:2264–2286.
472. Niedermeyer E, Riggio S, Santiago M. Benign occipital lobe
epilepsy. J Epilepsy. 1988;1:3–11.
473. Panayiotopoulos CP. Basilar migraine, seizures and severe epileptic
EEG abnormalities. Neurology (Minneapolis). 1980;30:1122–1125.
474. Riggio S, Santiago M, Niedermeyer E. Benign occipital lobe
epilepsy. Neurology. 1987;37(suppl I):106. Abstract.
475. Lugaresi E, Cirignotta F, Montagna P. Occipital lobe epilepsy with
scotosensitive seizures: the role of central vision. Epilepsia (New
York). 1984;25:115–120.
476. Tassinari CA, Bureau M. Epilepsy with myoclonic absences. In:
Roger J, Dravet C, Bureau M, et al., eds. Epileptic Syndromes in
Infancy, Childhood and Adolescence. London: John Libbey and
Co; 1985:121–129.
477. Bureau M, Tassinari CA. Epilepsy with myoclonic absences. Brain
Dev. 2005;27:178–184.
478. Manonmani V, Wallace SJ. Epilepsy with myoclonic absences.
Arch Dis Child. 1994;70:288–290.
479. Elia M, Guerrini R, Musumeci SA, et al. Myoclonic absence-like
seizures and chromosomal abnormalities. Epilepsia. 1998;39:
660–663.
480. Niedermeyer E. The Lennox–Gastaut syndrome: a severe type of
childhood epilepsy. Electroencephalogr Clin Neurophysiol.
1968;24:283. Abstract.
481. Oller-Daurella L. Sindrome de Lennox. Barcelona: Editorial
Espaxs; 1967.
482. Oller-Daurella L. Evolution et pronostic du syndrome de
Lennox–Gastaut. In: Lugaresi E, ed. Evolution and Prognosis of the
Epilepsies. Bologna: Gaggi; 1973:155–164.
483. Oller-Daurella L. Las fronteras entre et petit mal y el sindrome de
Lennox–Gastaut. Rev Esp Otoneurooftalmol Neurocir. 1976;34:
27–44.
484. Arzimanoglou A, French J, Blume WT, et al. Lennox–Gastaut
syndrome: a consensus approach on diagnosis, assessment,
management, and trial methodology. Lancet Neurol. 2009;8:
82–93.
485. Gastaut H, Gastaut JL, Gonçalves e Silva E, et al. Relative fre-
quency of different types of epilepsy: a study employing the clas-
sification by the International League against Epilepsy. Epilepsia
(New York). 1975;16:457–461.
486. Komai S. Lennox–Gastaut’s syndrome. Prognosis of the secondary
generalized epilepsies. Epilepsia (New York). 1977;18:131. Abstract.
487. Bauer G, Benke T, Bohr K. The Lennox–Gastaut syndrome in
adulthood. In: Niedermeyer E, Degen R, eds. The Lennox–Gastaut
Syndrome. New York: Alan R. Liss; 1988:317–327.
488. Olmos-Garcia de Alba G, Valdez JM, Crespo FV. West syndrome
evolving into the Lennox–Gastaut syndrome. Clin Electroen-
cephalogr. 1984;15:61–68.
489. Gastaut H, Roger J, Soulayrol R, et al., eds. L’Encéphalopathie
Myoclonique Infantile avec Hypsarrhythmie (Syndrome de West).
Paris: Masson; 1964.
490. Patry G, Lyagoubi S, Tassinari A. Subclinical “electrical status
epilepticus” induced by sleep in children. A clinical and electroen-
cephalographic study of six cases. Arch Neurol. 1971;24:242–252.
491. Tassinari CA, Daniele O, Dravet C, et al. Sleep polygraphic studies
in some epileptic encephalopathies from infancy to adolescence. In:
Degen R, Niedermeyer E, eds. Epilepsy, Sleep and Sleep Deprivation.
Amsterdam: Elsevier; 1984:175–189.
492. Niedermeyer E, Freeman JM, Long DM, et al. EEG studies in
recalcitrant and disabling focal motor seizures. Epilepsia (New
York). 1977;18:289. Abstract.
493. Niedermeyer E. Considerations of the centrencephalic (general-
ized) type of epilepsy. Del Med J. 1966;38:341–348.
494. Nogueira de Melo A, Niedermeyer E. Considerations of second-
ary temporalization. Clin Electroencephalogr. 1991;22:161–171.
 Chapter 26
495. Hughes JR, Patil VK. Long-term electro-clinical changes in the
Lennox–Gastaut syndrome before, during, and after the flow
spikewave pattern. Clin Electroencephalogr. 2002;33:1–7.
496. Ohtahara S, Ohtsuka Y, Kobayashi K. Lennox–Gastaut syndrome:
a new vista. Psychiatry Clin Neurosci. 1995;49:S179–S183.
497. Hirt HR. Zur Nosologie des Lennox–Gastaut Syndromes.
Nervenarzt. 1996;67:109–122.
498. Tassinari CA, Bureau M, Dravet C, et al. Electrical status epilep-
ticus during sleep in children (ESES). In: Sterman MB, Shouse
MN, Passouant P, eds. Sleep and Epilepsy. New York: Academic
Press; 1982:465–479.
499. Kramer U, Nevo Y, Neufeld MY, et al. Epidemiology of epilepsy in
childhood: a cohort of 440 consecutive patients. Pediatr Neurol.
1998;18:46–50.
500. Guilhoto LMFF, Morrell F. Electrophysiological differences
between Landau–Kleffner syndrome and other conditions show-
ing the CSWS electrical pattern. Epilepsia. 1994;35(suppl 8):126.
501. Hirsch E, Maquet P, Metz-Lutz M-N, et al. The eponym
“Landau–Kleffner syndrome” should not be restricted to child-
hood-acquired aphasia with epilepsy. In: Beaumanoir A, Bureaus
M, Deonna T, et al., eds. Continuous Spikes and Waves During
Slow Sleep. Electrical Status Epilepticus During Slow Sleep.
London: Libbey; 1995:57–62.
502. Landau W, Kleffner FR. Syndrome of acquired aphasia with con-
vulsive disorder in children. Neurology. 1957;7:523–530.
503. Stefanatos GA, Green GG, Ratcliff GG. Neurophysiological evi-
dence of auditory channel anomalies in developmental dyspha-
sia. Arch Neurol. 1989;46:871–875.
504. Kuhlo W. Petit-mal-Epilepsie. In: Niedermeyer E, ed. Epilepsy,
Recent Views on Theory, Diagnosis and Therapy of Epilepsy. Basel:
S. Karger; 1970:120–138.
505. Sauer H. Über gehäufte kleine Anfälle bei Kindern (Pyknolepsie).
Monatsschr Psychiatr Neurol. 1916;40:267–300.
506. Loiseau P, Duché B, Pédespan J-M. Absence epilepsies. Epilepsia.
1995;36:1182–1186.
507. Sadleir LG, Scheffer IE, Smith S, et al. Factors influencing clinical
features of absence seizures. Epilepsia. 2008;49:2100–2107.
508. Paal G. Katamnestische Untersuchungen und EEG bei Pyknolepsie.
Arch Psychiatr Nervenkr. 1957;196:48–62.
509. Livingston S, Torres I, Pauli L, et al. Petit mal epilepsy. Results
of a prolonged follow-up of 117 patients. JAMA. 1965;194:
227–232.
510. Janz D. Aufwach-Epilepsien. Arch Psychiatr Nervenkr. 1953;191:
73–98.
511. Janz D. The grand mal epilepsies and the sleeping-waking cycle.
Epilepsia (Amsterdam). 1962;3:69–109.
512. Waltz S, Beck-Managetta G, Janz D. Are there syndrome-related
genetically determined spike and wave patterns? A comparison
between syndromes of generalized epilepsy. Epilepsia 1990;31:
819 (abstract).
513. Camfield CS, Camfield PR, Sadler M, et al. Paroxysmal eyelid
movements: a confusing feature of generalized photosensitive
epilepsy. Neurology. 2004;63:40–42.
514. Tovia E, Goldberg-Stern H, Shahar E, et al. Outcome of children
with juvenile absence epilepsy. J Child Neurol. 2006;21:766–768.
515. Tezer FI, Sahin G, Ciger A, et al. Focal EEG findings in juvenile
absence syndrome and the effect of antiepileptic drugs. Clin EEG
Neurosci. 2008;39:33–38.
516. Tsuboi T. Primary Generalized Epilepsy with Sporadic Myoclonias
of Myoclonic Petit Mal Type. Stuttgart: Thieme; 1977.
517. Delgado-Escueta AV, Enrile-Bascal F. Juvenile myoclonic epilepsy
of Janz. Neurology. 1984;34:285–294.
■ Seizures and Epilepsy in Infants to Adolescents 539
518. Asconape J, Penry JK. Some clinical and EEG aspects of benign
juvenile myoclonic epilepsy. Epilepsia. 1984;25:108–114.
519. Schmitz B, Sander T, eds. Juvenile Myoclonic Epilepsy. Wrightson:
Petersfield; 2000.
520. Oguni H, Mukahira K, Oguni M, et al. Video-polygraphic analy-
sis of myoclonic seizures in juvenile myoclonic epilepsy. Epilepsia.
1994;35:307–316.
521. Montalenti E, Imperiale D, Rovera A, et al. Clinical features, EEG
findings, and diagnostic pitfalls in juvenile myoclonic epilepsy: a
series of 63 patients. J Neurol Sci. 2001;184:65–70.
522. Lancman ME, Asconape JJ, Penry JK. Clinical and EEG asymme-
tries in juvenile myoclonic epilepsy. Epilepsia. 1994;35:302–306.
523. So GM, Thiele EA, Sanger T, et al. Electroencephalogram and
clinical focalities in juvenile myoclonic epilepsy. J Child Neurol.
1998;13:541–545.
524. Aliberti V, Grunewald RA, Panayiotopoulos CP, et al. Focal elec-
troencephalographic abnormalities in juvenile myoclonic
epilepsy. Epilepsia. 1994;35:297–301.
525. de Carvalho Sousa NA, da Silva Sousa P, Garzon E, et al. EEG
recording after sleep deprivation in a series of patients with juvenile
myoclonic epilepsy. Arq Neuropsiquiatr. 2005;63(2-B):383–388.
526. Camfield CS, Camfield PR. Juvenile myoclonic epilepsy 25 years
after seizure onset: a population-based study. Neurology. 2009;73:
1041–1045.
527. Wirrell EC, Camfields CS, Camfield PR, et al. Long-term progno-
sis of typical childhood absence epilepsy: remission or progres-
sion to juvenile myoclonic epilepsy. Neurology. 1996;47:912–918.
528. Blume WT. Clinical and electroencephalographic correlates of
the multiple independent spike foci pattern in children. Ann
Neurol. 1978;4:541–547.
529. Yamatogi Y, Ohtahara S. Multiple independent spike foci with
epilepsy, with special reference to a new epileptic syndrome of
“severe epilepsy with multiple independent spike foci.” Epilepsy
Res. 2006;70(S1):S96–104.
530. Kotagal P. Multifocal independent Spike syndrome: relationship
to hypsarrhythmia and the slow spike-wave (Lennox–Gastaut)
syndrome. Clin Electroencephalogr. 1995;26:23–29.
531. Hauser WA, Rich S, Anderson VE. The multifocal spike pattern
and sibling risk for epilepsy. Electroencephalogr Clin Neurophysiol.
1984;57:44P–45P. Abstract.
532. Elger CE, Speckmann E-J. Interiktale epileptiforme Potentiale im
corticalen OberflŠchen-EEG und ihre Beziehungen zu spinalen
Feldpotentialen bei der Ratte. In: Doose M, Gross-Selbeck G, eds.
Epilepsie 1978. Stuttgart: Thieme; 1979:245–249.
533. Lugaresi E, Pazzaglia P. The EEG in the positive and differential
diagnosis of epilepsy. In: Remond A, ed-in-chief. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 13A.
Amsterdam: Elsevier; 1975:69–71.
534. Engel J Jr. A practical guide for routine EEG studies in epilepsy. J
Clin Neurophysiol. 1984;1:109–142.
535. Dinner, D.S., and Lueders, H. The electroencephalogram in the
routine evaluation of patients with epilepsy. In: Porter RJ, Morselli
PL, eds. The Epilepsies. London: Butterworth; 1985:142–173.
536. Hughes JR. The significance of the interictal spike discharges. A
review. J Clin Neurophysiol. 1989;6:207–226.
537. Ajmone Marsan C, Zivin LS. Factors related to the occurrence of
typical paroxysmal abnormalities in the EEG records of epileptic
patients. Epilepsia (Amsterdam). 1970;11:361–381.
538. Bauer G. Der Wert von EEG-Kontrollen mglichst bald nach
einem epileptischen Anfall. Z EEG-EMG. 1975;6:125–130.
539. Kaibara M, Blume WT. The postictal electroencephalogram.
Electroencephalogr Clin Neurophysiol. 1988;70:99–104.
540 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
540. Gastaut H, Gomez-Almanzar M, Taury M. Der provozierte
Mittagsschlaf: eine einfache erfolgreiche Methode zur Schlafak-
tivierung bei Epileptikern. Z EEG-EMG. 1983;14:1–5.
541. Gastaut H, Gomex-Almanzar M, Taury M. The enforced nap: a
simple effective method of inducing sleep activation in epileptics.
In: Degen R, Niedermeyer E, eds. Epilepsy, Sleep and Sleep
Deprivation. Amsterdam: Elsevier; 1984:75–83.
542. Niedermeyer E, Hšller L. Kurzschlaf im EEG-Eine Fundgrube sonst
tibersehener EEG-AbnormalitŠten. Z EEG-EMG. 1984;15: 57–66.
543. Binnie CD, Veldhuizen R, Beintema DJ. Evaluation of recording
after sleep deprivation in the diagnostic EEG assessment of epilepsy.
Electroencephalogr Clin Neurophysiol. 1982;54:21P–22P. Abstract.
544. Degen R, Degen H-E. Sleep and sleep deprivation in epileptology.
In: Degen R, Niedermeyer E, eds. Epilepsy, Sleep and Sleep
Deprivation. Amsterdam: Elsevier; 1984:273–286.
545. Ellingson RJ, Wilken K, Bennett DR. Efficacy of sleep deprivation
as an activation and procedure in epilepsy patients. J Clin
Neurophysiol. 1984;1:83–101.
546. Niedermeyer E. Generalized seizure discharges and possible pre-
cipitating mechanisms. Epilepsia (Amsterdam). 1966;7:23–29.
547. Niedermeyer E. Petit mal, primary generalized epilepsy and sleep.
In: Sterman MB, Shouse MN, Passouant P, eds. Sleep and
Epilepsy. New York: Academic Press; 1982:191–207.
548. Niedermeyer E. Awakening epilepsy (“Aufwach-Epilepsie”) revis-
ited 30 years later. In: Degen SR, Niedermeyer E, eds. Epilepsy,
Sleep and Sleep Deprivation. Amsterdam: Elsevier; 1984:85–94.
549. Halász P. Generalized epilepsy with spike-wave paroxysms as an
epileptic disorder of the function of sleep promotion. Acta
Physiol Acad Sci Hung. 1981;57:51–86.
550. Halász P. Sleep, arousal and electroclinical manifestations of gener-
alized epilepsy with spike wave pattern. In: Degen R, Niedermeyer
E, eds. Epilepsy, Sleep and Sleep Deprivation. Amsterdam: Elsevier;
1984:97–107.
551. Halász P, Dévényi E. Petit mal absences in night sleep with special
reference to transitional sleep and REM periods. Acta Med Acad
Sci Hung. 1974;31:31–45.
552. Nowack WJ. Dyshormia revisited: generalized seizures and
arousal. Clin Electroencephalogr. 1996;27:22–25.
553. Sato S, Dreifuss FE, Penry JK. The effect of sleep on spike wave
discharges in absence seizures. Neurology (Minneapolis). 1973;23:
1335–1345.
554. Passouant P, Cadilhac J. Décharges épileptiques. In: Niedermeyer
E, ed. Epilepsy. Recent Views on Theory, Diagnosis and Therapy of
Epilepsy. Basel: S. Karger; 1970:87–104.
555. Tomka I. Die funktionelle Beziehung zwischen Petit mal-
Epilepsie und Schlaf. Z EEG-EMG. 1983;14:154–159.
556. Kellaway P, Frost JD Jr, Crawley JW. The relationship between
sleep spindles and spike-and-wave bursts in human epilepsy. In:
Avoli M, Gloor P, Kostopoulos G, et al., eds. Generalized Epilepsy.
Boston, MA: Birkhauser; 1990:36–84.
557. Avoli M, Gloor P, Kostopoulos G, et al. An analysis of penicillin-
induced generalized spike and wave discharges using simultane-
ous recordings of cortical and thalamic single neurons. J
Neurophysiol. 1983;50:819–837.
558. Gloor P, Avoli M, Kostopoulos G. Thalamocortical relation-
ships in generalized epilepsy with bilaterally synchronous
spike-and-wave discharge. In: Avoli M, Gloor P, Kostopoulos
G, et al., eds. Generalized Epilepsy. Boston, MA: Birkhauser; 1990:
190–212.
559. Kostopoulos G, Gloor P, Pellegrini A, et al. A study of the tran-
sition from spindles to spike and wave discharges in feline gen-
eralized penicillin epilepsy: EEG features. Exp Neurol. 1981;73:
43–54.
560. Marciani MG, Gotman J, Andermann, F. Changes in interictal and
ictal activity during anticonvulsant withdrawal in epileptic patients.
Electroencephalogr Clin Neurophysiol. 1984;58:32P. Abstract.
561. Marks DA, Katz A, Scheyer R, et al. Clinical and electrographic
effects of acute anticonvulsant withdrawal in epileptic patients.
Neurology. 1991;41:508–512.
562. Van Parys JAP. The use of EEG as a method of control in epilepsy.
Electroencephalogr Clin Neurophysiol. 1981;52:41P. Abstract.
563. Auvin S, Pandit F, De Bellecize J, et al. Benign myoclonic epilepsy
in infancy: electroclinical features and long-term follow-up of 34
patients. Epilepsia. 2006;47:387–393.
564. Blume WT, Luders HO, Mizrahi E, et al. Glossary of descriptive
terminology for ictal semiology: Report of the ILAE task force on
classification and terminology. Epilepsia. 2001;42:1212–1218.
565. Dura-Trave T, Yoldi-Petri ME, Gallinas-Victoriano F. Epilepsy in
children in Navarre, Spain: epileptic seizure types and epileptic
syndromes. J Child Neurol. 2007;22:823–828.
566. Ferrie CD, Caraballo R, Covanis A, et al. Autonomic status epilep-
ticus in Panayiotopoulos syndrome and other childhood and
adult epilepsies: a consensus view. Epilepsia. 2007;48:1165–1172.
567. Legarda S, Jayakar P, Duchowny M, et al. High central and low
central subgroups. Epilepsia. 1994;35:1125–1129.
568. Legarda S, Jayakar P, Duchowny M, et al. Benign rolandic epilepsy:
high central and low central subgroups. Epilepsia. 1994;35:
1125–1129.

Epilepsy in Adults and the Elderly
BERNARD S. CHANG, DONALD L. SCHOMER AND ERNST NIEDERMEYER
INTRODUCTION
It is difficult to imagine the practice of modern epileptology
without the availability of electroencephalography (EEG), in all
of its various forms. The two fields have been closely inter-
twined over the past decades, the latter being the objective
accompaniment to the clinical art of the former. As some
epileptologists have remarked, “one feels that one knows a
patient a little better after seeing his EEG.” While electroen-
cephalography (EEG) clearly has many other important uses in
neurologic practice (detailed throughout this volume), it has
revolutionized the entire field of epileptology and continues to
do so with each technical advance.
This chapter defines and describes the various EEG abnor-
malities seen in patients with epilepsy, including both those that
occur during seizures and those that occur between seizures, as
well as similar EEG findings that have less certain significance
and relationship to seizure disorders.
PRINCIPAL DIFFERENCES BETWEEN
INTERICTAL AND ICTAL DISCHARGES
Naturally, one would like a distinction to be made between ictal
and interictal paroxysmal activities. This logical differentiation,
however, is beset with great difficulties. It is sometimes difficult
or even impossible to distinguish between ictal–clinical,
ictal–subclinical, and interictal paroxysmal EEG activities, even
with the use of neuropsychologic tests. We must acknowledge
the existence of a gray zone between interictal and ictal parox-
ysmal activities.
Sudden Change of Frequency
There are valid guidelines for the detection of true ictal events.
The onset of a clinical seizure may be characterized by a sudden
change of frequency. A new type of rhythm appears, hesitantly,
and then more distinctly; soon it boldly dominates the tracing.
This rhythm may be in alpha frequency or it may be faster or
slower; it clearly demonstrates a new element of the tracing,
indicative of a completely new electrophysiologic event. The
abnormal rhythm may or may not show spiky character. It
tends to become slower with increasing amplitudes and more
distinct spiky phases of the rhythmic waves.
Sudden Loss of Voltage
Sudden “desynchronization” of electrical activity is found in
electrodecremental seizures. The onset of these attacks may look
almost flat locally and/or diffusely, but extremely fast very low
CHAPTER
27
voltage activity may gradually increase in voltage with decreas-
ing frequencies. Ictal rhythmic activity may soon become pre-
ponderant, similar to that found in the sudden change of
frequency. Arroyo et al. (1) have stressed the role of the frontal
lobes in electrodecremental seizures. Electrodecremental
seizure represents the most common EEG expression of infan-
tile spasms occurring in the West syndrome (2,3).
Sudden Increase of Voltage
The classical example is the sudden steep rise of amplitude in a
classical petit mal absence with 3/sec spike waves. There is
hence considerable variability in the ictal EEG events. This has
been quite discouraging for those concentrating on automatic
devices for seizure detection. Even the eye of the experienced
electroencephalographer may have difficulties in the determi-
nation of ictal episodes.
Ultraslow and Ultrafast Frequencies
These two aspects of paroxysmal activity are playing an increas-
ing role in our understanding of epileptiform EEG activity, and
are discussed separately within this volume.
Problems of Terminology
Definitions of EEG events are indispensable, but they are also
quite difficult to arrive at and are often unsatisfactory. Attempts
to reassess the EEG terminology have been made from time to
time; they often result in neologistic construction of terms that
fail to receive general acceptance. The EEG terminology is filled
with widely used, popular, and, alas, often quite sloppy and
inaccurate terms.
A preliminary proposal by the International Federation of
Societies for Electroencephalography and Clinical Neuro -
physiology (4) was followed by a definitive proposal (5) and a
glossary of terms used by clinical electroencephalographers (6).
The handbook catalog of Dutertre (7) exudes Cartesian lucid-
ity and deserves admiration. Unfortunately, the definition of
electrophysiologic events can never do justice to all facets of the
phenomenon.
As to epileptic discharges, there has been considerable dis-
unity in the ranks of electroencephalographers. The term
epileptic discharge has been attacked and condemned on the
ground that such discharges may occur in the absence of clini-
cal seizure manifestations or in individuals who have never had
seizures. The same is true for the widely used term seizure dis-
charge; even paroxysmal discharge, certainly a more cautious
and unassuming term, has not found general acceptance.
541
542 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
INTERICTAL EPILEPTIFORM DISCHARGES
Spikes (Single or Random Spikes)
According to IFSECN (6), a spike is a transient, clearly distin-
guished from the background activity, with pointed peak at
conventional paper speed or screen display and a duration from
20 to under 70 msec; the main component is generally negative
(Fig. 27.1). Amplitude is variable.
The distinction from the background activity is based on
wave morphology and amplitude. In many cases, spikes stand
out against the background because of their high voltage. If the
voltage of spikes and background activity is approximately
equal, the faster character (i.e., the shorter duration) of the
spike is its distinctive feature. It is possible that a spike of
50-msec duration and moderate amplitude may be embedded
in 20/sec activity (50-msec wave duration), for instance, in a
person with epilepsy with considerable drug-induced fast activ-
ity. Under these circumstances, the spike activity may be unde-
tectable. In such a case, a fast paper speed or widened screen
display could demonstrate the morphologic features of the
spike (multiphasic and more pointed character with a domi-
nant negative phase) in contrast with the more monotonous
appearance of the fast waves.
Spikes have many characteristics in common and yet there
are also remarkable inter- and intraindividual variations and
inconsistencies between one spike and the next in the same lead
(8). Spikes appear to be hypersynchronous events due to exces-
sive simultaneous neuronal discharge. This view, however, is
not congruent with the desynchronized (low-voltage) EEG
aspect of “electrodecremental seizures” or the electrode-
incremental initiation of focal or generalized attacks.
Definitions such as “epilepsy is paroxysmal hypersynchrony”
(9) may be correct on the neuronal level but partly incorrect
from the viewpoint of the “macro-EEG.”
Basic neurophysiologists have been cautioning against the
casual use of terms such as synchronization and desynchroniza-
tion. Low-voltage fast rhythms are not necessarily desynchro-
Figure 27.1 Various wave morphologies of spikes recorded from the
same patient (age 6 years). Upper lead, T3 T5; lower lead, T5 O1.
(First and second tracing from the top, continued in tracings 3 and 4,
respectively, 5 and 6.) Note multiphasic character of spikes with pre-
dominance of the negative phase (see Chapter 8) and occasional forma-
tion of double spikes.
nized and may be associated with enhancement and synchro-
nization within intracortical networks (10).
On the basis of long-term monitoring and computer analy-
sis, Gotman (11) has raised the question of whether interictal
spikes are in reality postictal. Gotman (11) and Gotman and
Marciani (12) observed marked and prolonged increase of
interictal spikes in a patient after a seizure (awake as well as
asleep). Drug levels had little influence on the rate of spike
activity, and there was no increase of spiking prior to a seizure.
The multiphasic character of a single spike warrants particu-
lar emphasis. A sequence of a minor positive, a major negative,
and a second minor positive component (Fig. 27.1) is typical in
most instances. A slow negative component may trail the spike
discharge and often attain about the same amplitude as the neg-
ative main component of the spike. This trailing slow component
of a single spike should not be regarded as evidence of a spike-
wave complex. Dutertre (7) in his handbook catalogue shows
examples of single spike-wave complexes that one could just as
well describe as compounded or multiphasic single spikes. From
the vantage point of basic neurophysiology, there is some reason
to presume that the trailing slow wave is caused by the same type
of hyperpolarization that was demonstrated by Pollen et al. (13)
in the experimental analysis of the spike-wave discharge.
The electroencephalographer very seldom finds single spikes
on the scalp with predominant positive component (14,15).
Positive single spikes are more common in depth recordings; on
the scalp, predominant positivity of single spikes raises the ques-
tion of defective superficial cortical laminae. Following surgery
of cortical arteriovenous malformations (AVMs) or after trau-
matic laceration (16), such positive spikes may be present occa-
sionally. A special type of positive spiking over the vertex has
been described by Bergen (17) in a case that combined features
of sphingolipidosis and mucopolysaccharidosis. (Also see the
work of Engel et al. (18) on the cherry-red spot-myoclonus syn-
drome.) With dipole formation (especially in benign rolandic
epilepsy), a positive spike discharge may be found in a moderate
distance from the negative spike focus (19,20).
Spikes represent the basic element of paroxysmal activity in
the EEG. A unitarian view that all spikes mean a hidden or overt
paroxysmal event would be erroneous. The fine semiology of
spikes is extremely important and the EEG interpreter ought to
consider the following questions:
1. What is the precise wave morphology?
2. Where do the spikes occur?
3. What is the patient’s age?
4. What is the state of awareness?
5. Is there any possibility of an artifact of similar appearance?
6. Is there any possibility of a physiologic potential of similar
appearance?
Discrepancies between spikes and behavioral epileptic events
are not uncommon at all. This highly complex situation has
been reinvestigated by Niedermeyer (21).
Wave Morphology
The largest and most pronounced spikes are not necessarily
associated with more serious epileptic seizure disorders. On the
 Chapter 27 ■ Epilepsy in Adults and the Elderly 543

contrary, rolandic spikes in a child aged 4 to 10 are very promi-

nent; however, the seizure disorder is usually quite benign or
there may be no clinical seizures at all.
On the other hand, “small sharp spikes” are indeed marginal
spike discharges and will be discussed separately. A very
unusual type of rhythmically repetitive and yet interictal spike
activity in two epileptic brothers was described by Bauer and
Markoff (22).

Spatial Distribution
In childhood, occipital spikes are, in general, the most benign
spike discharges, with less than 50% having clinical seizures;
rolandic central–midtemporal–parietal spikes are also quite
benign, whereas frontal spikes or multifocal spikes are more
epileptogenic.

Age
The significance of the age factor is enormous. From the spikes
of an epileptic newborn to a seizure focus of old age, age-
determined varieties of spikes can be distinguished.

Level of Awareness
Random spikes may occur at any state of waking–sleeping cycle
and occur even in rapid-eye-movement (REM) sleep when
bilateral-synchronous bursts of spikes or spike waves are usu-
ally suppressed.
Relationship to Perceptual Function
Figure 27.2 Coexistence of rolandic spikes and physiologic vertex
waves in light sleep (8-year-old boy, attacks of abdominal pain and no
proven epileptic seizures). Right centroparietal spikes with occasional
spread to the left are marked X; typical examples of a vertex wave are
marked O.

Shewmon and Erwin (23,24) investigated the impact of spikes

located over the occipital region and its vicinity on visual per-
ception and reaction time in three subjects with posterior
spikes. Their work shows that the visuomotor reaction time was
prolonged (or the response was absent) when the task was
spike-locked. Furthermore, “spike-induced dysfunction was
most pronounced when either response hand or visual field of
stimulus was contralateral to the spike” (24). In a further study,
Shewmon and Erwin (25) emphasized the significance of the
trailing slow wave that follows a single spike discharge. The per-
ceptive disturbance was particularly pronounced when the task
was presented during the slow wave. These fascinating findings
indicate that even single spikes of rapid duration may be capa-
ble of momentary impairment of the cortical visual function.
This applies to visual tasks. During an ictal–subclinical episode
of repetitive fast spiking over the occipital area and its neigh-
borhood, correct mental arithmetic is possible (26).
Distinction from Similar Physiologic Patterns
This differentiation is particularly important in the case of ver-
tex sharp waves during deep drowsiness and stage 2 of light
non-REM sleep. In childhood (after age 4), these waves may
have a particularly spiky appearance and may be misinterpreted
as paroxysmal spikes.
Physiologic sharp or spiky vertex waves are usually quite eas-
ily distinguished from rolandic spikes in sleep (Fig. 27.2). Even
in the more uncommon case of paroxysmal spike discharges
over the vertex, the differentiation from physiologic vertex
sharp waves is not difficult.
The physiologic nature of the occipital lambda waves and
“lambdoid” activity (positive occipital transients of sleep) must
also be considered. The main component of this pattern is positive.
Distinction from Artifacts of Similar Appearance
This distinction depends on the electroencephalographer’s
experience and is usually an easy one.
Automatic Spike Detection
Since the pioneering work of Gotman et al. (27) considerable
controversy has arisen regarding computer-based automatic
detection of spikes in the EEG. The pivotal question is “What
should the computer imitate?” (28), and this question hinges on
the precise definitions of spikes (29): “A spike detected by any
reader is a spike, a spike detected by the majority of readers is a
spike, a spike detected by the weighted average of readers is a
spike, etc.” According to these authors, average spike attributes
are calculated, and the resulting database can serve as a “gold
standard” for testing computer algorithms or other readers.
With the use of an “off-line” artificial neural network, attempts
have been made at real-time detection of spikes (30).
Are Interictal Spikes Nothing but EEG Signals?
Our earlier view (found in this chapter in previous editions)—
namely that single spikes are an EEG signal and nothing else—
appears to be in need of some correction. This seems to be
particularly true for “negative myoclonus” (ictally suppressed
motor activity). Beautiful examples have been presented by
544 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 27.3 Recurrent sharp wave recorded from a 61-year-old patient.
Upper lead, F3 C3; lower lead, C3 P3. A prominent negative phase is
followed by a large and relatively slow trailing positive wave.
Gambardella et al. (31) and Werhahn and Noachtar (32) (also
see Ref. 33). A single spike may also cause changes in cerebral
blood flow and increase of metabolic demands (34), demon-
strated with the simultaneous use of EEG and functional mag-
netic resonance imaging (fMRI). Even behavioral changes with
impaired test performance may be associated with single inter-
ictal spikes (35,36). Further information is found in the work of
Fisch (33).
Sharp Waves
According to IFSECN (6), a sharp wave is a transient, clearly
distinguished from background activity, with pointed peak at
conventional paper speed or screen display and duration of 70
to 200 msec, that is, more than approximately 1/14 to 1/5 sec-
onds. The main component is generally negative (Fig. 27.3).
Jasper (37) pointed out that the rising phase of the sharp
wave is of the same order of magnitude as in spikes, but the
descending phase is prolonged. The configuration with a
steeper ascending phase, however, is not always present.
A very unusual type of repetitive slow sharp wave activity
has been reported in prematurely born infants with intraven-
tricular hemorrhage; these discharges show predominantly
positive polarity and are recorded mainly over the rolandic
region (38,39). The maximum of this activity is sometimes
found over the vertex.
Spikes and sharp waves are neurophysiologically closely
related phenomena; both of them are typical paroxysmal dis-
charges and highly suggestive of an epileptic seizure disorder,
although both phenomena may occur in patients without a his-
tory of seizure disorder.
Sharp waves are usually found as random focal discharges;
most anterior temporal spikes are, in a strict sense, sharp
waves. This is also true for most benign rolandic spikes of
childhood. Sharp waves are more seldom found in generalized-
synchronous bursts where single spikes, spike waves, and poly-
spikes predominate.
It has been contemplated that sharp waves on the scalp cor-
respond with spikes in the depth or on the cortex. Combined
depth and scalp recording clearly refutes this view (40). One can
detect spikes as well as sharp waves in depth recordings; a deep
sharp wave usually corresponds with a sharp wave on the scalp
if there is any corresponding scalp activity at all.
The long duration of a sharp wave permits better insight
into the multiphasic character of this potential. A small preced-
ing positive component may be very fast and qualify as a spike;
even a small biphasic spike discharge may precede the much
larger sharp wave. Some sharp waves even exceed the maximum
length of 200 msec of the IFSECN definition (6) (Fig. 27.4), also
Figure 27.4 An example of a very slow sharp wave (blunted sharp
wave), maximal over T6, obtained from a 6-year-old boy with cranio-
cerebral trauma at age 4 followed by seizures.
named “slow sharp waves” (41) or, in cases of particularly long
duration, “blunted sharp waves” (Fig. 27.4). Others become
very complex. They consist of a constantly varying number of
components; such compounded sharp waves may occur as the
periodic discharges (periodic lateralized epileptiform dis-
charges [PLEDs]) of Chatrian et al. (42).
It is certainly not incorrect to use the terms spikes and sharp
waves synonymously when a local paroxysmal event is dis-
cussed, although purists of nomenclature would regard this as
a breach of etiquette.
Polyspikes or Multiple Spikes
This discharge type represents a complex of spikes and may also
be called polyspike complex. In modern terminology (6), the
term multiple spike complex is preferred on the grounds of lin-
guistic considerations, because “polyspikes” is an etymological
hybrid. It has been defined as a complex paroxysmal EEG pat-
tern with close association of two or more diphasic spikes
occurring more or less rhythmically in bursts of variable dura-
tion, generally with large amplitudes (Figs. 27.5 and 27.6) (7).
Polyspike bursts are readily elicited by electrical stimulation
of single depth leads, especially in limbic regions. On the scalp,
however, polyspikes occur mostly as bilateral- or generalized-
synchronous discharges. Exceptional focal polyspikes are occa-
sionally encountered; these usually have a frontal maximum,
except for occipital accentuation in hypsarrhythmia. Polyspikes
and also polyspike-wave complexes are sometimes associated
with concomitant myoclonus, especially in primary generalized
epilepsy and in photosensitive individuals with this type of
seizure disorder. Children with Lennox–Gastaut syndrome may
also show association of polyspikes and myoclonus.

Figure 27.5 Polyspikes in light nonrapid-eye-movement (REM) sleep,
especially over central region, associated with K complexes. (Reproduced
with permission from AMA Archives of Neurology.)
Runs of Rapid Spikes
This pattern has been described as “grand mal discharge”
(43,44), “fast paroxysmal rhythms” (45), “rhythmic spikes” (46),
and “generalized paroxysmal fast activity” (47). It is seen only in
sleep and occurs in older children, adolescents, and younger
adults (Fig. 27.7). It consists of bursts of spike discharges at a
rate from 10 to 25/sec, usually generalized but with a well-
defined maximum over frontal regions; it may even be confined
to the frontal leads. The voltage is in the medium to high range,
often exceeding 100 or even 200 V. The discharge rate is in
most cases somewhat irregular. The bursts last for about 2 to
10 seconds; bursts of more than 5-second duration are usually
associated with tonic seizures and thus represent an ictal pattern.
The obvious misnomer “grand mal discharges” is based on
certain similarities with the ictal EEG of a generalized
tonic–clonic seizure. In a patient population with such seizures,
runs of rapid spikes as an interictal pattern are a very rare find-
ing. This has been confirmed by the data of Chayasirisobhon et
al. (48). The inappropriateness of the term grand mal discharge
was stressed by Rodin et al. (49), who thought that this pattern
occurred in patients with primary generalized epilepsy who suf-
fered from more than one type of seizure and especially from
akinetic seizures. As already noted by Gastaut et al. (46), this
Figure 27.6 The differentiation between polyspikes and drug-induced
fast activity may be difficult. The run of beta waves (center) certainly falls
into the category of fast activity, but the more outstanding and burst-like
run (right) is more likely to represent polyspike activity. A definitive
statement depends on the impressions derived from the entire record.
Chapter 27 ■ Epilepsy in Adults and the Elderly 545
Figure 27.7 A run of rapid spikes in a 19-year-old epileptic patient
(Lennox–Gastaut syndrome). Note anterior maximum of the discharge.
A few slow spike-wave complexes are also seen (right temporo-occipi-
tal region).
discharge is very typical in patients with Lennox–Gastaut syn-
drome; it is hardly ever found outside this clinical entity
(50–52). It may be found in very rare instances of posttrau-
matic epilepsy with EEG traits of the Lennox–Gastaut syn-
drome (53). Brenner and Atkinson (47), however, insist that
runs of rapid spikes (“generalized paroxysmal fast activity”)
may be found outside the Lennox–Gastaut syndrome. There is
evidence that runs of rapid spikes may occur in all “imitators”
of the Lennox–Gastaut syndrome (54).
We know little about the neurophysiologic underpinnings of
this pattern. It is clearly a frontal lobe discharge, and the associ-
ated tonic seizures are probably arising from the premotor and
supplementary motor portions of the frontal lobes. There could
be certain similarities to the decremental seizure patterns, which
also originate most commonly from the frontal lobe (1).
Unfortunately, those electroencephalographers who never
use sleep portions in routine records will never see this fascinat-
ing pattern, which is definitely in need of further basic research.
Three-Per-Second Spike-Wave Complexes
Classical 3/sec spike-wave complexes are widely known even
outside the community of electroencephalographers. The dis-
charge is officially termed spike- and slow-wave complex (6); this
term comprises all types of spike-wave complexes, which are
listed separately because of markedly differing associated clini-
cal-epileptologic conditions. The official definition (6) is quite
simple: “A pattern consisting of a spike followed by a slow
wave.” Older terms such as spike-and-dome complex, wave-
and-spike complex, and dart-and-dome complex are seldom used
these days and should be discouraged. The abridged term 3/sec
spike wave is certainly acceptable, because the omitted word
complex is automatically implied.
Spike-wave complexes of the classical variety (3/sec) have been
described as the EEG pattern of petit mal absences (Fig. 27.8) (44).
546 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 27.8 Three-per-second generalized synchronous spike-wave
complexes. There is a clear voltage maximum over the frontal regions
bilaterally. The spike component sometimes becomes less evident
toward the end of a run of discharges, and there can be a slight slow-
down of the usual frequency as well.
Since that time, the 3/sec spike-wave discharge has been equated
with absence epilepsy. Regrettable mistakes have occurred due to
lack of communication between electroencephalographers and
clinicians who initiated treatment with anticonvulsants specific
for absence seizures due to the use of the term 3/sec spike-wave
complexes alone. The term petit mal discharge (43) has reinforced
this type of reflex thinking. Fortunately, electroencephalographers
and epileptologists have always cautioned against the misinterpre-
tation of the term petit mal discharges; this should be interpreted
as “discharges that occur in petit mal absences but also in other
conditions.” The works of Silverman (55), Clark and Knott (56),
and Lundervold et al. (57) have clearly shown that the spike-wave
discharge as an interictal event correlates with the occurrence of
absence seizures only in a moderate percentage of patients (16%
according to Clark and Knott).
The 3/sec spike-wave discharge is a worthwhile subject for
more detailed discussion. As to wave morphology, there is good
evidence that the spike-wave complex is not simply an association
of a spike and a slow wave, although the above-mentioned defini-
tion is justifiable as a simplification. The spike-wave complex con-
tains hidden components that can be seen even with simple visual
analysis at 30 mm/sec paper speed or equivalent screen display;
special recording methods will reveal these extra components
much better. Cohn (58) and Weir (59) have elucidated the com-
ponents of this pattern. Most important is a spike of positive
polarity during the descending portion of the slow wave (59),
becoming evident as notching under usual recording conditions.
On the basis of experimental studies of cortical faradization,
Jung and Toennies (60) stressed the inhibitory character of
the slow wave following a spike discharge (“Bremswelle”).
Experimentally induced spike-wave-like activity in cats, elicited
by electrical stimulation of thalamic intralaminar structures, was
studied with intracellular cortical microelectrodes (13,61,62).
According to Pollen (61), all components of the spike-wave com-
plex were the “consequence of postsynaptic activity.” This author
points out that “the initial positivity which may precede a nega-
tive spike (depending upon the intensity of the stimulation)
results from depolarization of middle and deep lying neurons,
whereas the surface negative spike is the result of EPSP [excita-
tory postsynaptic potential] generation primarily upon apical
dendrites of vertically oriented pyramidal type neurons.”
According to Pollen (61), the long surface negative wave is gener-
ated by inhibitory postsynaptic potentials (IPSPs) predominantly
on or near the soma of similar pyramidal cells. Thus, the slow
wave indeed appears to be related to inhibitory impulses.
From this vantage point, the spike-wave discharge apparently
represents an alternating succession of excitation and inhibition.
The clinical–ictal activities are thus constantly curbed by interven-
ing inhibitory impulses that prevent the attack from progressing
into massive downward discharges with motor effects (polyspikes
with massive myoclonus) or into a full-blown tonic–clonic attack.
For this reason, motor manifestations of ictal episodes character-
ized by spike waves are almost always inconspicuous or modest.
Furthermore, spike-wave bursts rarely proceed into generalized
convulsions; such rare exceptions have been demonstrated by
Halász (63) and Niedermeyer (64).
The distinction between classical (3/sec), slow (2 to 2.5/sec),
and fast (4/sec) spike-wave complexes and the smaller 6/sec
spike-wave discharge is justified on the basis of different
clinical-epileptologic correlates of each spike-wave type; it is
most definitely not an example of electroencephalographic hair
splitting. The classical 3/sec spike-wave discharge is most typi-
cal and most pronounced in children with absence seizures.
Clinical absences are usually present when the burst lasts for
longer than 5 seconds. Thus, shorter bursts are usually subclin-
ical, but numerous psychophysiological attempts have been
made to demonstrate certain fluctuations of level of awareness
even in apparently subclinical spike-wave bursts. The classical
spike-wave complex is, in most cases, easily activated by hyper-
ventilation, whereas the slow and the fast forms of this dis-
charge show little or no enhancement.
A classical 3/sec spike-wave burst does not run exactly at a
rate of 3/sec. The complexes are faster at the onset of the burst
(mostly around 4/sec), and then slow down to 3.5 and 3/sec for
the main portion, and eventually slow to 2.5/sec at the end of
the burst. During a burst, the spike discharges become gradu-
ally smaller (Fig. 27.8), often shrinking to insignificance as in
drowsiness and sleep.
The spatial distribution of the bursts is very typical; the maxi-
mum almost always lies over the frontal midline region, whereas
the minimum is found over temporal and occipital areas.
Because somewhat different findings obtained by Dondey (65)
were based on EEG recordings in which no midline electrodes
were used, they do not invalidate the concept of a frontal mid-
line maximum, which has been confirmed by Rodin and
Ancheta (66) with the use of a brain mapping technique. The
understanding of this distribution type is highly conducive to a
better comprehension of the underlying mechanisms. Almost
all spike-wave bursts are bilateral synchronous or generalized

synchronous; the synchrony, however, is not perfect (58,67). The
age of distribution lies mainly in the range from 4 to 16 years.
The work of Lemieux and Blume (68) has shed more light
on the spatial distribution of the spike-wave complex on the
basis of computer analysis. According to these authors, field dis-
tribution of spikes differed from that of the ensuing slow waves.
Spikes “arose in one frontal region and propagated to the
homologous part of the other frontal region with a peak-to-
peak interval less than 15 msec. Less commonly, spikes first
moved anteriorly within the initiating frontal region before
contralateral propagation occurred.” By contrast, negative slow
waves were more diffuse, more symmetrical in evolution, and
more posteriorly centered than the spikes.
The onset of the spike component in one frontal lobe is con-
gruent with the earlier findings of Cohn (58) and Lüders et al.
(69), who noted asynchronies ranging from 5 to 20 msec. The
frontal lobe is a large structure and subdivisions are necessary.
Onset of spike-wave activity (and, in particular, of the spike
component) over the frontal midline means in reality that a
frontal portion in the immediate vicinity of the interhemi-
spheric fissure is the origin of the discharge; this area belongs to
the supplementary motor zone.
As to the neuropsychological correlates of subclinical general-
ized spike-wave bursts and clinical absences with spike waves, it
should be pointed out that the impact of apparently subclinical
3/sec spike-wave bursts might be stronger than one would expect
from routine clinical observation. With its maximum over the
frontal midline (frontal lobe close to interhemispheric fissure),
there might be some impact on prefrontal lobe functions and, in
particular, on working memory. The concept of working mem-
ory, introduced by Baddeley (70) and enormously refined by
Fuster (71,72), implies a system of constant perceptory afferences
to the prefrontal dorsolateral region whence motor impulses
emanate, resulting in readiness (set) for motor action. This sys-
tem could be blocked for a moment by a few spike-wave com-
plexes (or even by a single one) and interfere with working
memory. In a clinical 3/sec spike-wave absence, blocking of
working memory might be the most likely explanation for the
mild disturbance of consciousness and its immediate full recov-
ery with the termination of the attack (73). No “true disturbance
of consciousness”—epileptic or nonepileptic—can recover
immediately. In other words, the function of working memory is
simply suspended for the duration of the spike-wave complexes.
Due to the uniqueness of primary generalized epilepsy with
3/sec spike-wave complexes in the human (74), animal models
for this response have great limitations. The arrest reaction of
the monkey (75) produced by alumina cream injection into the
fronto-orbital cortex has some resemblance to the human
absence epilepsy. However, this cannot be said about the sudden
freeze of motility noted in the cat stimulation of the cat’s mesial
anterior thalamus (76), since there is no spike-wave activity
associated with this type of arrest reaction.
According to Sperling and Skolnick (77), there is a general
decrease of cerebral blood flow (by 133 xenon method) during
human generalized spike-wave activity, which, however, is less
pronounced in frontal regions than in the parietal lobes.
Transcranial magnetic stimulation of the motor cortex in
Chapter 27 ■ Epilepsy in Adults and the Elderly 547
patients with primary generalized epilepsy and 3/sec spike-
wave complexes produced motor evoked potentials of signifi-
cantly reduced size when the cortical stimulus was time-locked
to the slow-wave component of the spike-wave complex (but
slightly reduced or unchanged when the stimulus was time-
locked with the spike) (78). This underscores the special
inhibitory character of the slow-wave component, which has
been known since the work of Jung and Toennies (60).
Slow Spike-Wave Complexes
After the demonstration of the 3/sec spike-wave complex and
its relationship to the petit mal absence (44), the Gibbses and
William G. Lennox were struck by the occurrence of slow spike-
wave complexes in a much different type of patients with
seizures other than absence. The classical 3/sec spike-wave com-
plex was termed petit mal discharge. Consequently, the slow
spike-wave complex was termed petit mal variant discharge.
This term was used first by Gibbs et al. (79). In that study, hypo-
glycemia was induced in order to demonstrate an increase of
classical 3/sec spike-wave activity in the hypoglycemic state; an
increase of slow spike-wave activity, however, was not elicited
with this activation method when used in appropriate patients.
The wave morphology of this pattern varies considerably. In
the majority of the cases, the complex consists of a rather slow
spike (according to the definition, a sharp wave, lasting 70 msec
or longer) and a slow wave. In a sizable number of cases, true
spikes (60 msec or less duration) are followed by a slow wave
(Fig. 27.9).
The spatial distribution is, in most instances, quite similar to
that of the 3/sec spike-wave complex. In the vast majority of the
cases, the bursts are bilateral or generalized synchronous (with
imperfect synchrony); a frontal midline maximum is the rule in
these discharges. Lateralization or occasional focal occurrence
is sometimes observed, usually in children with severe residual
brain damage in certain areas where parenchymatous destruc-
tion abolishes the spike-wave discharge.
Figure 27.9 Generalized slow spike-wave complexes (mostly around
2/ sec) in a child with severe epileptic seizure disorder (Lennox–Gastaut
syndrome).
548 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Generalized slow spike-wave discharges are often quite pro-
longed. In some children or adolescents, the entire sleep por-
tion (light and moderately deep non-REM sleep) consists of
unabated generalized slow spike-wave activity. The diagnosis of
an “electrical status epilepticus” may or may not be justified in
such cases; there are usually no behavioral or polygraphic char-
acteristics to suggest an ongoing ictal event. In some cases, the
slow spike-wave activity may be found only in non-REM sleep.
Although the classical 3/sec spike-wave complex is seldom
seen before the age of 4 and almost never before 3.5 years, its
slow counterpart appears much earlier, sometimes before the
age of 6 months. At this early age, the frontal maximum may
not be readily demonstrable.
The slow spike-wave complex is almost always associated
with a severe and often uncontrollable type of childhood
epilepsy (seldom with onset between ages 11 and 20 years)
called Lennox–Gastaut syndrome (46,50,52,79,80). Most of
these children show a variety of minor attacks and evidence of
mental retardation.
Thus, the distinction between slow (1 to 2.5/sec) and classical
(3/sec) spike waves is of great clinical significance. One has to
keep in mind, however, that many children or adolescents with
slow spike waves also show series of 3/sec or even 4/sec spike-
wave complexes; the presence of these spike-wave types is obvi-
ously of no significance in such cases and what counts is the slow
type. On the other hand, patients with classical 3/sec spike-wave
complexes and a usually benign type of epileptic seizure disor-
der almost never show slow spike waves except when a 3 or 4/sec
spike-wave burst may slow down to 2.5/sec at the end.
observed transition from the former to the latter with deepening
drowsiness. Marshall (83) introduced the term wave and spike
phantom, which has found only limited acceptance.
The clinical correlates of 6/sec spike-wave complex have been
elucidated by Thomas (84), Gibbs and Gibbs (85), Hughes et al.
(86), and Thomas and Klass (87). It is usually a pattern of adult-
hood but may also occur in adolescents and children. About
50% to 60% of the patients suffer from clear epileptic seizures
(mostly generalized tonic–clonic); the remainder show a history
of syncopal attacks, posttraumatic states, or (as emphasized by
Small (87)) psychiatric problems.
The 6/sec spike-wave complex is an uncommon but not rare
pattern (about 0.5% to 1% in a central EEG laboratory). The
discharge may be recorded in waking state, drowsiness, and
light non-REM sleep; light drowsiness appears to be the opti-
mal recording condition (Fig. 27.10).
Investigations regarding the spatial distribution of this pat-
tern have led to an important dichotomy. Hecker et al. (88)
pointed out that a distinction must be made between frontal
accentuation and occipital accentuation. We are evidently deal-
ing with two different types of discharge; the frontal type is
most commonly associated with epileptic seizure disorders and
sometimes found in combination with other paroxysmal dis-
charge types. The occipital types are found predominantly in
patients with no evidence of epileptic seizure disorder.
In a study of 839 patients with 6/sec spike-wave complexes,
Hughes (89) placed particular emphasis on the distinction
between frontal and occipital maximum. He summarized the

Fast Spike-Wave Complexes

This pattern is closely related to the classical 3/sec spike-wave
complex; both discharge types are most commonly lumped
together. A differentiation, however, is justified on clinical
grounds.
Gastaut (81) deserves credit for an outline of the distinctive
features between the 3 and the 4/sec (or 4 to 5/sec) spike-wave
discharge. According to his investigations, the fast spike-wave
burst is usually of shorter duration (1 to 3 seconds), it occurs
in patients older than 15 years, the bursts are always subclinical,
and the associated seizure disorder is usually characterized by
myoclonic jerking, grand mal attacks, or a combination of both
seizure types, whereas petit mal absences are quite uncommon.
Paroxysmal flicker responses are common in such patients. A
positive family history of epileptic seizures disorder is frequently
found in this group of patients. The 4/sec spike-wave discharge is
spatially distributed in the same manner as the 3/sec spike-wave
discharge; the frontal midline maximum is quite prominent.

Six-Per-Second Spike-Wave Complexes

This is a marginal paroxysmal pattern, described by Gibbs and
Gibbs (43) as follows: “The spike in this pattern has a strong
positive component but the entire wave complex looks in gen-
eral like a miniature reproduction of the 3 per second spike and Figure 27.10 A short run of 6/ sec spike waves, posterior type,
wave of petit mal.” In view of the positive component of the recorded in a 52-year-old woman with a history of head injury 2 years
spike, Silverman (82) has stressed similarities to the 6/sec com- earlier and subsequent headache, dizziness, and memory loss. There
ponent of the 14 and 6/sec positive spike discharge; he also was computed tomography (CT) scan evidence of cortical atrophy.

principal features of both forms by the acronyms WHAM
and FOLD:
W: waking record F: females
H: high in amplitude O: occipital
A: anterior L: low in amplitude
M: males D: drowsy record
An interesting sidelight of this observation is the uneven sex
distribution, which is rather unique; no other EEG abnormali-
ties show any gender preference. According to Hughes (90),
patients with anterior 6/sec spike waves are more likely to have
epileptic seizures, and those with posterior discharges tend to
have neuroautonomic disturbances.
Kocher et al. (91) were struck by the occurrence of 6/sec
spike-wave complexes in the abstinence or withdrawal phase of
drug-dependent individuals. Tharp (92) produced the dis-
charge by intravenous diphenhydramine (50 mg) in normal
subjects. Hecker et al. (88) found that the occipital form of the
6/sec spike-wave complex is often related to drug dependence
(hypoanalgesics and barbiturates) and withdrawal. All this
underscores the need for a distinction of the two forms. Thus,
the frontal form will have to be regarded as a truly epilepsy-
related paroxysmal pattern, whereas the occipital form finds
itself on the fringe of paroxysmal discharge types.
INTERICTAL EPILEPTIFORM DISCHARGES
IN OTHERWISE HEALTHY INDIVIDUALS
The occurrence of focal or generalized paroxysmal discharges
in apparently healthy individuals is a puzzling and even annoy-
ing finding that requires some discussion. The EEG evaluation
of comparatively large healthy populations usually shows a cer-
tain percentage of abnormalities. The work of Thorner (93)
showed paroxysmal discharges in 0.3% of 1100 flying cadets;
Gibbs et al. (94) found epileptic abnormalities in 0.9% of nor-
mal adult controls, whereas Harty et al. (95) observed such
abnormalities in 3% of candidates for medical service. “Larval
epileptic disturbances” were noted in 9% of healthy controls by
Williams (96). The tumult of World War II was apparently con-
ducive to such large-scale studies, which have become more
scarce over the past decades.
A study of 682 Air Force applicants showed paroxysmal
changes in 2.6%; about one fourth were focal and the rest were
of nonfocal character (97). Further studies of this kind were
carried out in commercial flying personnel (98), candidates for
commercial airplane pilot training (99), and air cadets (100).
Some studies concentrated on the classical 3/sec spike-wave
pattern, which was found in 0.2% of 3070 normal controls (94)
and 2.7% in the subjects of Hill (101). Spike-wave complexes
were found in 0.9% to 1.5% of the groups studied by Dell and
Dell (98), in 1% of the subjects of Blanc et al. (99), and in 0.5% of
O’Connor’s subjects (100). No spike-wave complexes were found
in the normative study carried out for the National Aeronautics
and Space Administration by Baylor University in Houston,
Texas (102). Bennett (103) evaluated the EEG of military and
civil flying personnel (1332 persons) and found spike-wave
Chapter 27 ■ Epilepsy in Adults and the Elderly 549
discharges in only 0.6%. Gibbs and Gibbs (85) found 6/sec
spike waves in 1.3% to 2.0% of 3476 adults, peaking in the
range from 20 to 24 years, but were unable to detect any focal
spikes or sharp waves. Focal sharp activity was reported in 1%
of the material of Blanc et al. (99). The reports of Lennox-
Buchthal et al. (104) and Kitamura and Asakura (105) show a
relatively high incidence of spikes (6.4% and 4%, respectively).
Zivin and Ajmone Marsan (106) and Chatrian (107) have
contemplated the clinical significance of spikes in healthy per-
sons. Above all, the interpretation must take into consideration
age. In childhood, the occurrence of central–midtemporal (also
parietal) spikes is associated with overt seizures in only 50% to
70% of the cases; this pertains mainly to the age from 3 to 12
years. In occipital spikes (mainly age 3 to 5 years), the epilepto-
genicity is even lower. Major studies on this subject based on
large healthy populations were performed by Nekhorocheff
(108), Kellaway and Fox (109), Brandt and Brandt (110), Corbin
and Bickford (111), Trojaborg (112), Eeg-Olofsson et al. (113),
and Cavazzuti et al. (114). In general, “benign” focal spikes (such
as seen in benign rolandic epilepsy) outnumber generalized-
synchronous bursts of spikes or spike waves by a slight to con-
siderable margin. Cavazzuti et al. obtained EEG recordings of
3726 children from 6 to 13 years of age without evidence of
seizure or neurologic deficits. Paroxysmal discharges were found
in 131 cases (3.52%). Generalized, mainly polyspike, discharges
were found in 41 cases, midtemporal spikes in 50, centroparietal
spikes in 27, occipital spikes in 2, and bilateral (midtemporal or
parietal) spikes in 11 children. Change of spike localization
occurred in but a few subjects; shifts from focal to generalized
spiking and vice versa were also rare. The work of Cavazzuti et
al. is particularly important because it includes follow-up peri-
ods of up to 9 years. In most children, the abnormalities disap-
peared, usually during elementary school age or in early
adolescence. Only seven children developed clinical epileptic
seizures—five had generalized discharges, one had midtempo-
ral, and one had centroparietal spikes. In two subjects there was
evidence of epileptic seizures in the family, while six children
had siblings with spikes, either generalized or rolandic.
Both generalized-synchronous (spike wave and polyspike
wave) and rolandic (centroparieto-midtemporal) spikes in chil-
dren without epilepsy suggest a genetic predisposition if no neuro-
logic deficit and no history of insult to the central nervous system
(CNS) are present. In children with a history of cerebral palsy and
with no seizures but prominent spiking, the spike activity may
herald future epileptic seizures (85). Even in perfectly healthy chil-
dren with spikes, the possibility of future seizures cannot be com-
pletely ruled out, although the chances are slim.
EPILEPTIFORM EEG FEATURES
OF UNCERTAIN SIGNIFICANCE
Small Sharp Spikes, or Benign Epileptiform
Transients of Sleep (BETS)
This pattern has been delineated by Gibbs and Gibbs (43,85). It
is the most inconspicuous paroxysmal discharge and hence is eas-
ily overlooked. (Only the occipital type of the 6/sec spike-wave
550 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

et al. (120), Klass and Westmoreland (121), and Gutrecht (122)

have stressed the nonspecific nature of this pattern and do not
regard its occurrence as an abnormality. By contrast, Low et al.
(123) as well as Hughes and Gruener (124) have pointed out
that small sharp spikes are not a normal or meaningless pattern.
The data of these authors clearly show that this pattern indi-
cates “a moderate degree of epileptogenicity” (124). This has
been strongly confirmed by Saito et al. (125) and also by Molaie
et al. (126).

Needle-Like Occipital Spikes of the Blind

Spike discharges of a particularly fast and needle-like character
develop over the occipital region in most congenitally blind chil-
dren (127–129). These spikes appear to be completely innocu-
ous, are unrelated to epileptic seizure disorders, and may be due
to a state of functional deafferentation (130). The discharges dis-
appear during childhood or adolescence. The experimental basis
of visual deafferentation spikes, however, is controversial (131).
Congenitally blind children with a history of retrolental
Figure 27.11 Small sharp spikes (51-year-old patient). Note the subtle
fibroplasia often show evidence of accompanying cerebral
character and moderate voltage of the discharge; also note its predom-
impairment. Such children may show abundant and wide-
inance in the left nasopharyngeal lead. There is evidence of spread into
spread but chiefly occipital spike activity; concomitant clinical
T3, as well as into the right nasopharyngeal lead. The left section was
epileptic seizures are fairly common in this condition (85,132).
recorded in the waking state (transition to earliest drowsiness); the mid-
This type of spike activity must be differentiated from transient
dle and right sections were recorded in sleep. (Reproduced with permis-
occipital spiking in nonepileptic congenitally blind children.
sion from Clinical Electroencephalography.)
Fourteen-and-Six Positive Bursts
complex is equally small and unimpressive.) The main negative This discharge pattern has been defined as follows (6): “Bursts
and positive components are of about equally spiky character of arch shaped waves at 13–17 Hz and/or 5–7 Hz, most com-
(Fig. 27.11). The discharge is fairly widespread and is seen chiefly monly at 14 and/or 6 Hz, seen generally over the posterior tem-
over temporal and frontal areas, either shifting from side to side poral and adjacent areas of one or both sides of the head during
or synchronously firing. The pattern is almost exclusively found in sleep. The sharp peaks of its component are positive with
drowsiness and/or light non-REM sleep. It is essentially a pattern respect to other regions.” It is also pointed out in this definition
of adulthood with a peak between ages 30 and 60 years; occur- that the amplitudes are generally below 75 V, that the pattern
rence in adolescence and old age is somewhat less common. It is is best demonstrated by referential recording using contralat-
virtually absent in the first 10 years of life. This pattern is also eral ear lobes or other remote reference leads, and that the clin-
known as benign epileptiform transients of sleep (BETS). ical significance is controversial (Fig. 27.12). The synonymous
Koshino and Niedermeyer (115) found a prevalence of
1.36% with a peak of 2.9% between ages 30 and 40 years. In this
study, two thirds (67.4%) of the patients had a history of epilep-
tic seizures; this underscores the paroxysmal character of the
discharge. Small sharp spikes sometimes are precursors of typ-
ical anterior temporal spikes or sharp waves, which, in such
patients, simply occur somewhat later in drowsiness or sleep.
This association was also noted by Gibbs and Gibbs (85). In a
depth EEG study by Westmoreland et al. (116), an extensive
generator area of small sharp spikes was found.
Among those without epilepsy, the discharges may occur in
patients with cerebrovascular disorder, syncopal attacks, and
psychiatric problems. Small (87) and Small et al. (117) have
placed special emphasis on the occurrence of this spike dis-
charge in a psychiatric population and especially in patients
with manic-depressive illness. Figure 27.12 Various examples of 14/ sec and 6/ sec positive spikes
It is interesting to note that Gibbs and Gibbs (85) found a (underlined). Note posterior predominance for this pattern and shifting
rather high prevalence of small sharp spikes in presumed asymmetrics. Also note the sometimes blurred distinction between the
normal adult control subjects, reaching 7.9% in the range from 14 and 6 components, due to notch formation. The recording was
40 to 49 years. Reiher and Klass (118), Lebel et al. (119), White obtained from a 12-year-old patient; montages to ipsilateral ear.

term ctenoids was introduced by Lombroso et al. (133) with
regard to the comb-like shape. The 14 and the 6 component
may be observed independently. According to Gibbs and Gibbs
(85), the 6 component prevails in early childhood and adult-
hood, while the 14 component is more prominent in older chil-
dren and adolescents. Hughes (134,135) emphasized the
harmonic character of the major frequencies and preferred the
term 14 and 7/sec positive spikes.
This pattern was described first by Gibbs and Gibbs (136)
and regarded as “evidence of thalamic and hypothalamic
epilepsy.” The following 15 years produced exciting correlations
to autonomic nervous system dysfunctions and behavioral dis-
orders (2,137–147). Enthusiasm flagged when it became clear
that this pattern is often found in healthy individuals (133,148).
Little (149) even presumed that this pattern “is more likely to be
a sign of health than of disease.”
The “14 and 6 pattern” may be confined to the regions lying
beneath a skull defect (150) and hence could denote a “hidden
pattern” that is not detectable from the scalp unless it reaches
major proportions.
The pattern occurs most commonly in children after age 4
and adolescents and declines in adulthood; Gibbs and Gibbs
(85,151) recorded the 14 and 6 activity even in infants. Its
occurrence in the waking state is exceptional; deep drowsiness
and very light non-REM sleep are usually the ideal states for the
documentation of the pattern, although deep sleep might be
more conducive in very young children (85). Yürüker and
Menzi (152) elicited the 14 and 6 pattern with auditory stimuli
presented in light sleep. Okada et al. (153) noted 14 and 6/sec
positive spikes in association with simultaneous negative spikes
over the frontal area (suggestive of dipole formation). Depth
recordings at the thalamic as well as the basal ganglia level (154)
have shown this neurophysiologic discharge, and it has been
repeatedly recorded in the amygdaloid complex as well.
Not surprisingly, the prevalence of the pattern appears to
increase with the recorded length of sleep. Much of the contro-
versy about this pattern has arisen from differences of record-
ing techniques and the use of routine sleep tracings and, in
particular, the length of the sleep portion. Proven cases of
epileptic seizure disorder very seldom show 14 and 6/sec posi-
tive spikes as the only significant finding, and it appears to be a
marginal pattern of no or very little epileptologic significance.
The demonstration of 14 and 6/sec positive spikes in
advanced states of metabolic encephalopathies, especially in
hepatic coma (155), is an interesting exception. For completely
unknown reasons, this essentially innocuous discharge may
appear when slow activity becomes extremely pronounced in
such patients. A similar observation was reported by Ford and
Freeman (156). Drury (157) reported 121 children with hepatic
coma and mostly in association with Reye syndrome. Seven of
these children showed 14 and 6/sec positive spikes, occurring in
an almost periodic manner, in four children activated by stim-
ulation. In all of these cases, there was a background of severe
EEG abnormality. Repeat tracings failed to show any 14 and 6
discharges (with one exception). The “14 and 6” pattern may
also occur as a transient toxic effect of diphenhydramine
hydrochloride.
Chapter 27 ■ Epilepsy in Adults and the Elderly 551
Psychomotor Variant (Rhythmic Midtemporal
Theta Discharges)
This pattern is widely known as “psychomotor variant dis-
charge” (43,85) because of certain similarities to rhythmic ictal
activity occurring in psychomotor seizures (temporal lobe
seizures and complex partial seizures). This term has been dis-
couraged by IFSECN (6), which recommended the term rhyth-
mic theta bursts of drowsiness. The term rhythmic midtemporal
discharge of Lipman and Hughes (158) is also widely used.
The pattern consists of long runs of rhythmic activity in the
range of 5 to 6.5/sec with a maximum over the midtemporal
region, often with considerable spread into posterior temporal,
anterior temporal, and occipital areas (Fig. 27.13). The theta
activity shows a well-defined negative sharp component that
apparently stresses the paroxysmal character of the discharge.
These trains of sharp theta waves may occur in a unilateral
shifting, bilateral shifting, or synchronous distribution type.
The duration of a single run usually exceeds 10 seconds and
may reach 1 minute or more. Very often, the first run is noted
in early drowsiness; in deep drowsiness and light sleep, the pat-
tern tends to disappear. According to Egli et al. (159), patients
with rhythmic theta bursts show shortened periods of REM
sleep and also, to a lesser degree, of slow sleep in nocturnal sleep
studies at the expense of long drowsy periods with the rhythmic
theta pattern. Hughes (160) has emphasized that the rhythmic
midtemporal discharge is not limited by the drowsy state and
may be very active in wakefulness.
The pattern occurs mainly in younger or middle-aged
adults; it is also seen in adolescents and children. It is a rare
pattern; Egli et al. (159) found a prevalence of 0.1% in a large
laboratory.
The clinical significance is not clear. Despite its strongly
paroxysmal appearance, its epileptogenic properties seem to be
very low; most patients have no history of clinical seizures,
although Lipman and Hughes (158) found seizures in 36% of
their patients. Personality disorders and some autonomic nerv-
ous system dysfunctions are common in patients with this pat-
tern (85,158,161). This view is shared by Eeg-Olofsson and
Figure 27.13 Psychomotor variant discharge of the right midtemporal
region (T4). Tracing was obtained from a 21-year-old woman. The
record had low-voltage character; this necessitated the use of high gain.
552 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Petersén (1982; cited in Ref. 162), who also studied the genetic
transmission of this pattern and wondered about possible auto-
somal dominant inheritance.
Subclinical Rhythmic EEG Discharge
of Adults (SREDA)
This pattern was originally described by Naquet et al. (163) as a
“paroxysmal discharge of the temporo-parietooccipital junc-
tion,” facilitated or triggered by temporary hypoxic conditions. In
spite of certain paroxysmal features and its rhythmic sharp char-
acter, the possibility of an epileptic (ictal) event has been strongly
de-emphasized. Herranz and Lopez (164) have described this
pattern as “subclinical paroxysmal activity” in 31 patients and
stressed its rarity; it has an incidence of 0.02% to 0.045%.
Westmoreland and Klass (165,166), and Miller et al. (167)
deserve particular credit for the elucidation of this discharge
type. These authors disagree with the view of Naquet et al. (163)
concerning the major role of cerebrovascular insufficiency and
hypoxia, although the average age of their patients was 61 years.
The term “SREDA” has been proposed by Miller et al. (167).
This pattern occurs mainly in the waking state or in light
drowsiness. The onset may be fairly abrupt with widespread
sharp rhythmic theta (4 to 7/sec) and occasionally with delta
activity. In other cases, the onset is gradual with a few single
sharp discharges that increasingly develop rhythmic character,
first at delta, and then at theta frequency. As to the spatial dis-
tribution, a maximum of this discharge is usually found over
the centroparietal region and especially over the vertex.
Hyperventilation may induce this pattern.
In a “revisitation” of the SREDA pattern, O’Brien et al. (168)
analyzed this pattern in seven recent patients and only one of
them showed evidence of cerebrovascular disorder.
ICTAL EEG PATTERNS
Ictal EEG patterns may be a prolongation of a well-defined
interictal pattern that becomes ictal by virtue of its long dura-
tion (this augmenting impact on neuronal function) or com-
pletely different from preceding interictal discharges. A typical
example of the former case is the absence EEG activity with
short 3/sec spike-wave bursts as interictal and prolonged 3/sec
spike-wave bursts as ictal phenomena. The emergence of a
completely different ictal pattern is exemplified by partial forms
of seizures (e.g., by cases of temporal lobe epilepsy with typical
anterior temporal sharp waves in the interval and the sudden
appearance of completely different patterns during a seizure).
Ictal patterns may be clinical, with the typical motor and
behavioral changes of a seizure of whatever category, or sub-
clinical with no demonstrable motor or behavioral changes. In
such clinically silent ictal episodes, neuropsychological studies
may demonstrate some changes.
The following discussion of the ictal EEG patterns is cursory.
Special studies and atlases on epileptic seizures and their EEG
correlates have been presented by Gibbs and Gibbs (43), Ajmone
Marsan and Ralston (169), Gastaut and Broughton (170),
Karbowski (171,172), and Oller-Daurella and Oller-Ferrer-Vidal
(173). The atlas of Oller-Daurella and Oller-Ferrer-Vidal is
unique because of the beautiful cinematographic demonstration
of clinical and EEG features. Further work on ictal EEG patterns
was presented by Ralston (174), Ralston and Papatheodorou
(175), Anziska and Cracco (176), Geiger and Harner (177), and
Blume et al. (178,179).
Generalized Tonic–Clonic Seizures
The ictal EEG is invariably obscured by muscle artifact; it is
demonstrable only in patients treated with neuromuscular block-
ers while on mechanical ventilation. Fast rhythmic spikes are pres-
ent in all leads, with some accentuation in upper frontal leads.
This fast spike activity characterizes the tonic stage and becomes
discontinuous in the clonic stage. Rhythmic slow waves alternate
with bursts of polyspikes synchronously with clonic jerks (poly-
spikes) and brief relaxation (slow bursts). A period of postictal
electrical silence is usually fairly brief, but a longer phase of very
irregular slow activity follows during the ensuing minutes.
Absence Seizures
Typified by the generalized-synchronous 3/sec spike-wave com-
plex occurring in more prolonged bursts, absence seizures have
been discussed in the section “Three-Per-Second Spike-Wave
Complexes.”
Complex Partial Seizures
The ictal EEG patterns of these seizures are “as variable as the
clinical features” (170). The astounding variability of the ictal
patterns is impressively demonstrated in the atlas of Oller-
Daurella and Oller-Ferrer-Vidal (173). In early work of Gibbs et
al. (180,181), the ictal EEG activity of complex partial seizures
was described as a special type of seizure discharge characterized
by bursts of serrated slow waves, flat-topped 4/sec waves, and
high-voltage 6/sec waves (“seizure discharge of the psychomotor
type”). Much greater complexities of the ictal patterns were
stressed by Gastaut and Vigouroux (182) and Klass (183). In
about 5% of the cases, a complex partial seizure may occur with-
out any recordable ictal EEG activity (170).
Focal Motor and Other Simple Partial Seizures
Cortical focal (partial) seizures are expected to be associated
with a local discharge consisting of a sequence of repetitive
spikes over the area contralateral to the motor or sensory (visual
and auditory) clinical manifestations. This is correct in a large
number of cases but does not pertain to all cases. According to
Gastaut and Broughton (170), there is no demonstrable ictal
EEG discharge in about 30% of the cases. This is particularly
true for circumscript motor rolandic attacks; ictal discharges
arising from the precentral gyrus may be quite subtle and incon-
spicuous or not demonstrable at all in the EEG (43).
In cases with a well-documented EEG correlate of the
seizure, the attack is often initiated by local desynchronization,
that is, very fast and very low voltage spiky activity, which grad-
ually rises in amplitude with diminishing frequency (177).
Myoclonic Seizures
Myoclonus is a very complex phenomenon; a considerable
number of disorders with myoclonus do not fall into the epileptic

category (184). Epileptic myoclonus is classically characterized
by concomitant polyspikes or polyspike-wave discharges in the
EEG, of bilateral- or generalized-synchronous character, usu-
ally with maximum over the frontal regions. A review of
myoclonus and its relationship to epilepsy has been presented
by Hallett (185).
Tonic Seizures
The relationship between tonic seizures and massive fast spike
activity was discussed previously in this chapter. This pattern is
the typical correlate of tonic attacks occurring in patients with
the Lennox–Gastaut syndrome.
Some tonic seizures are characterized by simple flattening or
desynchronization of all activity during the attack (170). These
authors also describe rhythmic activity around 10/sec and a
very rare diffuse slow-wave pattern (mainly delta frequency) as
EEG concomitants of tonic seizures.
Atonic Seizures
These seizures are customarily divided into a short form, such
as simple drop attack, lasting only seconds, and a longer form,
lasting minutes and described as “inhibitory.” The EEG shows a
few polyspike waves or spike waves of generalized distribution
followed by large slow waves (short forms). More rhythmic
spike activity around 10/sec and intermixed slow activity in all
leads constitute the EEG correlate of the longer lasting attacks;
rhythmic slow spike-wave activity (1.5 to 2/sec) may also occur
(170,173).
Akinetic Seizures
These attacks are characterized by arrest of all motion, which,
however, is not caused by sudden loss of tone as in atonic
seizures. The patients, usually children, are in an absence-like
state; the EEG correlate is a slow (mostly 1 to 2/sec) spike-wave
discharge, generalized synchronous, and often with clocklike
rhythmicity. This type of seizure is rather poorly understood
and, in recent years, the use of this term has been discouraged.
PERIODIC EPILEPTIFORM DISCHARGES
ASSOCIATED WITH PARTICULAR
NEUROLOGIC DISORDERS
Some rhythmically (or nearly rhythmically) recurring dis-
charges or patterns are of paroxysmal character, but are usually
not associated with epileptic seizure disorders characterized by
chronically recurrent seizures. Periodic discharges or periodic
activities are most commonly an important EEG feature of a
severe ongoing CNS disease with certain paroxysmal or even
overt epileptogenic properties. They are hence disease-
suggestive and sometimes virtually disease-specific, rather than
suggestive of epileptic seizure disorders in general. A review of
periodic activities was presented by Gaches (186), as well as by
Bauer and Pieber (187).
The periodic character of these patterns in the presence of
severe CNS involvement remains enigmatic; such rhythmic
firing is markedly different from the predominantly random
Chapter 27 ■ Epilepsy in Adults and the Elderly 553
character of interictal seizure discharges (spikes, sharp wave,
etc.). There is no satisfactory model to explain the periodicity of
the discharges.
Periodic discharges are always of large amplitude, mostly in
the range of 100 to 300 V. These may be simple sharp waves,
but of a duration that usually exceeds 150 msec. Other periodic
discharges are compounded and polymorphic. Periodic dis-
charges may be focal, widely scattered, or generalized synchro-
nous. A fine tabulated review of periodic discharges has been
presented by Spehlmann (188).
Subacute Sclerosing Panencephalitis (SSPE)
The periodic complexes of SSPE dominate the second state of
this disease; this is a prolonged phase in which the clinical
diagnosis is usually made. The SSPE discharge has become
exceptionally rare in developed countries due to its relation-
ship with measles and the excellent preventive effect of measles
immunization. This type of complex discharge almost never
occurs in other clinical conditions and is hence almost disease-
specific. It is a very reliable and highly contributory diagnostic
finding in this disease. These discharges were first mentioned
by Balthazar (189) and extensively described by Radermecker
(190,191) and Cobb and Hill (192). The discharges show a
duration from 0.5 to 3 seconds and are formed by two or more
waves with mean amplitude of 500 V (100 to 1000 V)
(Fig. 27.14) (193). In some cases, the voltage may even reach
1400 V (8). The enormous height of this discharge may
impress the inexperienced as a movement artifact; this idea is
supported by frequent associated motor events such as
myoclonus or sudden loss of tone.
The periodicity becomes manifest as the disease progresses;
in its earlier stages, which are usually dominated by diffuse 1 to
3/sec activity of fairly rhythmic character, the compounded dis-
charge may be aperiodic. Sometimes, the periodic discharge is
present in the earliest stage of the disease and may persist to the
fatal outcome; more often, this pattern disappears in the termi-
nal or third stage. Vitrai et al. (194) investigated the SSPE com-
plex with the use of time functions of the auto- and
cross-correlation coefficients.
The elements of the discharge are variable; a giant slow wave
is usually mixed with several sharp waves. The discharge is gen-
eralized, with a maximum over the frontocentral areas and ver-
tex. The frequency of the complexes ranges from 4 to 16/min
(193). The discharge is more impressive in the waking state.
Accompanying myoclonus is fairly synchronous with the dis-
charge; the motor activity may precede or trail the discharge in
a range of 200 to 800 msec (195); this author showed minor
interhemispheric asynchronies (15 msec) and an earlier appear-
ance in parieto-occipital areas than in frontocentral areas,
where the discharge shows its most impressive voltage.
The origin of the periodic SSPE discharge is controversial.
Radermecker and Poser (196) have proposed a deep thalamic or
mesencephalic-reticular origin; this view has been supported by
Cobb (197,198), who performed depth EEG recordings, and by
Lombroso (199), who studied direct current (DC) potential
changes during the periodic discharge. Other authors favor the
idea of a cortical origin (200–202).
554 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 27.14 Periodic discharges in subacute sclerosing panencephalitis (SSPE). A: At an earlier state with some preser-
vation of other activities. B: At a more advanced state (recorded from two different patients, aged 8 and 9 years).
Herpes Simplex Encephalitis
In this extremely severe but not invariably fatal necrotizing
encephalitis, slow repetitive spikes have been noted by
Radermecker (191). These observations were subsequently con-
firmed by Millar and Coey (203), Perier et al. (204), Carmon et
al. (205), Rawls et al. (206), Adams and Jennett (207), Upton
and Gumpert (208), Gaches and Arfel (209), Elian (210), Kugler
et al. (211), and Cobb (212). The discharge was also found in
experimental disease inoculated rabbits by Gupta and Seth
(213).
In an earlier stage, local mostly unilateral temporal poly-
morphic delta waves are the most striking feature, but soon
large sharp waves emerge over the most affected region. These
discharges usually fire at intervals from 2 to 4 seconds. The dis-
charge may be quite slow and exceed 1000 msec; consider that
a sharp wave is defined as having a duration from 70 to
150 msec. The amplitudes are in the range from 100 to about
500 V. Similar periodic sharp discharges have been demon-
strated in neonatal herpes simplex encephalitis (214). In the
course of the disease, the originally regional discharge tends to
become generalized synchronous; asynchronous bilateral dis-
charges may also occur. Predominantly occipital localization of
the periodic discharges has been reported by Bergey et al. (215).
In contrast with the complexes found in SSPE, the periodic
discharge of herpes simplex encephalitis is of short periodicity,
with the greatest intervals of 4 seconds or less; according to
Gaches (186), this criterion is always met in the herpes simplex
discharge. In some cases, the periodic discharge can be detected
only when almost daily repeat records are carried out. This
explains the fact that periodic discharges are not reported in all
cases of herpes simplex encephalitis.
Creutzfeldt–Jakob Disease
The clinical and neuropathologic features of this rapidly pro-
gressive dementing illness were first established by the work of
Creutzfeldt (216), Jakob (217), and Kirschbaum (218,219); the
much later identification of pathogenic prion proteins and a
variant form of the disorder has only occurred in recent
decades.
In the most common, sporadic form of CJD, a typical peri-
odic discharge can be seen on EEG, consisting of a sharp wave
or a sharp triphasic complex of 100- to 300-msec duration with

Figure 27.15 Periodic discharges in Creutzfeldt–Jakob disease. Note
mild lateralization of the recurrent sharp waves to the left. The 68-year-
old patient was demented, with psychotic features; there were long tract
signs, fasciculation, and frequent myoclonus. Myoclonic jerking did not
alter the tracing.
a repetition rate of 0.5 to 2/sec or intervals of 500 to 2000 msec.
The discharges occur against a severely disordered background
of activity, mostly in generalized synchrony. Some cases show
unilateral onset with discharges over one hemisphere or lateral-
ized to one hemisphere for several days or weeks (220).
According to Tariska et al. (221), the periodic discharges may
even remain unilateral in the course of the disease.
The periodic activity usually shows a maximum over the
anterior region except for the Heidenhain form, which has a
posterior maximum (Fig. 27.15); in this special form, cortical
blindness is a common feature. In sleep, the periodic dis-
charges tend to disappear. The absence of periodic discharges
after 10 weeks of illness argues against sporadic
Creutzfeldt–Jakob disease as a diagnosis (222). The EEG stud-
ies of Chiofalo et al. (223) are based on the observation of 27
cases. The differential diagnosis of the combined clinical and
electroencephalographic picture of this disease can be quite
complex (224). Myoclonus is a typical clinical feature of this
disease, but the periodic patterns may or may not be associated
with myoclonus (184).
It has to be emphasized that the periodic discharges of
Creutzfeldt–Jakob disease (i) do not occur in every sporadic
case of this disease (even with numerous repeat recordings)
(225); (ii) are most common in the sporadic form and much
less commonly, if ever, seen in the familial, iatrogenic, and vari-
ant forms of CJD; and (iii) may occasionally be found in other
disorders.
Chapter 27 ■ Epilepsy in Adults and the Elderly 555
PERIODIC LATERALIZED EPILEPTIFORM
DISCHARGES (PLEDS)
In the wake of earlier work (226–231), Chatrian et al. (42) gave
an extensive account of periodic lateralized or focal discharge
that may occur in a variety of acute neurologic conditions. This
pattern is most often associated with acute cerebral infarctions
but may also occur in neoplastic and inflammatory pathology.
Periodic discharges occurring in herpes simplex encephalitis
are discussed earlier in this chapter. PLED activity, however,
may also occur in other types of encephalitis—among others, in
infectious mononucleosis encephalitis, according to Aminoff et
al. (232). Markand and Daly (233), Dauben and Adams (234),
and Bauer et al. (235) presented further important work on this
pattern, which also plays a major role in experimental cerebral
embolism (236).
PLEDs may be simple and large sharp waves or complex
(compounded) discharges with mixed spiky and slower ele-
ments. The amplitudes usually lie around 100 to 300 V, but
occasionally may be much higher. The firing rate may be as fast
as 3/sec or as slow as 12/min (Fig. 27.16) (42).
The discharge is found over the maximally involved area,
and the local background activity is almost always severely dis-
ordered. The term PLED implies lateralization. It is possible,
however, that PLED activity is generated independently over
both hemispheres. De la Paz and Brenner (237) have investi-
gated the bilateral independent type of PLEDs referred to as
BIPLEDs. Although strokes were found to be the leading cause
in the PLED group, CNS infection and epileptic seizure disor-
ders predominated in the group with BIPLEDs. Bilateral-
synchronous (generalized-synchronous) periodic activity
should not be listed as BIPLEDs in view of the original defini-
tion of Chatrian et al. (42), but is often termed generalized peri-
odic epileptiform discharges (GPEDs). The term cerebral
bigeminy has been used by Aldrich and Pugh (1984) for bilater-
ally independent PLED activity (238). Even TRI-PLEDs, arising
from three different areas, have been reported (left frontal, left
posterotemporal, and right posterotemporal) in an 85-year-old
patient with renal encephalopathy (239).
PLEDs are often associated with simultaneous focal motor
twitching in contralateral face or fingers, hand, arm, leg, foot,
etc. This underscores the paroxysmal character of the pattern. It
is usually a temporary pattern that changes into other abnor-
malities within 1 to 2 days. Patients with PLEDs are in most
cases acutely ill and often show a history of a variety of mixed
medical problems, such as cerebral arteriosclerosis plus renal
insufficiency or chronic alcoholism or diabetes mellitus.
Patients with PLEDs are commonly older adults, although
this pattern occasionally occurs in young adults and children
(240). Andriola (241) noted PLEDs as well as BIPLEDs in
12 children with various types of acute CNS disease; those with
bilateral activity (BIPLED) expired in the acute state. Ritaccio
and March (242) reported the association of BIPLEDs with
complex partial status epilepticus. Silbert et al. (243) described
another variant of PLEDs with independent ipsilateral periodic
lateralized discharges. A clinical curiosity appeared to be the
observation of Chabolla et al. (244): a young man suffering
556 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 27.16 A: Periodic lateralized epileptiform discharge (PLED) in
a 70-year-old patient, widespread over left hemisphere with left
midtemporal maximum. Note poorly developed background of activity.
Cerebrovascular accident associated with twitching of right arm. Patient
lethargic and obtunded. B: PLED over right central–midtemporal
region, sometimes also involving the right parietal area. Note well-
preserved physiologic activities of waking and light drowsiness. Also
note the complex and multiphasic character of the discharge. Patient is
68 years old with cerebrovascular disease with right centroparietal low-
density area in CT scan (infarction).
from multiple sclerosis (MS) with exacerbations ushered in by
complex partial status epilepticus associated with an enhancing
right frontal MRI lesion and PLEDs in the EEG. Another report
on PLEDs in MS (right temporal PLEDs and MRI lesion) has
been presented by Gandelman-Marton et al. (245).
A very thoughtful study of PLEDs deserves special attention:
a critical review of Pohlmann-Eden et al. (246). These authors
consider PLEDs as “an EEG signature of a dynamic pathophys-
iologic state in which unstable neurobiologic processes create
an ictal–interictal continuum, with the nature of the underlying
neuronal injury, the patient’s preexisting propensity to have
seizures, and the coexistence of any acute metabolic derange-
ments all contributing to whether seizures occur or not.”
Pohlmann-Eden et al. have strongly reemphasized the domi-
nant etiologic role of strokes in the generation of PLEDs. From
the evidence of functional neuroimaging studies, these authors
feel that PLEDs “might reflect a key pattern for focal hyperex-
citability in the penumbra zone of ischemic stroke.”
A nonlateralized variant of PLEDs is periodic epileptiform
discharges occurring in the midline (PEDIM), as cited by
Westmoreland et al. (247). These midline discharges are pre-
sumed to arise from vascular watershed areas, between the
anterior and middle and between the middle and posterior
cerebral arteries.
REFERENCES
1. Arroyo S, Lesser RP, Fisher RS, et al. Clinical and electroencephalo-
graphic evidence for sites of origin of seizures with diffuse elec-
trodecremental pattern. Epilepsia. 1994;35:974–987.
2. Kellaway P, Crawley JW, Kagawa N. A specific electroencephalo-
graphic correlate of convulsive equivalent disorders in children.
J Pediatr. 1959;55:582–592.
3. Ohtahara S, Yamatogi Y. Evolution of seizures and EEG abnormal-
ities in childhood onset epilepsy. In: Wada JA, Ellington RJ, eds.
Handbook of Electroencephalography and Clinical Neurophysiology.
Revised series. Vol 4. Amsterdam: Elsevier; 1990:457–477.
4. IFSECN. Preliminary proposal for an EEG terminology by the ter-
minology committee of the International Federation for
Electroencephalography and Clinical Neurophysiology.
Electroencephalogr Clin Neurophysiol. 1961;13:646–650.
5. IFSECN. Proposal for an EEG terminology committee of the
International Federation for Electroencephalography and Clinical
Neurophysiology. Electroencephalogr Clin Neurophysiol. 1966;30:
306–310.
6. IFSECN. A glossary of terms commonly used by clinical electroen-
cephalographers. Electroencephalogr Clin Neurophysiol. 1974;37:
538–548.
7. Dutertre F. Catalogue of the main EEG patterns. In: Remond A,
editor-in-chief. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 11A. Amsterdam: Elsevier; 1977:40–79.
8. Kiloh LG, McComas AJ, Osselton JW. Clinical Electroencephalography.
3rd ed. London: Butterworth; 1972.
9. Roger H, Cornil L, Paillas JE. Les Epilepsies. Paris: Flammarion;
1949.
10. Steriade M, Amzica F, Contreras D. Synchronization of fast
(30–40 Hz) spontaneous cortical rhythms during brain attention.
J Neurosci. 1996;16:392–417.
11. Gotman J. Could interictal epileptic spikes actually be postictal
epileptic spikes? Electroencephalogr Clin Neurophysiol. 1984;58:9P
[abstract].
12. Gotman J, Marciani MG. Electroencephalographic spiking activ-
ity, drug levels and seizure occurrence in epileptic patients. Ann
Neurol. 1985;17:597–603.
13. Pollen DA, Reid KH, Perol P. Micro-electrode studies of experi-
mental wave and spike in the cat. Electroencephalogr Clin
Neurophysiol. 1964;17:57–67.
14. Matsuo F, Knott JR. Focal positive spikes in electroencephalogra-
phy. Electroencephalogr Clin Neurophysiol. 1977;42:15–25.
15. Maulsby RL. Some guidelines for assessment of spikes and sharp
waves in EEG tracings. Am J EEG Technol. 1971;11:3–16.

16. Binnie CD, Polkey CE, Spencer S. Positive spikes over a cortical
laceration. Electroencephalogr Clin Neurophysiol. 1989;72:106P
[abstract].
17. Bergen D. A unique EEG pattern in mucolipidosis. Electro-
encephalogr Clin Neurophysiol. 1979;46:80.
18. Engel J Jr, Rapin I, Giblin DR. Electrophysiological studies in two
patients with cherry red spot-myoclonus syndrome. Epilepsia
(New York). 1977;18:73–87.
19. Gregory DL, Wong PK. Topographical analysis of the centrotem-
poral discharges in benign rolandic epilepsy of childhood.
Epilepsia (New York). 1984;25:705–711.
20. Wong PKH, Gregory D. Dipole fields in rolandic discharges. Am J
EEG Technol. 1988;28:243–250.
21. Niedermeyer E. Clinical relevance of EEG signals in epilepsies. In:
Wada JA, Ellingson RJ, eds. Handbook of Electroencephalography
and Clinical Neurophysiology. Revised series. Vol 4. Amsterdam:
Elsevier; 1990:237–261.
22. Bauer G, Markoff R. Familienuntersuchungen bei einem seltenen
paroxysmalen Muster. Z EEG-EMG. 1970;1:19–26.
23. Shewmon DA, Erwin RJ. The effect of focal interictal spikes on
perception and reaction time. I. General considerations.
Electroencephalogr Clin Neurophysiol. 1988;69:319–337.
24. Shewmon DA, Erwin RJ. The effect of focal interictal spikes on
perception and reaction time. II. Neuroanatomic specificity.
Electroencephalogr Clin Neurophysiol. 1988;69:338–352.
25. Shewmon DA, Erwin RJ. Focal spike-induced cerebral dysfunction
is related to the after-coming slow wave. Ann Neurol. 1988;23:
131–137.
26. Niedermeyer E, Krauss GL, Pacer CE. The electroencephalogram
and mental activation. Clin Electroencephalogr. 1989;20:215–227.
27. Gotman J, Gloor P, Schaul N. Comparison of traditional reading of
the EEG and automatic recognition of interictal epileptic activity.
Electroencephalogr Clin Neurophysiol. 1978;44:48–60.
28. Webber WRS, Litt B, Lesser RP, et al. Automatic spike detection:
what should the computer imitate? Electroencephalogr Clin
Neurophysiol. 1993;87:364–373.
29. Wilson SB, Harner RN, Duffy FH, et al. Spike detection. I.
Correlation and reliability of human experts. Electroencephalogr
Clin Neurophysiol. 1996;98:186–198.
30. Webber WRS, Litt B, Wilson K, et al. Practical detection of epilep-
tiform discharges (EDs) in the EEG using an artificial neural net-
work: a comparison of raw and parameterized EEG data.
Electroencephalogr Clin Neurophysiol. 1994;91:194–204.
31. Gambardella A, Aguglia U, Oliveri RI, et al. Negative myoclonic
status due to antiepileptic drug tapering: report of three cases.
Epilepsia. 1997;38:819–823.
32. Werhahn KJ, Noachtar S. Epileptic negative myoclonus. In:
Lueders HO, Noachtar S, eds. Epileptic Seizures: Pathophysiology
and Clinical Semiology. New York: Churchill Livingstone;
2000:473–483.
33. Fisch BJ. Interictal epileptiform activity: diagnostic and behavioral
implications. J Clin Neurophysiol. 2003;20:155–162.
34. Warach S, Ives JR, Schlaug G, et al. EEG-triggered echo-planar
functional MRI in epilepsy. Neurology. 1996;47:89–93.
35. Aarts JHP, Binnie CD, Smit AM, et al. Selective cognitive impair-
ment during focal and generalized epileptiform EEG activity.
Brain. 1984;107:293–308.
36. Hughes JR. The significance of interictal spike discharge: a review.
Clin Electroencephalogr. 1997;28:60 [abstract].
37. Jasper HH. Epilepsy. In: Penfield W, Erickson TC, eds. Epilepsy and
Cerebral Localization. Springfield, IL: Charles C Thomas;
1941:380–454.
Chapter 27 ■ Epilepsy in Adults and the Elderly 557
38. Cukier F, André, N., Monod N, et al. EEG contribution to
diagnosing intraventricular haemorrhage in premature
infants. Electroencephalogr Clin Neurophysiol. 1974;36:840
[abstract].
39. Werner SS, Stockard JE, Bickford RG. Atlas of Neonatal
Electroencephalography. New York: Raven Press; 1977.
40. Zobniw AM, Yaworth S, Niedermeyer E. Depth electroencephalo-
graphy. J Electrophysiol Technol. 1975;2:215–240.
41. Ribeiro M, Niedermyer E, Hertz S. Slow sharp waves. Clin
Electroencephalogr. 1999;30:114–117.
42. Chatrian GE, Shaw CM, Leffman H. The significance of periodic
epileptiform discharges in EEG: an electrographic, clinical and
pathological study. Electroencephalogr Clin Neurophysiol.
1964;17:177–193.
43. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 2.
Cambridge, MA: Addison-Wesley; 1952.
44. Gibbs FA, Davis H, Lennox WG. The electroencephalogram in
epilepsy and in conditions of impaired consciousness. Arch Neurol
Psychiatr (Chicago). 1935;34:1133–1148.
45. Jasper HH, Kershman J. Classification of the E.E.G. in epilepsy.
Electroencephalogr Clin Neurophysiol (suppl). 1949;2:123–131.
46. Gastaut H, Roger J, Soulayrol R, et al. Childhood epileptic
encephalopathy with diffuse slow spike-waves (otherwise known
as “petit mal variant”) or Lennox syndrome. Epilepsia (Amsterdam).
1966;7:139–179.
47. Brenner RP, Atkinson R. Generalized paroxysmal fast activity: elec-
troencephalographic and clinical features. Ann Neurol. 1982;11:
386–390.
48. Chayasirisobhon S, Cullis P, Sack R, et al. Grand mal discharge.
Clin Electroencephalogr. 1984;15:155–158.
49. Rodin E, Smid N, Mason K. The grand mal pattern of Gibbs, Gibbs
and Lennox. Electroencephalogr Clin Neurophysiol. 1976;40:
401–406.
50. Niedermeyer E. The Lennox–Gastaut syndrome: a severe type of
childhood epilepsy. Dtsch Z Nervenheilk. 1969;195:263–283.
51. Niedermeyer E. The Generalized Epilepsies. Springfield, IL: Charles
C Thomas; 1972.
52. Niedermeyer E. Compendium of the Epilepsies. Springfield, IL:
Charles C Thomas; 1974.
53. Niedermeyer E, Walker AE, Burton C. The slow spike-wave com-
plex as a correlate of frontal and fronto-temporal posttraumatic
epilepsy. Eur Neurol (Basel). 1970;3:330–346.
54. Niedermeyer E. The electroencephalogram in the differential diag-
nosis of the Lennox–Gastaut syndrome. In: Niedermeyer E, Degen
R, eds. The Lennox–Gastaut Syndrome. New York: Alan R. Liss;
1988:177–220.
55. Silverman D. Clinical correlates of the spike-wave complex.
Electroencephalogr Clin Neurophysiol. 1954;6:663–669.
56. Clark EC, Knott JR. Paroxysmal wave and spike activity and diag-
nostic sub-classification. Electroencephalogr Clin Neurophysiol.
1955;7:161–164.
57. Lundervold A, Henriksen GF, Fegersten L. The spike wave com-
plex. A clinical correlation. Electroencephalogr Clin Neurophysiol.
1959;11:13–22.
58. Cohn R. Spike–dome complex in the human electroencephalo-
gram. Arch Neurol Psychiatr (Chicago). 1954;71:699–706.
59. Weir B. The morphology of the spike-wave complex. Electro-
encephalogr Clin Neurophysiol. 1965;19:284–290.
60. Jung R, Toennies F. Über Entstehung and Erhaltung von
Krampfentladungen. Die Vorgänge im Reizort und die
Krampffähigkeit des Gehirns. Arch Psychiatr Nervenkr.
1950;1985:701–735.
558 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
61. Pollen DA. Intracellular studies of cortical neurons during thala-
mic induced wave and spike. Electroencephalogr Clin Neurophysiol.
1964;17:398–406.
62. Weir B, Sie PG. Extracellular unit activity in cat during the spike-
wave complex. Epilepsia (Amsterdam). 1966;7:30–43.
63. Halász P. The generalized epileptic spike-wave mechanism and the
sleep–wakefulness system. Acta Physiol Acad Scient Hung. 1972;42:
293–314.
64. Niedermeyer E. Immediate transition from a petit mal absence
into a grand mal seizure. Eur Neurol (Basel). 1976;14:11–16.
65. Dondey M. Transverse topographical analysis of petit mal
discharges: diagnostical and pathogenic implications.
Electroencephalogr Clin Neurophysiol. 1983;55:361–371.
66. Rodin E, Ancheta O. Cerebral electrical fields during petit mal
absences. Electroencephalogr Clin Neurophysiol. 1987;66:457–466.
67. Gotman J. Interhemispheric relations during bilateral spike-and-
wave activity. Epilepsia (New York). 1981;22:453–466.
68. Lemieux JF, Blume WT. Topographical evolution of the spike-wave
complexes. Electroencephalogr Clin Neurophysiol. 1983;56:30P
[abstract].
69. Lüders H, Daube J, Johnson R, et al. Computer analysis of general-
ized spike-and-wave complexes. Epilepsia (New York). 1980;21:183
[abstract].
70. Baddeley AD. Working Memory. Oxford: University Press; 1986.
71. Fuster JM. Memory and planning. In: Jasper HH, Riggio S, Patricia
Goldman-Rakic PS, eds. Epilepsy and the Functional Anatomy of
the Frontal Lobe. New York: Raven Press; 1995:9–19.
72. Fuster JM. Memory in the Cerebral Cortex. Cambridge, MA: MIT
Press; 1996.
73. Niedermeyer E. Frontal iobe functions and dysfunctions. Presented
at Symposium, Lima, Peru, 1997.
74. Niedermeyer E. Primary (idiopathic) generalized epilepsy. Clin
Electroencephalogr. 1996;27:1–21.
75. Walker A, Morello G. Experimental petit mal. Trans Am Neurol
Assoc. 1967;57–61.
76. Hunter J, Jasper HH. Effects of thalamic stimulation in unanes-
thetized animals. Electroencephalogr Clin Neurophysiol. 1949;1:
303–324.
77. Sperling MR, Skolnick BE. Cerebral blood flow during spike-wave
discharges. Epilepsia. 1995;36:156–163.
78. Gianelli M, Cantello R, Civardi C, et al. Idiopathic generalized
epilepsy: magnetic stimulation of motor cortex time-locked and
unlocked to 3-Hz spike-and-wave discharges. Epilepsia. 1994;35:
53–60.
79. Gibbs FA, Gibbs EL, Lennox WG. The influence of blood sugar
level on the wave and spike formation in petit mal epilepsy. Arch
Neurol Psychiatr (Chicago). 1939;50:111–128.
80. Lennox WG. Epilepsy and Related Disorders. Vol 1. Boston, MA:
Little, Brown & Co.; 1960.
81. Gastaut H. Clinical and electroencephalographic correlates of gen-
eralized spike and wave bursts occurring spontaneously in man.
Epilepsia (Amsterdam). 1968;9:179–184.
82. Silverman D. Phantom spike-wave and the fourteen and six per
second positive spike pattern: a consideration of their relationship.
Electroencephalogr Clin Neurophysiol. 1967;23:203–217.
83. Marshall C. Some clinical correlates of the wave and spike phan-
tom. Electroencephalogr Clin Neurophysiol. 1955;7:633–636.
84. Thomas JE. A rare electroencephalographic pattern: the six per second
spike and wave discharge. Neurology (Minneapolis). 1957;7:438–442.
85. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 3.
Reading, MA: Addison-Wesley; 1964.
86. Hughes JR, Schlagenhauff RE, Magoss M. Electro-clinical corre-
lations in the six per second spike and wave complex.
Electroencephalogr Clin Neurophysiol. 1965;18:71–77.
87. Thomas JE, Klass DW. Six-per-second spike and wave pattern in
the electroencephalogram: a reappraisal of clinical significance.
Neurology. 1968;18:587–593.
88. Hecker A, Kocher R, Ladewig D, et al. Das Minatur-Spike-Wave-
Muster. Das EEG Labor. 1979;1:51–56.
89. Hughes JR. Two forms of the 6/sec spike and wave complex.
Electroencephalogr Clin Neurophysiol. 1980;48:535–550.
90. Hughes JR. Changes in the annual incidence of the 6/sec spike
and wave pattern. Clin Electroencephalogr. 1991;22:71–74.
91. Kocher R, Scollo-Lavizzari G, Ladewig D. Miniature-spikewave-
Muster: Elektroencephalographisches Korrelat in der
Abstinenzphase bei Medikamentenabhängigkeit. Z EEG-EMG.
1975;6:78–82.
92. Tharp BR. The 6-per-second spike and wave complex. Arch
Neurol (Chicago). 1966;15:533–537.
93. Thorner MW. Procurement of Electroencephalograph Tracings in
1000 Flying Cadets for Evaluating the Gibbs Technique in Relation
to Flying Ability. USAD School of Aviation Medical Research
Report, No. 7–1; 1942.
94. Gibbs FA, Gibbs EL, Lennox WG. Electroencephalographic clas-
sification of epileptic patients and control subjects. Arch Neurol
Psychiatr. 1943;50:111–128.
95. Harty JE, Gibbs EG, Gibbs FA. Electroencephalographic study of
two hundred and seventy-five candidates for military service.
War Med (Chicago). 1942;2:923–930.
96. Williams D. The nature of transient outbursts in the electroen-
cephalogram of epileptics. Brain. 1944;67:10–37.
97. Buchthal F, Lennox M. The EEG effect of metrazol and photic
stimulation in 682 normal subjects. Electroencephalogr Clin
Neurophysiol. 1953;5:545–558.
98. Dell AR, Dell MB. L’électroencéphalogramme systématique du
personnel navigant. Étude préliminaire portant sur 1056 cas. Méd
Aéro. 1958;13:33–47.
99. Blanc C, Lafontaine E, Laplane R. Les activités EEG paroxystiques
généralisées ou focales chez les adultes non comitiaux: Études
évolutive correlations longitudinales. Rev Neurol (Paris). 1964;111:
362–369.
100. O’Connor P. Analysis of 500 routine EEGs of R.A.F. aircrew
cadets. Electroencephalogr Clin Neurophysiol. 1964;17:341
[abstract].
101. Hill D. EEG in episodic psychotic and psychopathic behaviour. A
classification of data. Electroencephalogr Clin Neurophysiol.
1952;4:419–442.
102. Baylor University College of Medicine and the Methodist
Hospital. The Normative Electroencephalographic Data Reference
Library. Final Report. Washington, DC: National Aeronautics and
Space Administration; 1966.
103. Bennett DR. Spike-wave complexes in “normal” flying personnel.
Aerospace Med. 1967;38:1276–1282.
104. Lennox-Buchthal M, Buchthal F, Rosenfalck P. Correlation of
electroencephalographic findings with crash rate of military jet
pilots. Epilepsia (Amsterdam). 1959;1:366–373.
105. Kitamura K, Asakura T. Clinical significance of spikes and sharp
waves in EEG. Electroencephalogr Clin Neurophysiol (suppl). 1958;
12:16
106. Zivin L, Ajmone Marsan C. Incidence and prognostic significance
of “epileptiform” activity in the EEG of nonepileptic subjects.
Brain. 1968;91:751–777.

107. Chatrian GE. Paroxysmal pattern in “normal” subjects. In: Remond
A, editor-in-chief. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 6A. Amsterdam: Elsevier; 1976:114–123.
108. Nekhorocheff MI. La stimulation lumineuse intermittante chez
l’enfant normal. Rev Neurol (Paris). 1950;83:601–602.
109. Kellaway P, Fox BJ. Electroencephalographic diagnosis of cerebral
pathology in infants during sleep. I. Rationale, technique and the
characteristics of normal sleep in infants. J Pediatr. 1952;41:
262–287.
110. Brandt S, Brandt H. The electroencephalographic patterns in
young healthy children from 0 to five years of life. Acta Psychiatr
Scand. 1955;30:77–89.
111. Corbin HPF, Bickford RG. Studies of the electroencephalogram
of normal children: comparison of visual and automatic fre-
quency analysis. Electroencephalogr Clin Neurophysiol. 1955;
7:15–28.
112. Trojaborg W. Focal spike discharges in children. Acta Paediatr
Scand (Suppl). 1966;168:1–13.
113. Eeg-Olofsson O, Petersén I, Selldén U. The development of the
electroencephalogram in normal children from the age of 1 through
15 years. Paroxysmal activity. Neuropaediatrie. 1971;2:375–404.
114. Cavazzuti GB, Cappella L, Nalin A. Longitudinal study of epilep-
tiform EEG patterns in normal children. Epilepsia (New York).
1980;21:43–55.
115. Koshino Y, Niedermeyer E. The clinical significance of small
sharp spikes in the electroencephalogram. Clin Electroencephalogr.
1975;6:131–140.
116. Westmoreland BF, Reiher J, Klass DW. Recording small sharp
spikes with depth electroencephalography. Epilepsia (New York).
1979;20:599–612.
117. Small JG, Small IF, Milstein V, et al. Familial associations with
EEG variants in manic depressive disease. Arch Gen Psychiatr.
1975;32:43–48.
118. Reiher J, Klass DW. Two common EEG patterns of doubtful clin-
ical significance. Med Clin North Am. 1968;52:933–940.
119. Lebel M, Reiher J, Klass D. Small sharp spikes (SSS). Reassessment
of electroencephalographic and clinical significance.
Electroencephalogr Clin Neurophysiol. 1977;43:463 [abstract].
120. White JC, Langston JW, Pedley TA. Benign epileptiform tran-
sients of sleep: clarification of the small sharp spike controversy.
Neurology. 1977;27:1061–1068.
121. Klass DW, Westmoreland BF. Nonepileptogenic epileptiform
electroencephalographic activity. Ann Neurol. 1985;18:627–635.
122. Gutrecht JA. Clinical implications of benign epileptiform transients
of sleep. Electroencephalogr Clin Neurophysiol. 1989;72:486–490.
123. Low MD, Henrikson K, Lamont D, et al. The relationship of small
sharp spikes to seizures. Electroencephalogr Clin Neurophysiol.
1984;58:11P [abstract].
124. Hughes JR, Gruener, GT. Small sharp spikes revisited: further
data on this controversial pattern. Electroencephalogr Clin
Neurophysiol. 1984;15:208–213.
125. Saito F, Fukushima Y, Kubota S. Small sharp spikes: possible rela-
tionship to epilepsy. Clin Electroencephalogr. 1987;18:144–119.
126. Molaie M, Santana HB, Cavanaugh WA. Benign epileptiform
transients of sleep (BETS) revisited. Electroencephalogr Clin
Neurophysiol. 1989;72:19P [abstract].
127. Gibbs EL, Gibbs FA. Das Elektroenzephalogramm bei congeni-
taler Anophthalmie. Z EEG-EMG. 1981;12:171–173.
128. Gibbs FA, Gibbs EL, Gibbs TJ. Relation between specific types of
occipital dysrhythmia and visual defects. Johns Hopkins Bull.
1968;122:343–349.
Chapter 27 ■ Epilepsy in Adults and the Elderly 559
129. Lairy GC, Harrison A, Leger EM. Foyers EEG biocciptaux asyn-
chrones de pointes chez l’enfant mal voyant et aveugle d’âge sco-
laire. Rev Neurol (Paris). 1964;111:351–353.
130. Cannon WB, Rosenblueth A. The Supersensitivity of Denervated
Structures: A Law of Denervation. New York: Macmillan; 1949.
131. Kellaway P, Bloxsom A, MacGregor M. Occipital spike foci associ-
ated with retrolental fibroplasia and other forms of retinal loss in
children. Electroencephalogr Clin Neurophysiol. 1955;7:469–470.
132. Gibbs EL, Fois A, Gibbs FA. The electroencephalogram in retro-
lental fibroplasia. N Engl J Med. 1955;253:1102–1106.
133. Lombroso CT, Schwarts IH, Clark DM, et al. Ctenoids in healthy
youth. Controlled study of 14 and 6-per-second positive spiking.
Neurology (Minneapolis). 1966;16:1152–1158.
134. Hughes JR. The 14 and 7 per second positive spikes—a reap-
praisal following a frequency count. Electroencephalogr Clin
Neurophysiol. 1960;12:495–496.
135. Hughes JR. A review of the positive spike phenomenon. In:
Wilson WP, ed. Application of Electroencephalography in
Psychiatry. Durham: Duke University Press; 1965:54–101.
136. Gibbs EL, Gibbs FA. Electroencephalographic evidence of thala-
mic and hypothalamic epilepsy. Neurology (Minneapolis). 1951;1:
136–144.
137. Gibbs FA, Gibbs EL. Fourteen and six per second positive spikes.
Electroencephalogr Clin Neurophysiol. 1963;15:353–358.
138. Henry EE. Positive spike discharges in the EEG and behavior
abnormality. In: Glaser GE, ed. EEG and Behavior. New York:
Basic Books; 1963:315–344.
139. Hughes JR, Gianturco D, Stein W. Electro-clinical correlations in
the positive spike phenomenon. Electroencephalogr Clin
Neurophysiol. 1961;13:599.
140. Millen FJ, White B. Fourteen and six per second positive spike
activity in children. Neurology (Minneapolis). 1954;4:541–549.
141. Niedermeyer E, Knott JR. Über die Bedeutung der 14 and 6/sec-
positiven Spitzen im EEG. Arch Psychiatr Nervenkrankh.
1961;202:266–280.
142. Poser CM, Ziegler DK. Clinical significance of 14 and 6 per second
positive spike complexes. Neurology (Minneapolis). 1958;8:903–912.
143. Refsum S, Presthus J, Skulstad AA, et al. Clinical correlates of the
14 and 6 per second positive spikes. An electroencephalographic
and clinical study. Acta Psychiatr Scand. 1960;35:330–344.
144. Schwade ED, Geiger SG. Matricide with electroencephalographic
evidence of thalamic or hypothalamic disorder. Dis Nerv Syst.
1953;14:18–20.
145. Shimoda Y. The clinical and electroencephalographic study of the
primary diencephalic epilepsy or epilepsy of brain stem. Acta
Neurovegetat (Vienna). 1961;23:181–191.
146. Stephenson WA. Intracranial neoplasm associated with fourteen
and six per second positive spikes. Neurology (Minneapolis).
1951;1:372–376.
147. Walter RD, Colbert EG, Koegler RR, et al. A controlled study of
the fourteen-and-six-per-second EEG pattern. Arch Gen
Psychiatr. 1960;2:559–566.
148. Long MT, Johnson LC. Fourteen and six positive spikes in
nonpatient population. Neurology (Minneapolis). 1967;17:316
[abstract].
149. Little SC. A general analysis of the fourteen and six per second
dysrhythmia. In: Proceedings of the 6th International Congress of
Electroencephalography and Clinical Neurophysiology, Vienna.
Vienna: Wiener Med. Akad.; 1965:313–315.
150. Beydoun A, Drury I. Unilateral 14 and 6 Hz positive bursts.
Electroencephalogr Clin Neurophysiol. 1992;82:310–312.
560 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
151. Gibbs FA, Gibbs EL. Atlas of Electroencephalography. Vol 4. Menlo
Park, CA: Addison-Wesley; 1978.
152. Yürüker N, Menzi W. 14 and 6/s positive-Spikes im Schlaf nach
akustischem Reiz. Z EEG-EMG. 1971;2:121–124.
153. Okada S, Kato T, Miyashita K, et al. 14 and 6 Hz positive spikes:
relationship with a negative spiky phase in the frontal area.
Electroencephalogr Clin Neurophysiol. 1990;75:S108 [abstract].
154. Niedermeyer E, Ray CD, Walker AE. Depth studies in a
patient with fourteen and six per second positive spikes.
Electroencephalogr Clin Neurophysiol. 1967;22:86–89.
155. Silverman D. Fourteen and six per second positive spike pattern
in a patient with hepatic coma. Electroencephalogr Clin
Neurophysiol. 1964;16:395–398.
156. Ford RG, Freeman AM. Positive spike bursts in a comatose adult.
Electroencephalogr Clin Neurophysiol. 1982;53:29P [abstract].
157. Drury I. 14- and 6-Hz positive bursts in diverse encephalopathies
of childhood. Electroencephalogr Clin Neurophysiol. 1989;72:12P
[abstract].
158. Lipman IL, Hughes JR. Rhythmic mid-temporal discharges. An
electroclinical study. Electroencephalogr Clin Neurophysiol.
1969;27:43–47.
159. Egli M, Hess R, Kuritzke G. Die Bedeutung der “rhythmic mid-
temporal discharges.” Z EEG-EMG. 1978;9:74–85.
160. Hughes JR. The continuous rhythmic mid-temporal discharge.
Clin Electroencephalogr. 2001;32:10–13.
161. Garvin JS. Psychomotor variant pattern. Dis Nerv Syst. 1968;
29:59–76.
162. Boutros NN, Hughes JR, Weiler M. Psychiatric correlates of the
rhythmic mid-temporal discharges and the 6/sec spike and wave
complexes. Biol Psychiatr. 1986;21:94–99.
163. Naquet R, Louard C, Rhodes J, et al. À propos de certaines décharges
paroxystiques du carrefour temporo-pariéto-occipital: leur activa-
tion par l’hypoxie. Rev Neurol (Paris). 1961;105:203–207.
164. Herranz F, Lopez S. Subclinical paroxysmal rhythmic activity.
Electroencephalogr Clin Neurophysiol. 1984;4:419–442.
165. Westmoreland BF, Klass DW. A distinctive rhythmic EEG discharge
of adults. Electroencephalogr Clin Neurophysiol. 1981;51:186–191.
166. Westmoreland BF, Klass DW. Unusual variants of subclinical rhyth-
mic electrographic discharge of adults (SREDA). Electroencephalogr
Clin Neurophysiol. 1997;102:1–4.
167. Miller CR, Westmoreland BF, Klass DW. Subclinical rhythmic
EEG discharge of adults (SREDA): further observations. Am J
EEG Technol. 1985;25:217–224.
168. O’Brien TJ, Sharbrough FW, Westmoreland BF, et al. Subclinical
rhythmic electrographic discharges of adults (SREDA) revisited: a
study using digital EEG analysis. J Clin Neurophysiol. 1998;15:
493–501.
169. Ajmone Marsan C, Ralston BL. The Epileptic Seizures. Springfield,
IL: Charles C Thomas; 1957.
170. Gastaut H, Broughton R. Epileptic Seizures. Springfield, IL:
Charles C Thomas; 1972.
171. Karbowski K. Das Elekroenzephalogramm im epileptischen Anfall.
Bern: Huber; 1975. ^
172. Karbowski K. Epileptische AnfSlle. Berlin: Springer; 1985.
173. Oller-Daurella L, Oller-Ferrer-Vidal L. Atlas de Crisis Epilepticas.
Barcelona: Geigy Division Farmaceut; 1977.
174. Ralston BL. The mechanism of transitions of interictal spiking
foci into ictal seizure discharges. Electroencephalogr Clin
Neurophysiol. 1958;10:217–232.
175. Ralston BL, Papatheodorou CA. The mechanisms of transition of
interictal spiking foci into ictal seizure discharges. Part II.
Observations in man. Electroencephalogr Clin Neurophysiol.
1960;12:297–304.
176. Anziska B, Cracco R. Changes in frequency and amplitude in
electrographic seizure discharges. Clin Electroencephalogr.
1977;8:206–210.
177. Geiger LR, Harner RN. EEG patterns at the time of focal seizure
onset. Arch Neurol (Chicago). 1978;35:276–286.
178. Blume WT, Lemieux JF, Young GB. EEG morphology of early
stage epileptic seizures. Electroencephalogr Clin Neurophysiol.
1983;56:25P [abstract].
179. Blume WT, Young GB, Lemieux JF. EEG morphology of partial
epileptic seizures. Electroencephalogr Clin Neurophysiol. 1984;57:
295–302.
180. Gibbs FA, Lennox WG, Gibbs EL. The electro-encephalogram in
diagnosis and localization of epileptic seizures. Arch Neurol
Psychiatr (Chicago). 1936;36:1225–1235.
181. Gibbs FA, Gibbs EL, Lennox WG. Epilepsy, a paroxysmal cerebral
dysrhythmia. Brain. 1937;60:377–388.
182. Gastaut H, Vigouroux M. Electro-clinical correlations in 500 cases
of psychomotor seizures. In: Baldwin M, Bailey P, eds. Temporal
Lobe Epilepsy. Springfield, IL: Charles C Thomas; 1958:118–128.
183. Klass DW. Electroencephalographic complications in complex par-
tial seizures. In: Penry JK, Daly DD, eds. Complex Partial Seizures
and Their Treatment. New York: Raven Press; 1975:113–140.
184. Niedermeyer E, Fineyre R, Riley T, et al. Myoclonus and the elec-
troencephalogram. A review. Clin Electroencephalogr. 1979;
10:75–95.
185. Hallett M. Myoclonus: relation to epilepsy. Epilepsia (New York).
1985;26(suppl 1):567–577.
186. Gaches J. Les activitiés périodiques en EEG. Rev EEG
Neurophysiol. 1971;1:9–33.
187. Bauer G, Pieber R. Über periodische Komplexe im EEG. Z EEG-
EMG. 1974;5:75–86.
188. Spehlmann R. EEG Primer. Amsterdam: Elsevier; 1981.
189. Balthazar K. Zur Kenntnis der Panencephalitis nodosa (Pette).
Arch Psychiatr Nervenkr. 1944;117:666.
190. Radermecker FJ. Aspects électroencéphalographiques dans trois
cas d’encéphalite subaigue. Acta Neurol Psychiatr Belg.
1949;49:222–232. ^ ^ ^
191. Radermecker FJ. SystZ matique et Z lectroencZphalographie des
^ ^
encZ phalites et encZ phalopathies. Amsterdam: Elsevier; 1956
(Electroencephalogr Clin Neurophysiol., suppl 5).
192. Cobb WA, Hill D. Electroencephalogram in subacute progressive
encephalitis. Brain. 1950;73:392–404.
193. Rabending G, Radermecker FJ. Subacute sclerosing panen-
cephalitis (SSPE). In: Remond A, editor-in-chief. Handbook of
Electroencephalography and Clinical Neurophysiology. Vol 15A
Amsterdam: Elsevier; 1977:28–35.
194. Vitrai J, Czobor P, Marosfi S, et al. A study of periodicity and
structure of EEG complexes in a case of SSPE. Electroencephalogr
Clin Neurophysiol. 1980;50:11–18.
195. Celesia GG. Pathophysiology of periodic EEG complexes in sub-
acute sclerosing panencephalitis (SSPE). Electroencephalogr Clin
Neurophysiol. 1973;35:293–300.
196. Radermecker FJ, Poser CM. The significance of repetitive paroxys-
mal electroencephalographic patterns. Their specificity in suba-
cute sclerosing leukoencephalitis. World Neurol. 1960;1:422–433.
197. Cobb WA. The periodic events of subacute sclerosing leucoen-
cephalitis. Electroencephalogr Clin Neurophysiol. 1966;21:278–294.
198. Cobb WA. Depth recording in subacute sclerosing leucoencephali-
tis. In: Kellaway P, Petersén I, eds. Clinical Electroencephalography

of Children. New York: Grune & Stratton (Stockholm: Almquist
and Wiksell); 1968:275–286.
199. Lombroso CT. Remarks on the EEG and movement disorder in
SSPE. Neurology (Minneapolis). 1968;18:69–75.
200. Bogacz J, Castells C, San Julian J, et al. Nonepidemic progressive
subacute encephalitis (type Van Bogaert). II. Serial EEG abnor-
malities and deep electrography. Acta Neurol Lat-Am. 1959;5:
158–183.
201. Lesse S, Hoefer PFA, Austin JH. The electroencephalogram in dif-
fuse encephalopathies. Arch Neurol Psychiatr (Chicago). 1958;79:
359–375.
202. Storm van Leeuwen W. Electroencephalographical and neuro-
physiological aspects of subacute sclerosing leuco-encephalitis.
Psychiatr Neurol Neurochir (Amsterdam). 1964;67:312–322.
203. Millar JHD, Coey A. The EEG in necrotizing encephalitis.
Electroencephalogr Clin Neurophysiol. 1959;11:582–585.
204. Perier O, Parmentier B, Brihaye J, et al. A case of inclusion body
necrotizing encephalitis. In: Van Bogaert L, Radermecker J, Hozay
J, et al., eds. Encephalitides. Amsterdam: Elsevier; 1961:235–242.
205. Carmon A, Behar A, Beller A. Acute necrotizing encephalitis
presenting clinically as a space-occupying lesion. A clinico-
pathological study of six cases. J Neurol Sci. 1965;2:328–343.
206. Rawls WE, Dyck PJ, Klass DW, et al. Encephalitis associated with
herpes simplex virus. Ann Intern Med. 1966;64:104–115.
207. Adams JH, Jennett WB. Acute necrotizing encephalitis: a prob-
lem in diagnosis. J Neurol Neurosurg Psychiatr. 1967;30:248–260.
208. Upton A, Gumpert J. Electroencephalography in diagnosis of
herpes simplex encephalitis. Lancet. 1970;1:650–652.
209. Gaches J, Arfel G. Certitude et suspicions d’encéphalites herpé-
tiques. Aspects électroencéphalographiques. Encéphale. 1972;
61:510–549.
210. Elian M. Herpes simplex virus encephalitis and EEG. J
Electrophysiol Technol. 1975;1:161–170.
211. Kugler J, Marltin JJ, Radermecker RJ, et al. Periodische Komplexe
im EEG bei nekrotisierender Herpes-Enzephalitis. Z EEG-EMG.
1976;7:63–71.
212. Cobb WA. Evidence on the periodic mechanism in herpes sim-
plex encephalitis. Electroencephalogr Clin Neurophysiol. 1979;
46:345–350.
213. Gupta PC, Seth P. Periodic complexes in herpes simplex
encephalitis. A clinical and experimental study. Electroencephalogr
Clin Neurophysiol. 1973;35:67–74.
214. Fineyre R. Periodic discharge of unusually slow character in
neonatal status epilepticus caused by necrotizing encephalitis
(herpes simplex). Electroencephalogr Clin Neurophysiol. 1979;46:
80 [abstract].
215. Bergey GK, Coyle PK, Krumholz A, et al. Herpes simplex
encephalitis with occipital localization. Arch Neurol (Chicago).
1982;39: 312–313.
216. Creutzfeldt HG. Über eine eigenartige herdförmige Erkrankung
des Zentralnervensystems. Z Ges Neurol Psychiatr. 1920;57:1
(quoted after Kirschbaum (219)).
217. Jakob A. Über eigenartige Erkrankungen des Zentralnervensystems
mit bemerkenswerten anatomischen Befunden (spastische
Pseudosklerose, Encephalomyelopathie mit disseminierten
Degenerationsbeschwerden). Dtsch Z Nervenheilk. 1921;70:132
(quoted after Kirschbaum (219)).
218. Kirschbaum WR. Zwei eigenartige Erkrankungen des
Zentralnervensystems nach Art der spastischen Pseudosklerose
(Jakob). Z Ges Neurol Psychiatr. 1924;92:157 (quoted after
Kirschbaum (219)).
Chapter 27 ■ Epilepsy in Adults and the Elderly 561
219. Kirschbaum WR. Jakob–Creutzfeldt’s Disease. New York:
American Elsevier; 1968.
220. Drury I, Beydoun A. Evolution of periodic complexes in
Creutzfeldt–Jakob disease. Am J EEG Technol. 1996;36:230–234.
221. Tariska P, Gereby G, Majtenyi K. The clinical significance of
periodic lateralized epileptiform discharges (PLEDs).
Electroencephalogr Clin Neurophysiol. 1985;61:11P–12P [abstract].
222. Chiappa K, Young R. The EEG as a definitive diagnostic tool early
in the course of Creutzfeldt–Jakob disease. Electroencephalogr
Clin Neurophysiol. 1978;45:26 [abstract].
223. Chiofalo N, Fuentes A, Galvez S. Serial EEG findings in 27 cases of
Creutzfeldt–Jakob disease. Arch Neurol (Chicago). 1980;37:143–145.
224. Poole EW. Jakob–Creutzfeldt disease. Patients who might have
had this. Electroencephalogr Clin Neurophysiol. 1980;49:89P.
225. Bortone E, Bettoni L, Giorgi C, et al. Reliability of EEG in the
diagnosis of Creutzfeldt–Jakob disease. Electroencephalogr Clin
Neurophysiol. 1994;90:323–330.
226. Alajouanine T, Lecasble R, Remond A. Eléments graphiques
paroxystiques lents de survénue périodiques. Corrélations elec-
trocliniquis. Rev Neurol (Paris). 1955;93:477–478.
227. Barolin G, Scherzer E, Naquet R, et al. Étude électro-clinique des
états de mal épileptiques survenants chez les apoplectiques. Rev
Neurol (Paris). 1962;107:242–243.
228. Chatrian GE. EEG pattern: pseudo-rhythmic recurrent sharp
waves. Its relationship to local cerebral anoxia and hypoxia.
Electroencephalogr Clin Neurophysiol. 1961;13:144 [abstract].
229. Fischer-Williams M. Burst-suppression activity as an indication
of undercut cortex. Electroencephalogr Clin Neurophysiol. 1963;
15:723–724 [abstract].
230. LeBeau FE, Dondey M. Importance diagnostique de certaines
activités électroencéphalographiques lateralisées périodiques ou
à tendance périodique au cours des abscès de cerveau.
Electroencephalogr Clin Neurophysiol. 1959;11:43–58.
231. Pagni CA, Vitale A, Cassinari V. Su alcuni casi di tracciati carat-
terizzati da elementi parassistici a ricorrenza ritmica periodica o
pseudoperiodica. Riv Neurol. 1960;30:580–591.
232. Aminoff MJ, Greenberg DA, Wiekle DJ. Periodic EEG complexes
in infectious mononucleosis encephalitis. Electroencephalogr Clin
Neurophysiol. 1982;53:28P [abstract].
233. Markand ON, Daly DD. Pseudoperiodic lateralized paroxysmal
discharges in the electroencephalogram. Neurology (Minneapolis).
1971;21:975–981.
234. Dauben RD, Adams AH. Periodic lateralized epileptiform dis-
charges in EEG. A review with special attention to etiology and
recurrence. Clin Electroencephalogr. 1974;8:116–124.
235. Bauer G, Aichner F, Hengl W. Der diagnostische Wert periodisch
lateralisierter Komplexe im EEG. Z EEG-EMG. 1981;12:135–141.
236. Naquet R, Vigouroux RA. The EEG in relation to pathology in
simple cerebral ischaemia. In: Remond A, editor-in-chief.
Handbook of Electroencephalography and Clinical Neurophysiology.
Vol 14A. Amsterdam: Elsevier; 1972:38–44.
237. De la Paz D, Brenner RP. Bilateral independent periodic
lateralized epileptiform discharges—clinical significance.
Electroencephalogr Clin Neurophysiol. 1982;53:27P [abstract].
238. Aldrich MS, Pugh JE. Cerebral bigeminy: alternating periodic
epileptiform discharges [abstract]. Electroencephalogr Clin
Neurophysiol. 1984;58:35P.
239. Hughes JR, Taber J, Uppal H. TRI-PLEDs: a case report. Clin
Electroencephalogr. 1998;29:106–108.
240. PeBenito R, Cracco JB. Periodic lateralized epileptiform dis-
charges in children. Ann Neurol. 1979;6:47–50.
562 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
241. Andriola M. PLEDS and bi-PLEDS in children. Electroencephalogr
Clin Neurophysiol. 1982;53:88P [abstract].
242. Ritaccio AL, March G. The significance of BIPLEDs in complex
partial status epilepticus. Am J EEG Technol. 1993;33:27–34.
243. Silbert PL, Radhakrishnan K, Sharbrough FW, et al. Ipsilateral
independent periodic lateralized epileptiform discharges.
Electroencephalogr Clin Neurophysiol. 1996;98:223–226.
244. Chabolla DR, Moore JL, Westmoreland BF. Periodic lateralized
epileptiform discharges in multiple sclerosis. Electroencephalogr
Clin Neurophysiol. 1996;98:5–8.
245. Gandelman-Marton R, Rabey JM, Flechter S. Periodic lateralized
epileptiform discharges in multiple sclerosis: a case report. J Clin
Neurophysiol. 2003;20:117–121.
246. Pohlmann-Eden B, Hoch DB, Cochius JI, et al. Periodic lateralized
epileptiform discharges—a critical review. J Clin Neurophysiol.
1996;13:519–530.
247. Westmoreland BF, Frere RC, Klass DW. Periodic epileptiform
discharges in the midline. J Clin Neurophysiol. 1997;14:495–498.

Convulsive Status Epilepticus
FRANK W. DRISLANE, SUSAN T. HERMAN AND PETER W. KAPLAN
HISTORY
Status epilepticus (SE) may have been described as early as 600
to 700 BC in cuneiform writings, one of which notes: “If the
possessing demon possesses him many times during the middle
watch of the night, and at the time of his possession his hands
and feet are cold, he is much darkened, keeps opening and shut-
ting his mouth, is brown and yellow as to the eye. It may go on
for some time, but he will die” (1,2). Caelius Aurelianus, a
scholar from the late Roman empire, noted that “fits can recur
… even in the same day” (3).
In the 17th century England, Thomas Willis wrote, “… when
as fits are often repeated, and every time grow more cruel, the
animal function is quickly debilitated … [and] …. the vital
function is by little and little enervated, till at length, the whole
body languishing, and the pulse loosened, and at length ceasing,
at last the vital flame is extinguished” (4).
Reports of SE are few through the 18th century, but in the
1800s, the study of epilepsy flourished in Paris at the Salpêtrière
that, with 8000 patients, was Europe’s largest asylum. Physicians
there, including Charcot, provided extensive clinical descrip-
tions of SE. The topic was brought to modern attention in the
university dissertation of Calmeil in Paris, who reported that
the phrase derives from the patients’ use of the term: “C’est ce
que les malades appellent entre eux l’état de mal” (5). [“It’s what
the patients among themselves called a bad condition (of
epilepsy).”] He detailed the illness’s severity and a sequence of
seizures without pause that forebode a poor outcome. The use
in (Latinized) English of SE appeared when Bazire translated
the lectures of Trousseau, who wrote, “You, however, have heard
of circumstances where fits have lasted two or three days, and
ended in death. It is in these cases that one spoke of, at the
Salpêtrière or at Bicêtre, of an état de mal (status epilepticus)”
(6). Rather than simply repeated seizures, SE was then assumed
to be a distinct entity representing the “maximum expression of
epilepsy with its own particular characteristics, although
restricted at the time to tonic–clonic seizures” (1).
In 1903, Clark and Prout described the clinical and patho-
logical appearance of SE in 38 patients, detailing a fusion of
successive convulsions over 2 to 9 days, to the point of coma
and exhaustion, along with changes in respiration, temperature,
and pulse (7). The presentation was “… composed of delirium,
stupor or coma …. and a variety of psychic states, which have
for their basis cortical discharges.” They listed three stages of SE,
with the third described as stupor—likely corresponding to the
nonconvulsive SE that follows generalized convulsive status
epilepticus (GCSE) with “electromechanical dissociation” and
CHAPTER
28
continued electrochemical seizure activity in the brain after ces-
sation of clinically observable convulsions, sometimes now
referred to as “subtle” SE (8).
Relatively little progress in the management of SE was made
until the 20th century. The first modern meetings on SE, the
Marseille Colloquia of 1962 and 1964, promulgated better def-
initions and classification for seizures and SE and suggested
that SE was the prolongation of many different types of epilep-
tic seizures. Gastaut, the organizer of the meetings, stated that
there were “as many forms of SE as there were types of seizures.”
Today, most literature recognizes SE’s own particular patho-
physiology, not always identical to that of prolongation of an
individual seizure (9).
Convulsive SE can be characterized clinically without elec-
trophysiologic data, but Berger’s development of the electroen-
cephalography (EEG) in 1929 led to valuable understanding of
the different forms of SE, especially for nonconvulsive SE (10)
(see Chapter 29). Since the 1960s, increasing scientific and clin-
ical attention has turned to the physiologic underpinnings,
neurochemistry, and pharmacology of SE, most recently with
genetic investigations and functional imaging.
DEFINITIONS
Gastaut defined SE as “an epileptic seizure that is so frequently
repeated or so prolonged as to create a fixed and lasting
epileptic condition” (11). No duration was specified. The
International League against Epilepsy specified “a single epilep-
tic seizure of 30-min duration or a series of epileptic seizures
during which function is not regained between ictal events in a
30-minute period” (12). Although without a definite scientific
basis, 30 minutes has been the standard criterion for duration
in most clinical studies (13).
SE is often a series of individual seizures, without recovery
between seizures. Alternatively, a single convulsion can be very
prolonged, and many convulsive seizures progress to become
nonconvulsive SE, in which the EEG can demonstrate ongoing
(nonconvulsive) seizure activity. If recovery occurs between
individual convulsions, this is considered a clustering of
seizures rather than SE.
While experimental studies on SE show evidence of neuro-
logic damage after 30 minutes, a shorter duration of seizures is
currently felt to warrant urgent treatment. Prospective clinical
trials must define SE as diagnosable expeditiously, and urgent
treatment is a clinical necessity. Thus, the largest trial of differ-
ent antiepileptic drugs (AEDs) for the treatment of GCSE (14)
considered it mandatory to begin treatment (and make the
563
564 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
diagnosis) within 10 minutes. Later, an “operational” definition
of GCSE proposed that 5 minutes was the time by which SE
should be interrupted to avoid morbidity, mortality, or refrac-
tory SE (15). The duration considered to characterize SE has
been (just barely) agreed upon for GCSE; it is reasonable to use
30 minutes as a clinical definition for other types of SE.
PATHOPHYSIOLOGY OF STATUS
EPILEPTICUS
Episodes of SE appear clinically to begin in the same way as indi-
vidual seizures of the same type, but there are many different
types of seizures and SE, and it is rare to capture the beginning of
SE on an individual EEG. Following initiation, almost half of
seizures lasting from 10 to 29 minutes cease spontaneously, with-
out treatment (16), and most individual secondarily generalized
convulsions cease on their own in under 2 minutes and the
patient recovers thereafter (17). Some seizures, however, continue
despite the brain’s intrinsic capacity to terminate shorter seizures,
and they progress to the longer episodes of SE.
It is unclear exactly what facilitates the transition from indi-
vidual seizures to seemingly self-sustaining prolonged seizures
or SE, but this transition appears to depend substantially on the
relative balance of excitatory and inhibitory electrochemical
function at the time—at intracellular, membrane, extracellular,
and interneuronal levels. To a large extent, SE implies a failure
of the typical inhibitory function of the brain in terminating
seizures as usual (18).
There are probably several pathophysiologic mechanisms
that facilitate the continuation of seizures, and in some seizures,
the balance of electrochemical function appears to shift, unfa-
vorably, toward excessive excitatory activity as seizures progress.
In experimental animals, one of the primary cellular changes in
the transition from prolonged seizures to refractory SE is a
marked and progressive impairment of gamma-amino butyric
acid (GABA) agonist–mediated inhibitory function, in turn
favoring recurrent excitatory circuits (19). During seizures, neu-
rotransmitter receptor changes occur almost immediately, with
gene transcription and new peptide production after several
minutes. GABA receptor modification occurs within 45 min-
utes, showing rapid functional plasticity (20). The electrochem-
ical activity of ongoing seizures also induces gene transcription
leading to the production of numerous molecular factors with
excitatory or inhibiting effects on continuing seizures. Among
the many are galanin (which inhibits seizures and SE) (21), sub-
stance P, neuropeptide Y, somatostatin, and dynorphin.
Early in SE, there is a marked and maladaptive diminution
in the number of GABA-A receptors on neuronal surfaces (22),
in large part due to progressive internalization or endocytosis of
those receptors (23). This likely contributes to the reduced
effectiveness of GABA agonists in terminating seizures. As
seizures progress, in both experimental animals and in human
studies, they appear to become resistant to many AEDs (24),
especially those acting through the GABA system, such as ben-
zodiazepines (BDZs) and barbiturates (20), and those drugs
become less useful in treating SE.
As SE continues (concomitant with the decay in GABA
receptor function), there is increasing excitatory activity of glu-
tamate receptors, especially the N-methyl- D-aspartate (NMDA)
receptor, promoting and prolonging the hyperexcitable state
(18,25,26). When glutamate binds to the NMDA receptor,
membrane ion flow can still be blocked by magnesium, but
with sufficient repetitive depolarization (as occurs during
seizures), the magnesium is released and the ion channel opens
more completely in a “use-dependent” facilitation of gluta-
matergic activity (27). NMDA receptor antagonists are not
effective in blocking the initiation of SE, but as seizures con-
tinue, with repetitive depolarization, they appear to become
more effective anticonvulsants in the same “use-dependent”
manner (28). At that point, glutamate antagonists (such as ket-
amine) may become more effective treatments (29).
As seizures continue, stimulation of glutamate (especially
NMDA) receptors leads to excessive sodium and calcium entry
into neurons, and this may cause cell death. Gene activation
during SE may lead to the production of kinases and phos-
phatases that accelerate the harmful process. Later, inflamma-
tion, glial cell changes, and neuronal sprouting or cell loss may
contribute to the refractoriness of subsequent seizures (18).
The potential neuronal injury from prolonged seizures or SE
may depend somewhat on the intensity (or frequency) of the
epileptiform discharges (30).
There is abundant experimental (if much more modest clini-
cal) evidence that neuronal damage results from GCSE, particu-
larly when prolonged. A century ago, this damage was attributed
to ischemia, but in the 1970s, Meldrum and colleagues showed
that excessive excitatory neuronal activity was the cause of nerve
cell damage, at least in experimental animals with prolonged
GCSE (31–33). In baboons given bicuculline to induce SE, pro-
longed convulsions with continued EEG spike discharges for 1 to
5 hours caused neuronal damage in the neocortex, cerebellum,
and hippocampus (31–34). By “clamping” blood flow, brain oxy-
genation, and glucose supply to preserve homeostasis during
SE, they showed that hypoperfusion was not crucial in the devel-
opment of such damage. Neuronal injury still occurred in the
hippocampus, and it appeared to be due to excitatory neurotrans-
mitter activity. Thus, the prolonged ongoing cellular electrochem-
ical activity of SE itself (rather than concomitant physiologic
derangements) was sufficient to cause neuronal injury, although
systemic abnormalities also contributed to this injury. This hip-
pocampal damage may correlate with prolonged memory and
other cognitive dysfunction in humans following SE.
It is very difficult to prove that SE leads to neuronal damage
in humans (as it clearly does in experimental animals exposed to
prolonged, intense electrical stimulation), in part because SE is
seldom fatal without causation by other severe neurologic ill-
nesses that themselves cause neuropathologic damage inde-
pendent of the SE. Nevertheless, a disproportionately decreased
neuronal cell density was seen in the hippocampus of some
patients who died from GCSE (but not NCSE) (35). Forms of SE
other than GCSE may not be as harmful and may not require
exactly the same treatment, but concern for the possibility of
such damage leads to a clear mandate for urgent treatment.
 Chapter 28 ■ Convulsive Status Epilepticus 565

EPIDEMIOLOGY AND CAUSES Tabl e 2 8 . 1

Studies of the incidence of SE vary tremendously by definitions Different Forms of Status Epilepticus
and different populations studied. The corrected, age-adjusted
incidence of SE was 17.1 cases per 100,000 population per year Generalized Focal
in Germany and 54.5/100,000 in subjects greater than 60 years
Convulsive
old; the case fatality rate was 9.3% (36). This study included all
forms of SE. Most were due to remote symptomatic causes such Generalized convulsive Focal motor SE
as cerebrovascular disease; the mean age of adult patients was SE (GCSE)
65 years. SE is more common in children under 1 year and in Primary generalized Epilepsia partialis
the elderly. The incidence of SE is several times that in develop- GCSE continua (EPC)
ing countries, particularly in children (37).
Secondarily generalized
The incidence of all forms of SE was 41/100,000 in
GCSE (focal onset)
Richmond, Virginia, but 20/100,000 in the white population,
similar to that in Germany (38). The incidence was higher in Myoclonic (see Table 28.2)
African Americans and also likely higher because of ascertain- Tonic (may actually start
ment in an urban medical setting. Extrapolation from the same as focal)
database suggests an incidence of up to 50/100,000 per year,
Clonic (may start as
substantially higher than most estimates (13). In another study,
focal seizure)
mortality of SE was 19% within the first 30 days, highly associ-
ated with the acute cause of SE (39). Short-term case fatality is Atonic (very rare for SE)
generally 20% to 25% when considering GCSE alone, and in all Nonconvulsive
series it is primarily due to the underlying illness causing SE
“Classic” absence SE Complex partial SE, with
(13). In another study of patients with afebrile SE, 32% had
prolonged or repeated
recurrence of SE within 10 years of follow-up (40). Recurrence
CP seizures
is higher in patients with progressive neurologic disorders.
Causes of SE are extremely numerous. They may be grouped Other primary generalized
into larger categories similar to the causes of seizures and NCSE
epilepsy syndromes (12): “symptomatic,” that is, a consequence “Atypical” absence SE
of a known or suspected cerebral dysfunction such as head
Other generalized NCSE Other focal SE with
injury, vascular disease, infection, tumor or other mass lesion,
(often secondarily generalized) nonmotor features:
toxic or metabolic disturbances, or following surgery; “idio-
SE in coma; SE in critically for example, aphasic,
pathic,” such as in the idiopathic generalized epilepsies (IGEs),
ill patients; electrographic SE. sensory
almost always related to a known or presumed genetic-based
syndrome; or “cryptogenic,” in which there is a presumed
underlying cause, but it cannot be identified. The IGEs lead to
SE relatively infrequently, and the SE is usually nonconvulsive
(see Chapter 29), except for the unusual case of generalized which clinical syndrome a patient has. SE, like seizures, can be
convulsions upon awakening (41). organized clinically into those with a focal onset and those with
The most common cause of SE in most larger series, espe- a generalized onset—although many generalized cases cannot
cially in nonhospitalized patients, is “remote symptomatic” be divided easily into those of a primarily generalized nature
(with the cause present for more than a week), often due to ear- versus those with a focal onset and secondary generalization
lier injuries such as stroke and often with more immediate (Table 28.1). Most epidemiologic studies do not specify the type
exacerbation by some systemic problem such as infection, med- of SE, but those that do indicate that GCSE is the most com-
ications, or AED withdrawal (24,42). The next largest category mon and accounts for about half of all SE. Forms of SE with
of causes is probably an exacerbation of earlier epilepsy, also prominent motor signs, whether convulsive or myoclonic, are
commonly with an acute precipitant such as lowered AED lev- covered in this chapter; those with primarily nonmotor mani-
els, infection, or metabolic disturbance. In acute care hospitals, festations are described in the Chapter 29.
most SE patients have “acute symptomatic” causes, that is, due
to a recent (within a week) neurologic injury or illness, the most GENERALIZED CONVULSIVE STATUS
common example of which is a new stroke (43). Most have no EPILEPTICUS
history of seizures and are often older adults with acute and
severe medical or neurologic illnesses. Presentation, Clinical Observations
In order to discuss diagnosis, management, and outcome, it GCSE is the best described form of SE and has the greatest mor-
is important to recognize the tremendous variety of clinical sit- bidity, mortality, and clinical urgency for treatment (44). It may
uations that fall under the rubric of SE and to specify exactly begin as primary generalized convulsive seizures or evolve from
566 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
a focal or complex partial seizure and progress to generalized
convulsions. Most GCSE has some evidence of a focal onset or
focal lesion (13), that is, about 70% to 80% is secondarily gen-
eralized (42). Some patients have recurrent seizures without
recovery of consciousness between seizures; others evolve from
the first convulsion into a more prolonged seizure, with motor
manifestations becoming less regular clonic movements and
regular or irregular myoclonic jerking activity, in a continuous
epileptic (“ictal”) state (44). Primary GCSE is relatively uncom-
mon, often occurring as a result of AED discontinuation or the
consequences of inappropriate AEDs or other drugs, or sys-
temic illness in patients with IGEs (41).
Clinically, GCSE is readily recognizable by its motor mani-
festations, which are often dramatic. Typically, convulsive
activity consists of paroxysmal or continuous bilateral, usually
rhythmic, tonic movements or clonic shaking of the limbs,
commonly with facial and truncal manifestations as well, and
with loss of consciousness. Convulsions may have some later-
alized emphasis, even though the physiologic seizure activity is
bilateral or generalized. If the generalized convulsion is pri-
mary, the patient may exhibit no movement for a second or
two, followed by the tonic phase of a seizure in which muscle
contraction occurs throughout the whole body, including a
tightening of chest muscles giving rise to the “epileptic cry.”
The tonic phase typically lasts about 10 seconds (45,46) and is
interrupted by the clonic phase, with particularly rapid and
forceful flexion contractions of the arms. Sometimes move-
ments are more subtle, and there may be a faster quivering.
Typically, there is a progressive lessening of the violence of the
convulsions, often within the first minute of SE. Eventually the
frequency of clonic contractions lessens, and they cease
(45,46). The patient is then flaccid, but there may be a large
generalized myoclonic jerk at the end of the motor phase. The
tonic, and then clonic, phases of an individual seizure seldom
last more than 2 minutes (17). As GCSE progresses, motor
manifestations may lessen and include just relatively subtle
eyelid or facial or other myoclonic jerking, or motor phenom-
ena may disappear entirely (8). Progression to this “subtle”
GCSE (below) is not characteristic of primary GCSE (44).
Repeated seizures also diminish in duration and intensity over
time (44,47).
Consciousness is lost completely at the beginning of GCSE
with the first convulsion, or earlier, and the patient is apneic
during the tonic and clonic phase. The eyes are usually open
and may show pupillary dilation. Tachycardia, hypertension,
and other autonomic changes are common. Following an iso-
lated convulsion, muscle tone and consciousness return gradu-
ally, usually with some improvement within minutes, but in SE
recovery does not occur before the next generalized convulsion
or progression to more subtle or nonconvulsive SE.
THE EEG OF GENERALIZED CONVULSIVE
STATUS EPILEPTICUS
EEG patterns in SE undergo evolution, usually paralleling the
clinical manifestations. Generalized convulsions typically show
bilaterally symmetric activity (behavioral and EEG) from the
onset and may progress through a stereotyped sequence of clin-
ical and electrographic manifestations (8,47) (Fig. 28.1). The
initial brief interruption in behavior is often accompanied by
widespread voltage attenuation with fast frequencies at 20 to 40
Hz, sometimes called “electrodecremental.” This may be a wide-
spread desynchronization of the background, but with
implanted EEG electrode recording, very rapid ( 200 Hz)
activity may be evident in the area of seizure onset (not evident
in routine, surface EEG), suggesting a role for pathologic organ-
ization of networks or circuits of neurons in the generation of
seizures (48). Within seconds, as the initial tonic contraction
occurs, generalized rhythmic electrographic seizure activity,
including rapid spike discharges, gradually increases in voltage
and decreases in frequency to approximately 10 Hz.
Not all episodes of GCSE begin this way. Some start with
early electrographic signs of an IGE that, within seconds,
assumes the clinical and EEG picture of a tonic and then clonic
seizure (Fig. 28.2A–D). Still others begin as focal-onset seizures
and generalize rapidly. Within seconds, all three can appear very
similar clinically and on EEG.
During the clonic phase of the seizure, there is rhythmic
slow activity, sometimes with interspersed spikes, for 10 sec-
onds or more (Fig. 28.2B). Generalized spike and slow waves or
polyspike and slow-wave activity correspond to the frequency
of the clinical manifestations.
Many of the initial EEG rhythms of GCSE are obscured by
the muscle activity of the convulsion, unless the patient has
received paralyzing drugs. It is often only when the motor man-
ifestations lessen that the EEG becomes interpretable.
Sometimes, epileptiform discharges can be seen at the vertex.
The superimposed electromyographic activity, however, may
exhibit its own characteristic rhythmic appearance, synchro-
nous with the clinical jerking activity, and this muscle activity
on the EEG still indicates generalized convulsions (Fig. 28.2C).
If the seizure ends (or between seizures) the discharging
activity gives way to a slow, encephalopathic background in all
areas, or sometimes a very suppressed background, before a
gradual resumption of a normal drowsy and then waking pat-
tern (Fig. 28.2D). In GCSE characterized by repetitive seizures
with impaired consciousness, the intervening EEG shows dif-
fuse slowing, voltage attenuation, and disorganization. Between
seizures, there are often persistent generalized epileptiform dis-
charges with a rhythmic appearance, but slower than during the
convulsion (49).
As GCSE becomes prolonged, the EEG becomes more dis-
continuous, and clinical manifestations become more subtle,
often with relatively minimal eyelid or facial or other myoclonic
jerking, nystagmoid eye movements, or even absence of motor
movements, all while the EEG continues to show repetitive,
rhythmic epileptiform discharges. When the visible motor man-
ifestations of GCSE cease, this can be termed “subtle” GCSE
(below) (8).
Some have proposed that there is a characteristic sequence
of EEG changes during GCSE, based on recordings from ani-
mals and humans during various stages of SE (Fig. 28.3) (8):
(i) Seizures are initially discrete and followed by background
slowing and attenuation until the next seizure begins. (ii)
 Chapter 28 ■ Convulsive Status Epilepticus 567

Fp1-Av12
F3-Av12
C3-Av12
P3-Av12
O1-Av12
Fp2-Av12
F4-Av12

C4-Av12
P4-Av12
O2-Av12

F7-Av12
T3-Av12
T5-Av12

F8-Av12
T4-Av12
T6-Av12

Fz-Av12
Cz-Av12
Pz-Av12
RLC
LUC
75 uV 1 sec

Figure 28.1 A generalized tonic–clonic seizure in a 20-year-old woman with cryptogenic generalized epilepsy. At the
onset there is diffuse frontally maximal low-to-moderate voltage beta activity, followed by diffuse muscle artifact as low-
amplitude clonus begins.

The seizures merge gradually, with waxing and waning of volt-
age and frequency. (iii) Eventually, seizures become continuous,
although ictal discharges are frequently asymmetric, reflecting
the partial onset of many of the seizures. (iv) With time, seizures
become discontinuous; there are brief (0.5 to 8 seconds) periods
of generalized voltage attenuation or flattening. (v) In late SE,
generalized periodic epileptiform discharges (GPEDs) or bursts
of polyspikes with frequencies from 0.5 to 4 Hz arise from a rel-
atively flat background. “Subtle” SE often corresponds to EEG
stages 4 and 5 (8). Patients are typically comatose, with minimal
or no motor manifestations at this point (8,49).
Sequential EEG changes, such as the decreased prominence
of ictal discharges and increase in intervening background
attenuation (in parallel with the diminution of clinical mani-
festations), represent progressive neuronal dysfunction, possi-
bly due to metabolic depletion (8). As the EEG “deteriorates,”
motor manifestations become more subtle or disappear
entirely.
Other investigators have not found the same sequence of
EEG changes (50–52). While these EEG patterns were seen fre-
quently, they often occurred out of sequence or in only some
patients, and some patients persist in one EEG pattern
throughout the course of SE. Evolution of the EEG patterns
depends on the etiology of SE and the baseline condition of
each patient, time from SE onset, type of SE, duration of
seizure activity, and treatment. Such variation makes classifica-
tion of SE based purely on EEG criteria difficult and adds to
the controversy regarding which patterns are “ictal.” Stage 4
above usually has subtle or no motor manifestations but has
very frequent and rhythmic epileptiform discharges and is
usually considered to represent NCSE. Stage 5 is far more con-
troversial as to its ictal nature.
There is no definite difference in clinical significance if SE is
manifested on EEG by recurrent discrete electrographic
seizures or by continuous epileptiform discharges. In one series
of focal SE, there was no significant difference in the etiologies
and outcome of patients with either continuous discharges or
recurrent discrete seizures on EEG (53). In another series,
patients with continuous clinical seizure activity appeared to be
sicker overall than those with discrete seizures (54), but the
management and outcome does not differ much between the
two patterns.
A
Fp1-F3

F3-C3

C3-P3

P3-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

ECG K2-K1
100 uV1 sec

B
Fp1-F3

F3-C3

C3-P3

P3-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

ECG K2-K1
100 uV1 sec

Figure 28.2 A: A generalized tonic–clonic seizure in a 43-year-old man with idiopathic generalized epilepsy and frequent
episodes of status epilepticus. The seizure begins with widespread frontally maximal alpha activity, then 5-Hz spike-and
slow-wave discharges. B: Continuation of the seizure in A. Slower rhythmic activity, with admixed spike and wave dis-
charges, gradually becomes obscured by muscle activity.
C

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

ECG K2-K1
100 uV 1 sec

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

ECG K2-K1
100 uV 1 sec

Figure 28.2 (Continued ) C: After a period of EEG activity nearly entirely obscured by muscle activity (not shown here),
there is evolution into generalized 1.5-Hz bursts of EMG artifact corresponding to the patient’s generalized clonic jerking.
D: The seizure ends and is followed by diffuse background voltage attenuation.
570 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 28.3 Stages of status epilepticus, illus-
trated by single-channel recordings showing EEG
evolution (see text for details). Stage 1: Discrete
seizures with intervening normal EEG. Stage 2:
Merging seizures, with some intervening normal
EEG. Stage 3: Continuous ictal activity. Stage 4:
Discontinuous ictal activity with intervening
periods of a flat background. Stage 5: Somewhat
periodic epileptiform discharges on a low-voltage
background.
COMPLICATIONS, MORBIDITY
GCSE engenders major physiologic changes, several of which
may contribute to its significant potential for morbidity and
mortality (42). At the beginning of GCSE (as with isolated con-
vulsions) heart rate and blood pressure increase rapidly, along
with a substantial surge of catecholamines into the blood
stream. Cardiac arrhythmias, including bradyarrhythmias and
tachyarrhythmias, are common and sometimes serious; cardiac
arrest may occur (55). Blood pressure and glucose concentra-
tions increase early but return to baseline levels after several
minutes and may even decrease further (56). Hypoperfusion
can affect the liver and pancreas. Respiratory rhythm distur-
bances are common, as is aspiration pneumonia, often leading
to fever. Prolonged treatment may be associated with infections
or sepsis, thrombophlebitis, and pulmonary emboli.
Metabolic complications include electrolyte disturbances
such as hyponatremia. Peripheral leukocytosis and cerebrospinal
fluid (CSF) pleocytosis are common. Increased intracranial
pressure and cerebral edema may develop, especially in children
(57). Disseminated intravascular coagulation can lead to multi-
organ failure. Fractures of long bones are common, particularly
with falls and trauma, as are vertebral compression fractures.
Falls also cause facial, dental, skull, and many other injuries.
Violent muscle contractions increase lactate levels, con-
tributing to a systemic acidosis, and systemic temperatures
often rise significantly (42). Prolonged muscle contraction can
also lead to rhabdomyolysis, and myoglobinuria can cause renal
failure. Hyperthermia may be due to continued convulsions or
infection, for which the underlying cause must be sought and
treated. Hyperthermia due to the SE itself usually resolves with
prompt treatment of the SE.
Imaging studies, including CT, MRI, and PET scans, have
shown substantial changes in brain tissue with SE, although
many resolve with time, and their clinical implications are often
uncertain (58). There are some well-detailed cases of perma-
nent and severe tissue damage, such as hippocampal atrophy,
reasonably attributed to very prolonged SE alone—and accom-
panied by neuropsychologic deficits (59).
Studies (most of which are pediatric and retrospective) have
found minimal cognitive morbidity following SE. Several found
negative consequences, but it is unclear whether they are due to
the underlying illness or SE itself (60,61). Comprehensive neu-
rologic and neuropsychologic evaluations before and after SE
are seldom available; patients with progressive illnesses worsen,
whether or not related to the SE; and neurologic deficits may
fluctuate with time. It is also difficult to control for the influ-
ence of AEDs and other medications.
One group studied 193 children with SE, almost all GCSE,
and concluded that “the major neurologic sequelae are usually
due to the underlying insult rather than to the prolonged
seizure itself” (60). Another group performed thorough neu-
ropsychologic testing on 15 adults with prior epilepsy (without
underlying lesions or symptomatic causes) before and after
episodes of SE, finding no significant cognitive morbidity
accruing from the SE and suggesting that the morbidity in
other series comes from the underlying etiology, not from the
SE itself (62). Long-term cognitive complications of SE appear
to be relatively uncommon, occurring primarily in the most
prolonged and severe cases.
About 10% of chronic epilepsy presents with an episode of
SE (63). Conversely, about 30% to 40% of survivors of GCSE
develop subsequent epilepsy (60,64,65), four times the rate fol-
lowing a single acute symptomatic seizure. SE might lead to
epilepsy, but alternatively, patients more likely to have later
epilepsy may present with more severe seizures at the onset, that
is, SE may represent a marker for a more severe epileptic
process that has already begun (66). While chronic epilepsy may
follow SE and SE is a risk factor for such epilepsy (67), it does
not appear to be causal in most.
OUTCOME, MORTALITY
Mortality in “the status epileptici” (with all forms taken together)
is approximately 20% to 25% (13) with most studies consisting
primarily of GCSE. Even for GCSE, outcome varies considerably
and depends heavily on etiology and possibly on other factors. In
one of the largest prospective, population-based studies of SE,

case fatality was 3% for children and 26% for adults—using a 30-
minute definition and including all types of SE (13). Almost all
studies conclude that most of the morbidity and mortality of SE
derives from the illness precipitating the SE rather than from the
SE itself (68,69). By one estimate, no more than 2% of fatalities
comes from the SE itself (70).
Etiology is overwhelmingly the most important prognostic
factor. Short-term mortality is highest when SE occurs in the
setting of an acute insult, especially stroke and anoxia
(13,39,65,71). Patients with multiple medical problems, includ-
ing sepsis, fare poorly, while those with SE due to tumors,
trauma, alcohol abuse, or other drugs have an intermediate or
lower mortality (13,42). Generally, the most favorable etiology
is epilepsy itself with some exacerbating factor (reduced AEDs,
fever, sleep deprivation, or an intercurrent illness or other pre-
cipitant) prompting the SE. Older patients with a symptomatic
cause of GCSE have a mortality of 50% (43). Presentation in
coma (vs. stupor or confusion) confers a poor prognosis (72),
but this may also be related to the severity of the etiology.
Children with SE have a low mortality but the elderly a
much higher one. In one series, SE patients over age 60 had a
mortality of nearly 40% and those over 80 nearly 60% (73).
Most SE types in the elderly are focal, or focal with secondary
generalization (totaling 74%; (73)), reflecting the focal pathol-
ogy, particularly stroke—the most common cause of SE in the
elderly. Younger patients with earlier epilepsy had a much
lower mortality (74). Mortality of SE increases with age, but
this may be explained largely by the increasing incidence of
acute medical illness, such as strokes and other concomitant
conditions (68,75).
The duration of SE may influence outcome. Experimentally
induced seizures in rodents become more refractory to treat-
ment with BDZs after 45 minutes (compared to those lasting
10 minutes) (19). In clinical studies (not always controlled for
etiology) patients with 2 or 4 hours of SE had more morbidity
than those with shorter episodes (24,76). In the largest epi-
demiologic study of duration and SE, patients whose SE lasted
an hour or more had a 32% mortality within 1 month versus
2.7% for shorter durations, even corrected for other factors by
multivariate analysis (77), but there was no definite correlation
between duration and mortality beyond 1 hour. When con-
trolled for etiology and other factors, duration of SE was not a
prognostic factor in at least two studies (39,78), indicating that
prolonged SE should not be considered untreatable—unless
the etiology is particularly ominous. Patients with shorter SE
durations do better, at least in large part, because the SE has
more benign causes. It has been impossible to separate dura-
tion (beyond the first hour) conclusively from the influence of
etiology.
Appropriate treatment of SE appears to reduce morbidity
and mortality, but it is impossible to carry out a randomized
trial of treatment and nontreatment. A study comparing two
centers found that poorer quality treatment (primarily the use
of inadequate BDZ doses early in SE) led to worsened outcome
(79). Another found that most treatment was inadequate and
correspondingly ineffective, also suggesting that adequate doses
of appropriate AEDs led to better outcomes (80).
Chapter 28 ■ Convulsive Status Epilepticus 571
The mortality of SE may be declining over the past few
decades—often attributed to faster recognition and treatment
(81). It is possible, however, that a wider variety of SE types
is now being diagnosed, including more benign cases with
lower morbidity and mortality, thus increasing the reported
incidence of SE and improving the reported outcome. One
study reported a recent decrease in mortality, at least when
myoclonic SE (usually following cardiac arrest and anoxia) was
excluded (81).
TREATMENT OF GENERALIZED
CONVULSIVE STATUS EPILEPTICUS
Clinical management of SE is beyond the scope of a neurophys-
iology chapter but will be alluded to briefly. Although there is a
substantial difference between the duration of SE that leads to
neuronal damage in experimental animals and the shorter
duration of SE now felt to warrant prompt treatment in
patients, seizures should be treated before they last 30 minutes,
and certainly by the time of “impending status” (18). The
“operational” definition above recommended treatment within
5 minutes to avoid morbidity, mortality, or refractory SE (15).
This applies primarily to GCSE.
As an emergency condition, GCSE warrants admission and
intensive care unit (ICU) care in most cases, with urgent atten-
tion to airway, breathing, circulation, and basic medical care,
addressing the underlying cause of SE and the medical illness
causing or worsening it. There has been just one large random-
ized controlled trial of different AEDs for GCSE (14). Based pri-
marily on that trial, several other studies, and common practice,
treatment recommendations usually begin with sufficient doses
of intravenous (i.v.) BDZs, the most rapidly acting medications
(82). With its longer duration of action, lorazepam is frequently
preferred (82–84). This should almost always be followed by i.v.
infusion of a longer acting AED such as phenytoin, fos-pheny-
toin, phenobarbital, valproate, or levetiracetam (82,84–86). If
the first agent is unsuccessful in treating SE, a second drug (often
one of the other i.v. AEDs) is used frequently (87). Often adjunc-
tive use is made of enterally administered carbamazepine, topi-
ramate, and other AEDs unavailable in i.v. form.
In a large series, the treatment success rate in patients with
overt GCSE was 56%, but only 15% with subtle SE (14).
Lorazepam had the greatest success, terminating seizures in
65% of cases of overt GCSE. Phenobarbital (58.2%), diazepam
followed by phenytoin (55.8%), and phenytoin alone (43.6%)
followed, but these differences were not statistically significant
(except for lorazepam vs. phenytoin). If SE continues after the
first AED, a second AED succeeds in stopping SE less than 10%
of the time (and even less with subtle or purely electrographic
SE) (88). At this point, SE is considered refractory, and most
experts move on to more definitive or aggressive treatment with
continuous intravenous (C-IV) AEDs.
Common causes of treatment failure include inadequate dos-
ing of the initial AEDs, failure in diagnosis of the cause or
underlying etiology (which must be treated), inaccurate diagno-
sis with psychogenic nonepileptic seizures, and failure to main-
tain adequate AED therapy after the initial treatment—leading
572 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
to a relapse of SE, a common complication and often one signi-
fying or predicting worsened outcome (89).
REFRACTORY STATUS EPILEPTICUS
Most SE responds to appropriate use of AEDs, but over time,
with continued seizures, SE may become refractory to treat-
ment. The primary causes of refractory status epilepticus (RSE)
are similar to those of SE overall, including stroke or central
nervous system (CNS) tumor (20%), epilepsy-related (20%),
toxic-metabolic encephalopathy (19%), CNS infection (19%),
hypoxia-ischemia (12%), and traumatic brain injury (5%) (90).
The minimal duration of seizure activity to define RSE varies,
from unspecified to 2 hours (84,91,92). Most epileptologists use
the term refractory without regard to temporal duration, but
rather for SE that has not responded successfully to treatment
with two AEDs (some require three), usually including one
BDZ and a longer acting AED—administered promptly and
adequately (75,84,91–93).
The prevalence of RSE varies from 10% to over 30% of all
SE, depending on definitions (84,90). In the VA Cooperative
Study of GCSE treatment, RSE, defined as continued seizures
after two AEDs, occurred in 38% of 384 patients with overt
GCSE and 82% of 134 patients with subtle GCSE (14). In a ret-
rospective study of all types of SE, RSE (defined as lasting more
than 60 minutes, with failure of two AEDs) occurred in 31% of
83 episodes of SE in 74 patients (84). RSE was more common
in patients with nonconvulsive SE (88% vs. 26% with other
forms of SE) and focal motor SE. In children, about 40% of SE
becomes RSE, defined as 60 minutes of seizures unresponsive to
two AEDs (94). In one series of ICU patients (with SE identified
and treated quickly) SE became refractory in 25% and was fatal
in 21%; continuous SE had a higher case fatality rate than inter-
mittent SE (95).
Patients with particularly prolonged SE have been labeled as
having “malignant” SE, or SE persisting despite high-dose anti-
convulsant anesthetics. Many cases occur in young adults with
encephalitis, usually following GCSE with a continuation clini-
cally or electrographically (96–98). They remain in RSE or
relapsing and remitting SE for up to weeks. Malignant SE has a
poor outcome, with mortality rates often over 30% (90,99).
Similarly, the prognosis for most patients treated with pro-
longed courses of pentobarbital or other aggressive treatment
for RSE is dismal, with the very high mortality usually attrib-
uted to the almost invariably severe underlying etiology
(89,92,100,101). Nevertheless, there are several reports of criti-
cally ill patients surviving SE (most of it NCSE) lasting over a
month (93,102,103), and recovery can be good, depending on
the etiology (93). Postanoxic SE, on the other hand, is a fatal
condition in essentially all cases, but hypothermia may help to
save some patients, especially if brainstem reflexes, somatosen-
sory evoked potentials, and some reactivity of the EEG back-
ground are preserved (82,104). A longer duration of SE is rarely
a good sign, but it is also unwarranted to conclude that the sit-
uation is hopeless when SE has continued for hours or days,
especially if the cause is relatively more benign (105).
TREATMENT OF REFRACTORY STATUS
EPILEPTICUS
For refractory SE, ICU admission and mechanical ventilation are
necessary. Most neurologists choose a BDZ followed by a longer
acting AED (106), but more definitive treatment is often neces-
sary. This can be provided by three commonly used i.v. medica-
tions for RSE, all labeled as “coma-inducing” or anesthetic:
midazolam, pentobarbital (or often thiopental in Europe), and
propofol (82,107). In one meta-analysis, pentobarbital had a
lower incidence of short-term treatment failure, breakthrough
seizures, and need for another treatment, but a higher incidence
of hypotension (90). The study concluded that pentobarbital was
more effective in short-term seizure control but was associated
with more adverse effects, without improved outcome. A few
cases require treatment with inhaled anesthetics, usually isoflu-
rane and desflurane, typically used to produce a burst-suppres-
sion pattern (108). The glutamatergic antagonist, ketamine is
another option (109). Anesthetic treatment should generally be
continued for days, or at least for 24 hours after seizures stop,
then tapered gradually, with concurrent EEG monitoring (82).
Many cases of RSE are difficult to treat, but most yield even-
tually to i.v. medications, at times augmented with enteral
AEDs. There remain infrequent cases, sometimes of focal RSE
from focal cortical lesions (such as cortical maldevelopments),
that are amenable to surgical treatment. Typically, the seizure
focus is identified by neuroimaging or electrocorticography
(110), although this can be difficult. When scans are normal,
surgical localization may be guided by ictal single-photon emis-
sion computed tomography (SPECT) (111). In one series of 10
children with focal RSE (all with focal imaging abnormalities)
who failed at least 2 weeks of high-dose suppressive therapy, SE
was terminated by surgery in all, and 7 of 10 became seizure-
free (112). Surgery may consist of focal resection of the epilep-
togenic zone; wider, lobar or even multilobar resections, even to
the point of functional or anatomic hemispherectomy; occa-
sional larger operations such as corpus callosotomy and hemi-
spherectomy, this last operation used most often in cases of
devastating epilepsies of childhood, including Rasmussen
encephalitis (RE); callostomy (113); and multiple subpial tran-
sections (114), with or without focal resections (115).
EEG USE IN REFRACTORY STATUS
EPILEPTICUS
EEG is not necessary to make a diagnosis of GCSE, but if a
patient does not awaken quickly following treatment, an expla-
nation must be sought. In the VA Cooperative study, only 17%
of patients with overt GCSE (and none with “subtle” GCSE)
regained normal alertness within 12 hours of treatment (14).
The cause may be the electrographic persistence of seizure
activity, that is, NCSE—which most epileptologists consider a
continuation of GCSE and warranting urgent treatment.
As GCSE progresses, seizure manifestations become subtle,
or completely absent, and clinical assessment (e.g., cessation of
motor activity) is inadequate in determining treatment effect. It
 Chapter 28 ■ Convulsive Status Epilepticus 573

is often impossible to distinguish whether GCSE has entered a
postictal phase or if the lack of recovery is due to continuing SE.
Both are possible. Electrographic seizures lasting under 30 min-
utes occur in about half of patients after treatment for GCSE
(51). For patients with refractory SE or persistently impaired
mental status, only EEG can distinguish patients who are in
drug-induced coma from the 14% to 20% of patients with con-
tinued nonconvulsive seizure activity (51). Patients who have
received long-acting neuromuscular paralyzing agents also
require EEG for accurate diagnosis.
After diagnosis, and during prolonged intensive treatment
with C-IV AEDs, continuous EEG (C-EEG) is essential to assess
the effectiveness of that treatment and for recognition of later
seizure activity and managing its re-treatment (8). C-EEG is
essential to watch for relapse of nonconvulsive seizures or
NCSE, which is not rare, especially in the first 24 hours (51);
this occurs in up to 68% of patients as C-IV AEDs are tapered
(99). In a retrospective study of RSE treatment, nonconvulsive
seizures were found in 18% of patients within the first 6 hours
of i.v. midazolam infusion, and breakthrough seizures occurred
in 56% of patients later; they were clinically undetectable (i.e.,
purely electrographic) in 89% (99).
Electrographic seizures during C-EEG predict a relapse of
clinical seizures and SE (116,117) and usually warrant an increase
in AED treatment. Most epileptologists re-treat with higher doses
of definitive treatment, or for longer at doses that were successful
earlier, and then have more AEDs (or higher levels of other
AEDs) on board for the next attempt at tapering. This can lead to
a long ICU course. Clearly, these seizures cannot be suppressed
unless they are detected. Isolated epileptiform discharges, how-
ever, do not appear to necessitate more treatment (116).
Continued or recurrent SE should be suppressed for some
time, usually 12 to 24 hours or more (especially during definitive
treatment with midazolam, pentobarbital, or propofol) to ensure
the absence of electrographic (and of course, clinical) seizures.
Nevertheless, the optimal electroclinical goal of treatment (sim-
ple cessation of seizures; both clinical and electrographic seizure
control; or a certain degree of suppression of cerebral activity—
often a burst-suppression pattern (Fig. 28.4)) has never been
studied well prospectively. One retrospective study of patients
with RSE found that patients whose EEGs were suppressed with
pentobarbital to the point of a “flat” background did better than
those with a burst-suppression pattern, but three other patients
whose EEGs showed simply freedom from seizures also survived
(116). The study was retrospective, and patients thought more
likely to survive may have been treated more intensively. Another
study found no clear difference in outcome depending on the
depth of EEG suppression (118).

FP1-F3

F3-C3

C3-P3

P3-O1

FP2-F4

F4-C4

C4-P4

P4-O2

FP1-F7

F7-T3

T3-T5

T5-O1

FP2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
75 uV
2 sec

Figure 28.4 Burst-suppression EEG in a 15-year-old boy with refractory status epilepticus, after treatment with pentobarbital.
574 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
In a meta-analysis, patients treated with the goal of EEG
background suppression (mostly with pentobarbital) had a 4%
chance of breakthrough seizures versus 53% for patients treated
to control clinical and electrographic seizures only (mostly with
midazolam or propofol) (90). Patients treated to EEG back-
ground suppression had a 76% chance of significant hypoten-
sion versus 29% for those treated to suppress seizures only.
Whatever the depth of suppression, mortality was 48%, always
attributed to the severity of the underlying illness causing SE.
The duration of treatment (and monitoring with C-EEG) is
also inadequately studied. Many investigators have used 12 to 24
hours of seizure or EEG suppression and then a taper of medica-
tion, typically over 12 to 24 hours (92,119), but several trials of
midazolam have been shorter (90). One study of patients on
pentobarbital raised the possibility that a prolonged period of
seizure and EEG suppression could be beneficial (116).
During treatment of RSE, EEG should be reviewed immedi-
ately until SE is terminated or a burst-suppression record is
induced with C-IV AEDs. Use of further EEG monitoring is dic-
tated by the patient’s condition. If seizures are suppressed, less
frequent evaluation may be sufficient, while continued or recur-
rent seizures require more intensive monitoring. Optimally, C-
EEG monitoring should be performed and reviewed from the
time aggressive therapy (i.e., general anesthesia) is started, until
seizures are stopped and the patient returns to normal con-
sciousness or until electrographic seizures are controlled for 24
hours. If C-EEG is not available, EEGs might be repeated every
6 to 12 hours, until seizure activity does not return, but this may
miss nonconvulsive seizures.
EEG findings can also help to predict outcome after SE. After
carotid artery distribution strokes, subarachnoid hemorrhage,
meningitis, encephalitis, or closed head injury, a diagnosis of SE
increases the mortality rate above that associated with the pre-
cipitating event itself (120).
There are persistent “ictal” discharges (rhythmic spikes, sharp
waves, or spike-wave discharges) lasting 10 seconds to minutes,
in 48% of patients after control of GCSE (51). Discharges lasting
longer than 30 minutes are associated with higher mortality
rates, even after controlling for age and etiology of SE. Delayed
ictal discharges occurring more than 30 minutes after SE has
been controlled completely did not worsen outcome. Patients
with normal EEGs after SE had no significant morbidity or
mortality.
Periodic lateralized epileptiform discharges (PLEDs) after SE
are associated with an increased mortality, typically attributed to
more severe etiologies (52,121). The presence of GPEDs late in
SE was associated with poor outcome in one study (52), but not
in another (122). Burst-suppression patterns are also associated
with worse outcomes, whether pharmacologically induced or
due to hypoxia (121,123). In children over 2 years, lack of back-
ground reactivity after SE portended a poorer prognosis (124).
Recording EEGs in the ICU is fraught with numerous diffi-
culties (125) (see also Chapter 29). Patients are, by definition,
critically ill and may need to be moved by nursing staff, produc-
ing plentiful artifact during EEG, and electrodes are more likely
to become disconnected. Machinery in the ICU, including bed
equipment, ventilators, fluid infusion pumps, and other moni-
tors, can interfere with EEG recording by introducing electrical
artifact. Also, patients are often sedated with BDZs, barbitu-
rates, general anesthesia, and other medications, with their own
effects, often obscuring EEG findings, or even suppressing EEG
rhythms entirely. Head injury introduces additional challenges,
including lacerations, surgical bandages, skull defects, subgaleal
fluid collections, and edema, plus dressings, drains, and shunts,
all of which can reduce the voltage recorded from the underly-
ing brain and distort waveform morphologies. C-EEG lasting
days require significantly more attention from the EEG technol-
ogist to maintaining integrity of EEG leads than does routine
EEG recording on cooperative patients.
FOCAL STATUS EPILEPTICUS
Focal Status Epilepticus
Focal status epilepticus (FSE) has impressively varied manifesta-
tions, ranging from the readily recognized focal motor seizures
to focal sensory and aphasic seizures that remain markedly
under-diagnosed. FSE without impairment of consciousness or
abnormal movements appears relatively infrequent, but it may
remain under-recognized due to the lack of motor findings. The
clinical manifestations of simple partial status epilepticus
(SPSE) reflect ictal involvement of discrete brain regions, such as
motor, sensory, special sensory, autonomic, psychic, language, or
other cognitive function; its diagnosis requires preserved con-
sciousness (126). The incidence of SPSE has been estimated as
no more than 10 cases per million population per year (127).
This chapter considers examples of FSE with motor manifesta-
tions; those without motor phenomena are considered in the
chapter on nonconvulsive status.
There are many causes of FSE. Infectious (e.g., encephalitis)
and vascular (e.g., stroke, hemorrhage, or vasculitis) are the
most common, but mass lesions, trauma, multiple sclerosis, and
rarely, mitochondrial or degenerative disorders can be respon-
sible (128). Occasionally, benign idiopathic (genetic-based)
focal epilepsies, such as rolandic epilepsy, lead to SE (129).
Medication and metabolic changes may precipitate an episode,
presumably with an underlying focal lesion.
Many cases of focal SE occur in the setting of acute stroke—
in one series causing 41% of all cases of SE—and 61% in the
elderly (71). In another series, 5% of stroke patients had new
seizures and 1/5 of these had SE (130). These patients had a
high mortality rate, just over 50%. In another series, the most
common cause (75%) of FSE was stroke, and most patients had
a history of epilepsy; all except for one with aphasic SE had
motor-predominant seizures, and about half had epilepsia par-
tialis continua (EPC) (131). The median duration of SE was 4
hours, but three patients had EPC continuing for weeks.
Duration did not correlate with outcome, which was deter-
mined primarily by the underlying cause.
The morbidity and mortality of focal SE are determined
largely by etiology. Among 47 FSE patients, acute strokes caused
the only poor (fatal) outcome, in four patients (131). Persistent
morbidity occurred in another 10 patients, attributed to the
underlying lesion (usually stroke) in almost all. One study found
 Chapter 28 ■ Convulsive Status Epilepticus 575

a synergistic (for the worse) effect of acute stroke and SE in 83
patients (71). Rarely, FSE becomes extremely refractory and even
life-threatening and necessitates surgical treatment, for example,
some FSE due to focal cortical dysphasias (111).
In a large epidemiologic study, SE with focal features had the
same mortality as generalized SE (20% to 30%) (77), but this
large database included (relatively benign) absence SE in the
generalized group and patients with secondarily GCSE in the
focal group. For SPSE without motor manifestations, there is no
evidence of mortality or long-term morbidity.
EEG
In part because a limited volume of brain tissue is involved,
only 20% to 35% of simple partial seizures have an ictal electro-
graphic correlate detectable by scalp EEG (132). The EEG may
be normal or show nonspecific changes such as focal slowing.
In other patients, a variety of ictal patterns may occur, such as
focal high-frequency discharges, rhythmic waveforms with
evolving morphology, and repetitive epileptiform discharges
(Fig. 28.5) (53,128,133). Seizure activity may be either continu-
ous or discontinuous, with discrete seizures showing evolving
frequencies of 2 to 6 Hz lasting 30 to 60 seconds or more (53).
Some EEGs show discrete electrographic seizures with wide-
spread fields that are more pronounced over one hemisphere.
Although clinical manifestations of a seizure may be focal, the
EEG may show generalized discharges later, and an EEG initi-
ated in the middle of SE may be unable to distinguish between
focal and generalized SE.
Epilepsia Partialis Continua
Focal motor SE with very prolonged jerking activity is called
EPC. EPC has been defined as “clonic muscular twitching
repeated at fairly regular intervals in one part of the body for a
period of days or weeks” (134). It includes continuous or
nearly continuous, relatively slow (every 0.5 to 2 seconds),
often somewhat irregular, clonic jerking motions of groups of
muscles on one side of the body (134). EPC can be remarkably
persistent, with episodes lasting days, weeks, or even years
(135). It is readily diagnosed clinically because of its promi-
nent and long-lasting motor manifestations. While EPC has

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

EKG
100 uV 1 sec

Figure 28.5 A 78-year-old woman with colon cancer had repetitive seizures originating in the right temporal lobe, evolv-
ing gradually to clonic jerking of the left face and neck. These seizures occurred every 5 to 10 minutes for over 12 hours,
gradually increasing in duration, becoming focal motor status epilepticus. This EEG is from a period of 3 hours of contin-
uous seizure activity. Rhythmic epileptiform activity is seen in the right posterior temporal region, with contralateral EMG
artifact from the left face and with neck muscle clonus.
576 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
become close to a synonym for focal motor SE, it is this pro-
longed continuous jerking activity that is EPC, while repetitive
isolated discrete seizures are usually termed simply focal motor
SE. The prevalence of EPC is estimated to be under one patient
per million population in the United Kingdom (136).
EPC was described by Kozhevnikov in 1895 in cases of
Russian spring–summer encephalitis, which included spreading
focal motor seizures (137). Most cases involved children, often
appearing months after a presumed infectious illness and with
continuation for years. Similar cases followed other encephali-
tides. EPC in the modern era is far more often the result of
strokes, both hemorrhagic and ischemic, and other vascular
lesions (including venous infarcts) in adults (134,135). Tumors
are a very uncommon cause of new-onset EPC (138). Cortical
migration abnormalities, or heterotopias, and other develop-
mental abnormalities are possible causes and occasionally lead
to refractory focal SE (111). Other causes include focal infec-
tions such as tuberculosis, syphilis, Toxoplasmosis, and HIV,
progressive multifocal leukoencephalopathy, tuberous sclerosis,
subacute sclerosing panencephalitis, Sturge–Weber syndrome,
perinatal injury, congenital abnormalities including cortical
dysplasias, and trauma (136). Unusual causes include multiple
sclerosis; mitochondrial disorders including mitochondrial
encephalopathy with lactic acid and stroke-like episodes
(MELAS) (139); and paraneoplastic syndromes (140). In chil-
dren, RE is the characteristic abnormality causing EPC.
In addition to laboratory studies to look for infectious and
metabolic causes, all patients with new-onset EPC need MRI to
look for focal lesions. There is usually an identifiable brain
lesion, although it may be small. Fluid-attenuated inversion
recovery (FLAIR) sequences are often best at demonstrating the
lesion (141). A particularly important cause for recognition and
treatment is hypernatremia and nonketotic hyperglycemia,
which may precipitate EPC especially in a patient with an ear-
lier focal lesion, most often stroke (142). This EPC is usually
treatable readily with correction of the serum glucose and vol-
ume status. EPC often occurs in patients with chronic neuro-
logic illness and is often relatively stable clinically and not
always in need of urgent intervention.
Pathophysiology
Most investigators conclude that EPC results from focal cortical
epileptic discharges, often involving a relatively small area of
cortex near the rolandic region or motor strip.
Physiologic studies show cortical dysfunction (135,143,144),
with the pathophysiology of epileptogenic foci in hyperexcitable
motor cortex (145). Focal cortical hyperexcitability can be
demonstrated by correlation with focally enhanced somatosen-
sory evoked potentials (145,146). Responsible lesions, however,
may be subcortical and engender hyperexcitability of the overly-
ing cortex, possibly through deafferentation (134,138,147).
EPC can be considered to take two forms (148). Type 1 arises
from a focal, nonprogressive lesion, for example, due to the
classic Russian spring–summer encephalitis. Repetitive jerking
tends to occur in a restricted group of muscles but can progress
to complex partial seizures or generalized convulsions. Type 2 is
associated with progressive brain lesions, and there is a greater
variety of seizures and occasionally, of muscle groups involved.
Type 2 EPC may have multifocal motor activity and may be
associated with other seizure types, more disturbance of the
EEG background, and a progressive neurologic deficit, such as
the hemiparesis of Rasmussen syndrome (below).
Treatment and Outcome
The continued focal jerking of EPC is often refractory to AED
treatment (82). Valproate, clonazepam, and levetiracetam are
common treatments, as are steroids. Intravenous immunoglob-
ulin (IVIG) and plasma exchange are used increasingly (82).
Resective surgery can be curative if the lesion is small enough,
but it is often ineffective (136). Multiple subpial transections
are another possibility when the epileptogenic area is in corti-
cal structures and motor or language function is at risk from
hemispherectomy (115,149,150). Occasionally, EPC resolves
without treatment (136). While EPC may remain refractory,
AEDs may help prevent more dangerous secondarily general-
ized convulsions.
The long-term prognosis for patients with EPC depends pri-
marily on the underlying lesion or illness. It is often poor. Long-
term morbidity of EPC in terms of weakness, sensory and
visual loss, and language dysfunction is substantial, and many
patients have severe cognitive deficits (136). In one series,
nearly half the patients died over a follow-up averaging 3 years,
usually because of the lesion causing EPC (135).
EEG
While EPC is a subtype of SPSE characterized by continuous
motor seizures, the relationship between clinical manifestations
and EEG findings is inconsistent. EPC is often associated with
irregular focal spikes and sharp waves and sometimes with
sharp-and-slow wave discharges with a frequency of 0.5 to 3
Hz, centering on about 1 Hz (Fig. 28.6). Depending on the pop-
ulation studied, epileptiform discharges on scalp recordings
contralateral to the motor manifestations may occur in most or
in a minority of patients (135,136), presumably because the
responsible lesions are small or, more likely, because of the
“inconvenient” orientation of discharges with respect to the
cortical surface. One series reported focal discharges on scalp
EEG in only 22% of patients (136), but additional electrodes
can show focal discharges in up to 71% (135). Discharges not
seen on scalp EEGs may be more evident on corticography or
other invasive recordings (144). Back-averaging of EEG time-
locked to clinical myoclonus can demonstrate a relationship
between scalp-recorded discharges and muscle jerks in 37% to
45% of patients (135,136). PLEDs are seen in 22% to 71% of
patients (135,151).
The EEG background rhythm may remain normal in EPC,
especially in areas away from the lesion, but EPC is often asso-
ciated with structural lesions that produce some focal slowing
or attenuation of the EEG background, if not epileptiform dis-
charges. Under 10% of patients have completely normal EEGs
(135). Clinically, the EEG is often surprisingly unrevealing, but
it is not always necessary for management.
 Chapter 28 ■ Convulsive Status Epilepticus 577

Figure 28.6 A 56-year-old woman with epilepsia partialis continua following resection of a right parietal meningioma.
EEG shows periodic lateralized epileptiform discharges at approximately 2 Hz in the right hemisphere, more posteriorly,
with EMG artifact from rhythmic facial twitching over the left temporal channels.

Rasmussen Syndrome glutamate receptor antibodies (157), immune therapies such as

RE is itself uncommon, but it is a relatively common cause of
EPC. It was first described at the Montreal Neurologic Institute in
1958, usually occurring in children and adolescents and almost
always restricted to one hemisphere (152). It often follows some
earlier, viral infection (153). It has an inflammatory basis and is
posited to represent a persistent and chronic focal encephalitis of
infectious or autoimmune basis (154,155). In most patients, RE
is a progressive illness with neurologic deterioration, often with
unremitting focal seizures, contralateral hemiparesis, and wors-
ening cognitive function. The neurologic deficit may progress for
years and then stabilize, with persistent EPC (Fig. 28.7). For
unclear reasons, some cases stop progressing spontaneously, and
occasionally, the disease is self-limiting (154). On EEG, there is
prominent focal slowing, correlating with the underlying and
widespread inflammatory or immune lesion (156).
The EPC of RE includes jerking at 1 to 2 Hz of both agonist
and antagonist muscles in same area of the body. It is typically
refractory to treatment with AEDs and persists for years. BDZs,
valproate, and ethosuximide may be useful in treating
myoclonic jerks and seizures, but even high doses of AEDs usu-
ally fail to suppress all seizures. Because RE appears related to
high-dose corticosteroids and IVIG have been attempted, with
variable success in the short-term reduction of seizures, but the
hemiparesis persists (158).
Resective surgery becomes necessary in many patients (155)
and is usually the best treatment (150), especially for seizure
control; sometimes there is improved neurologic function post-
operatively. A minority of patients may have seizures reduced
or even eliminated by hemispherectomy, but this may entail
major morbidity. Often, a partial hemispherectomy is the only
helpful treatment (159). Multiple subpial transections, with or
without focal resections, are a possible option (115,149,150).
PERIODIC LATERALIZED EPILEPTIFORM
DISCHARGES
PLEDs were first described by Chatrian as spike or sharp-
and-slow-wave complexes, with repetition rates of about 1 to 2
Hz (151). PLEDs are surface-negative bi-, tri-, or polyphasic dis-
charges consisting of spike, sharp, or polyspike components, with
variable following slow-wave complexes lasting 60 to 600 msec
(mean 200 msec), of 50 to 150 V (occasionally up to 300 V),
578 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Figure 28.7 A: EEG showing periodic lateralized epileptiform dis-

charges in the left parietal and posterior temporal regions, following focal
status epilepticus in a 26-year-old woman with adult onset Rasmussen
encephalitis. B: MRI following status epilepticus in the same patient with
recurrent secondarily generalized seizures shows cortical T2 hyperinten-
sity and focal thickening (edema) of the cortex on the left.
 Chapter 28 ■ Convulsive Status Epilepticus 579

FP1-F3

F3-C3

C3-P3

P3-O1

FP2-F4

F4-C4

C4-P4

P4-O2

FP1-F7

F7-T3

T3-T5

T5-O1

FP2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
75 uV 1 sec

Figure 28.8 Left hemisphere periodic lateralized epileptiform discharges in a 72-year-old woman following a left middle
cerebral artery territory stroke.

usually recurring at 0.5 to 2 Hz (ranging from 0.2 to 3 Hz ); occa-
sionally, intervals are up to 10 seconds (Fig. 28.8). Epileptiform
discharges are broadly distributed over most of one hemisphere
and may reflect to the opposite hemisphere as well. Between dis-
charges, the background activity is usually markedly attenuated
and slow. EEG waveform morphologies stay fairly constant for a
given patient and EEG, but the discharge frequency may decline
with time, and most PLEDs will resolve within weeks (151).
Rarely, discharges persist for months or even years (160).
PLEDs occur most commonly after acute large structural
destructive lesions such as stroke (the most common cause),
tumor, or infection, but also in chronic seizure disorders and static
lesions (160). One series of 147 cases of PLEDs found the etiology
to be stroke in 27%, anoxia or old stroke in 12% each, infection
8%, tumor in 7%, and other in 34% (161). Patients with PLEDs
are often obtunded, with focal neurologic deficits and often focal
motor seizures. EPC is common. Clinical seizures occur in at least
80% of patients with PLEDs, often before the EEG is obtained,
and electrographic seizures in even higher proportions. Many had
prior SE (161,162). Half of patients without prior epilepsy who
survive the acute illness develop long-term epilepsy (163).
Most electroencephalographers do not consider PLEDs to be
“ictal” (i.e., a manifestation of clinical seizures or SE), at least at
the time of the EEG recording (151,163,164). PLEDs are often
considered “the terminal phase of status epilepticus” (161).
Nevertheless, they are highly associated with clinical seizures
and may lie somewhere along an “ictal-interictal continuum”
(165). PLEDs may also occur on limited segments of an EEG
(e.g., 20 to 30 minutes) with definite electrographic seizures later
in the same recording (166) (Fig. 28.9A,B), indicating that
PLEDs may not be “ictal” when seen but are recorded in patients
who have clear seizures at other times (161). Seizures typically
begin with rhythmic fast activity, sometimes with a field distinct
from that of underlying PLEDs; the PLEDs may then disappear.
Patients with “PLEDs plus” (with lower voltage rhythmic epilep-
tiform discharges or other rhythmic patterns between the higher
voltage periodic sharp waves) (167) (Figs. 28.9 and 28.10) or
PLEDs with intermittent electrographic seizures are more likely
to have epileptic seizures (Fig. 28.9B) and should be considered
for more vigorous treatment with AEDS.
Periodic discharges are generally considered ictal if they occur
consistently with stereotyped clinical behavior that appears
epileptic. More rapid discharges (at least 1 Hz and certainly 1.5
Hz) would also be interpreted as representing seizures (and if
prolonged enough, SE) by most EEG-ers. In a patient with confu-
sion, PLEDs or BiPLEDs (bilateral-independent PLEDs) are likely
A
FP1-F3

F3-C3

C3-P3

P3-O1

FP2-F4

F4-C4

C4-P4

P4-O2

FP1-F7

F7-T3

T3-T5

T5-O1

FP2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
75 uV 1 sec

B
Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz
Cz-Pz
EKG

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz
Cz-Pz
EKG 150 uV2 sec

Figure 28.9 A: Same patient as in Figure 28.8. Later EEG shows “PLEDs plus,” a faster repetition rate of PLEDs with inter-
vening faster frequency patterns. B: Same patient; PLEDs evolving into a focal left hemisphere seizure.

Figure 28.10 Another example of “PLEDs plus” (PLEDs with intervening faster components) in a 56-year-old man with
HIV infection and a left temporal ring-enhancing lesion.
to be ictal if the EEG discharges resolve and clinical symptoms
improve after BDZ treatment. Also, some studies have found focal
hyperperfusion on SPECT scans (168,169) or evidence of focally
increased metabolic activity on PET at the time of the PLEDs
(170), suggesting that PLEDs may be ictal in those cases.
Sometimes, PLEDs are definite signs of seizures! One report
of seven patients described recurrent confusional episodes
associated with PLEDs (171). There were no structural lesions,
and EEG discharge intervals were as long as 4 seconds. Clinical
deficits resolved, with slowing of EEG discharges, sponta-
neously or in response to BDZs. Carbamazepine appeared to
prevent recurrences, but patients relapsed when it was
decreased. The authors considered PLEDs an “unusual status
epilepticus of the elderly.” Another group recorded PLEDs dur-
ing clinically well-defined SE, with an EEG evolution in fre-
quency and amplitude terminating in a more typical seizure
pattern—with a favorable response to BDZ treatment (172).
They concluded that PLEDs was an ictal EEG pattern when
there are appropriate clinical signs. PLEDs may be seen during
seizures or SE, and their clinical significance differs in individ-
ual cases. EEG findings must be discussed and reinterpreted in
the clinical setting.
Similar discharges, bilateral-independent pseudoperiodic lat-
eralized epileptiform discharges (BiPLEDs), occur independently
and asynchronously over the two hemispheres (173). Discharges
Chapter 28 ■ Convulsive Status Epilepticus 581
over each hemisphere have different amplitudes, fields, and repe-
tition rates (Fig. 28.11). Background activity is often severely
attenuated bilaterally. BiPLEDs are typically due to anoxia (28%),
CNS infections (28%), chronic seizure disorders (22%), or vascu-
lar causes and are associated more often with impaired mental
status (coma in 72% vs. 24%) and higher mortality rates (61% vs.
29%) than PLEDs, but focal seizures are less common (173).
GPEDs, also considered ictal sometimes and interictal at other
times, are discussed further in the Chapter 29.
MYOCLONIC STATUS EPILEPTICUS
Persistent myoclonus may or may not be epileptic in origin.
Without evidence of an epileptic cause, it is typically (and prob-
ably best) labeled status myoclonicus or sometimes myoclonus
status. It includes prolonged, but frequently nonrhythmic,
myoclonic jerking, usually of large amplitude, often involving
the face, trunk, and limbs, but sometimes multifocal or asyn-
chronous. The cause is usually an acute and severe encephalopa-
thy, often anoxia or possibly metabolic disturbances, particularly
renal failure (174). The EEG may show widespread slowing
indicative of an encephalopathy and have no spikes or sharp
waves correlating with the myoclonic jerks (Fig. 28.12). In more
severe encephalopathies, there may be a burst-suppression pat-
tern or periodic discharges.
582 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz
50 uV 1 sec

Figure 28.11 A 68-year-old woman with herpes encephalitis and status epilepticus. EEG after control of clinical seizures
showed bilateral-independent pseudoperiodic lateralized epileptiform discharges (BiPLEDs), with different and independ-
ent discharge frequencies over the two hemispheres.

Continuing myoclonic jerks with an EEG showing persistent
rapid and rhythmic epileptiform discharges are persuasive for an
epileptic origin, in which case the diagnosis is better termed
myoclonic status epilepticus (MSE) (175). MSE is characterized
by prolonged myoclonic jerking activity, by usual definition over
30 minutes in duration. Myoclonus may be isolated to one par-
ticular muscle group or, particularly in primary generalized
epilepsies, appear bilaterally and symmetrically, particularly in
flexor muscle groups in the arms. Myoclonus may be rhythmic,
but this is not uniform, even with generalized discharges on the
EEG. Pathophysiologically, the myoclonus is of the cortical reflex
type (176). MSE has a remarkable variety of causes although
clinical presentations may be similar from one to the next. The
prognosis depends on the etiology (49,175). Although MSE is
discussed in the convulsive SE chapter because movements are a
prominent component of the SE, some myoclonic jerks are of
minimal amplitude and the MSE can appear largely nonconvul-
sive clinically (177).
An insightful review of the different forms of MSE divides
cases into those due to epilepsy itself and those symptomatic
of other brain illnesses (178) (Table 28.2). The pure, or epilep-
tic, forms include MSE occurring in the primary “idiopathic”
generalized epilepsies (IGEs) in which myoclonus is a promi-
nent manifestation both ictally and interictally, such as juvenile
myoclonic epilepsy (JME) (Fig. 28.13); childhood and juvenile
absence epilepsy; and grand mal seizures upon awakening
(Fig. 28.14) (179,180). With MSE in the IGE syndromes,
episodes may include myoclonic jerking activity up to five
times a second (although often every several seconds), partic-
ularly in proximal muscles, often with bilateral synchrony, and
can last for hours—somewhat surprisingly without necessarily
affecting consciousness (179). Some patients have frequent
prolonged myoclonic seizures with normal consciousness
between individual myoclonic jerks (180), even with frequent
3 to 6-Hz epileptiform discharges on the EEG. Myoclonus can
also consist primarily of eyelid myoclonia (181). The individ-
ual syndrome predicts the frequency of SE (41,182).
In JME, MSE may occur upon awakening and begin with
irregular and isolated myoclonus with an accelerating or
crescendo pattern, ending in SE. Patients with generalized
myoclonic epilepsies (such as JME and some other IGEs) may go
into MSE when medications are changed (175). Sleep depriva-
tion is another common precipitant, and some AEDs may lead to
an exacerbation of epilepsy, sometimes to MSE (Fig. 28.14A,B)
 Chapter 28 ■ Convulsive Status Epilepticus 583

Fp1-F3

F3-C3
C3-P3

P3-O1
Fp1-F7
F7-T3

T3-T5
T5-O1

Fp2-F4
F4-C4
C4-P4
P4-O2

Fp2-F8
F8-T4

T4-T6
T6-O2
Fz-Cz

Cz-Pz
EogL-A1

EogR-A2

ECG
75 uV 1 sec

Figure 28.12 Irregular generalized and multifocal sharp waves, in an 87-year-old man with acute renal failure, altered
mental status, and frequent myoclonic jerking activity. The EEG sharp waves are less rhythmic than would be considered
indicative of seizures by most electroencephalographers, and the background indicates an encephalopathy.

Tabl e 2 8 . 2

Myoclonic Status Epilepticus

“Primary,” in generalized epilepsy syndromes such as absence epilepsy, juvenile myoclonic

epilepsy, and other myoclonic epilepsies
“Secondary” in other epilepsy syndromes:
“Minor” epileptic status (of Brett)
Severe myoclonic epilepsy of infancy
MSE in myoclonic astatic epilepsy, Lennox–Gastaut syndrome
Epilepsy with myoclonic absences
SE with negative myoclonus and brief atonic episodes
“Symptomatic” of acute neurologic illness (e.g., anoxia, injuries, and encephalopathies)

Adapted from Ohtahara S, Ohtsuka Y. Myoclonic status epilepticus. Chapter 63. In: Engel J Jr, Pedley TA, eds. Epilepsy: A
Comprehensive Textbook. 2nd ed. Philadelphia: Lippincott-Raven Publishers; 2008:725–729.
584 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7

F7-T3
T3-T5

T5-O1
Fp2-F8
F8-T4
T4-T6

T6-O2
Fz-Cz
Cz-Pz
LUC
RLC
EKG
100 uV1 sec

Figure 28.13 Myoclonic status epilepticus in an 18-year-old woman with juvenile myoclonic epilepsy. EEG after a night
of sleep deprivation showed irregular bursts of generalized spike-and-wave and polyspike-and-wave discharges for nearly
one hour. Most bursts were associated with clinical myoclonic jerking activity.

(183). Overall, the IGE syndromes lead to MSE relatively infre-
quently, and the MSE is usually easy to treat and the long-term
outcome good (179).
MSE also occurs in other “secondary” epilepsy syndromes in
which myoclonus is not a prominent feature of the interictal
baseline. In these syndromes, the myoclonic jerks are more often
asymmetric and asynchronous than in the IGE syndromes, and
they may be of smaller amplitude and less rhythmic (11).
Secondary MSE causes include the progressive myoclonus
epilepsies caused by storage diseases (Fig. 28.15) such as lipi-
doses and lipofuscinosis, and Lafora disease (184), infectious ill-
nesses such as encephalitis, and severe childhood epilepsies such
as Dravet syndrome (185). MSE is more common in the second-
ary forms of childhood myoclonic epilepsy than in the primary
syndromes (11), and they are frequently refractory to AEDs.
Many episodes of MSE occur in very young children with
early life genetic or structural deficits or inborn errors of
metabolism and severe baseline encephalopathies. These
include Lennox–Gastaut syndrome (LGS), where MSE may be
mixed with other forms of SE such as “myoclonic astatic”
seizures with irregular facial and limb myoclonus and epilepsy
with myoclonic absences (Fig. 28.16). SE is not rare in LGS, but
myoclonus is usually not the dominant feature. MSE in LGS is
often associated with the slow spike and wave EEG patterns
seen in atypical absence SE (see Chapter 29). Altered con-
sciousness may be mixed with myoclonic jerking activity,
which by itself does not necessarily affect consciousness. MSE
also occurs in younger children with infantile spasms and hyp-
sarrhythmia.
Severe myoclonic epilepsy of infancy (SMEI) or Dravet
syndrome may include myoclonic jerking activity, but
impairment of consciousness (“obtundation status”), some-
times prolonged, is usually the greater clinical problem (185).
The irregular myoclonic activity often involves the face or
limbs and may persist for hours. It usually presents in infancy
but may lead to NCSE around the age of 2 years. The EEG
may be normal at first but later includes focal or multifocal
epileptiform discharges interictally, with eventual generalized
spike-wave discharges on a very disturbed, slow, and irregu-
lar background (186). The long-term outcome is often poor,
reflective of the underlying illness rather than because of the
MSE itself.
Finally, MSE may occur as a manifestation of underlying
encephalopathies unrelated to an earlier epilepsy syndrome,
A
FP1-F3

F3-C3
C3-P3

P3-O1
FP2-F4
F4-C4
C4-P4
P4-O2
FP1-F7
F7-T3
T3-T5

T5-O1
FP2-F8
F8-T4
T4-T6

T6-O2
Fz-Cz
Cz-Pz
LUC
RLC
EKG 200 uV
1 sec

B
Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz
Cz-Pz
LUC
RLC
EKG

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz
Cz-Pz
LUC
RLC
EKG
Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz
Cz-Pz
LUC
RLC
500 uV
EKG 2 sec

Figure 28.14 A: EEG during myoclonic status epilepticus in a 15-year-old girl with epilepsy with generalized tonic–clonic
seizures on awakening. She had no earlier history of myoclonus but was admitted with nearly continuous myoclonic jerk-
ing after being started on carbamazepine. B: Same patient as in A. After 20 minutes, the individual generalized myoclonic
jerks evolve into rhythmic generalized clonic jerking (clonic status epilepticus).
586 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz
50 uV 1 sec

Figure 28.15 Myoclonic status epilepticus in 32-year-old man with progressive myoclonus epilepsy due to
Unverricht–Lundborg disease. The background activity is slow and disorganized, and EEG shows generalized and multifo-
cal spikes and polyspikes. Clinically, the patient was more confused than at baseline and showed continuous fragmentary
myoclonus.

particularly in severe encephalopathies due to anoxia or meta-
bolic derangements (Fig. 28.17) (174,175,187). The MSE, per
se, is usually less important than the devastation caused by the
encephalopathy, particularly in the case of anoxia. MSE may be
part of an acute postanoxic syndrome following cardiac arrest,
often beginning 8 to 24 hours after the insult, with continuous
irregular generalized or fragmentary myoclonic jerks (188).
This MSE is often refractory to therapy and almost always has a
poor outcome (175,189,190). Sometimes, the MSE may be
abolished by AEDs, but without improvement in outcome.
Chronic postanoxic action myoclonus is the Lance–Adams syn-
drome: segmental or generalized myoclonic jerks typically pre-
cipitated by action or stimulation (191,192). EEG patterns vary
but often include irregular centrally maximal polyspike-and-
wave discharges.
The EEG can be extremely helpful in distinguishing among
different causes of MSE, whether related to a particular primary
or secondary epilepsy syndrome (in which the EEG findings are
often an important component of the syndrome) or due to a
nonepileptic encephalopathy (Figs. 28.13–28.17). In the IGE
syndromes, the EEG is often definitive and may show frequent
generalized, frontally predominant spikes and polyspikes with
faster frequencies of 4 Hz or so interictally on a normal back-
ground (Fig. 28.13). Spike discharges may precede the
myoclonic jerking activity immediately (178). During seizures,
there are generalized, frontally maximal irregular spike-and-
wave or polyspike-and-wave discharges, sometimes correlating
with clinical myoclonic jerks, occurring singly or in brief clus-
ters (193). In MSE, the EEG often shows intermittent 1- to 2-
second bursts of epileptiform discharges, with an intervening
normal background.
In secondary epilepsy forms, the EEG background is often
slow, reflecting the encephalopathy (Fig. 28.15). Spike-wave dis-
charges may come in irregular bursts, often with a slower 2 to
2.5 Hz appearance. In both primary and secondary cases, there
is often a poor correlation between the epileptiform discharges
and the myoclonic jerking activity. In symptomatic generalized
epilepsies with MSE, such as LGS, there is usually a background
encephalopathy and more prolonged irregular spike-and-wave
or polyspike-and-wave discharges, but with prominent inter-
vening frontally maximal delta activity (Fig. 28.16). In MSE due
to encepahalopathies, the EEG might have no simple spike-
 Chapter 28 ■ Convulsive Status Epilepticus 587

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

Figure 28.16 Generalized polyspikes at approximately 2 Hz during myoclonic status epilepticus in a 16-year-old boy with
Lennox–Gastaut syndrome.

wave discharges but often shows somewhat arrhythmic, slower
spikes on a slow background. In postanoxic MSE, the EEG
background shows severe voltage attenuation or a burst-sup-
pression pattern, often with superimposed GPEDs (Fig. 28.17).
The characteristic suppressed (often nearly flat) background
augurs poorly for prognosis.
Treatment
Myoclonic SE in the IGE syndromes usually responds well to
AEDs (175), and episodes are often interrupted expeditiously
by BDZs (178). Intravenous valproate may also be effective
(193), and enteral ethosuximide is another possible treatment.
Primary generalized epilepsies with prominent myoclonic
components may be aggravated by use of inappropriate AEDs,
particularly including sodium channel-blocking agents such
as carbamazepine and phenytoin, as well as those with puta-
tive GABA-ergic mechanisms including gabapentin, tiagabine,
and others (194,195).
Appropriate management of other forms of MSE depends
overwhelmingly on etiology and can be much more difficult.
The MSE is often interruptible, but the prognosis corresponds
to that of the underlying syndrome or illness. For MSE in the
setting of anoxia, BDZs and valproic acid are among medica-
tions that can diminish the myoclonic jerking, but they do not
alter the ultimate outcome.
Prognosis and Outcome
The prognosis in status myoclonicus (without evidence of
epileptic pathophysiology) follows the etiology. It is particularly
ominous after anoxia but also poor in metabolic encephalopathies
and infection (174). In one study of 107 patients with comatose
after cardiac resuscitation, myoclonus status was present in 37%
(188). All patients died, although 70% of resuscitated patients
without myoclonus status also died. The authors argued that
this was a reliable sign that anoxic status myoclonicus had a
hopeless prognosis and should not be treated as epilepsy. There
have been a few reports of patients recovering from status
myoclonicus, but most of those were resuscitated almost imme-
diately after primarily respiratory (rather than cardiac) failure,
and several had a subsequent Lance–Adams syndrome, with
action myoclonus, but with a reasonably favorable cognitive
outcome (191,196,197).
The outcome of MSE is also determined strongly by the eti-
ology. Patients with MSE in the IGE syndromes tend to
respond to AEDs well, and the prognosis is usually good. In
“secondary epileptic syndromes” such as the myoclonic
588 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2
75 uV 1 sec

Figure 28.17 A 50-year-old man with myoclonic status epilepticus after cardiac arrest. Both generalized polyspikes
and EMG artifact from myoclonic jerks can be seen. The EEG between myoclonic jerks shows severe diffuse voltage
attenuation.

epilepsies of childhood, episodes can remit and the patient
returns to baseline—but those baselines may be quite abnor-
mal. In storage diseases and progressive illness, the SE may
stop, but the underlying disease progresses. The prognosis is
worst for patients with MSE due to an acute new illness and is
particularly poor following anoxia (189). Anoxic MSE is
nearly always fatal—determined by the anoxic damage, rather
than by any epileptic phenomenon (190). Under 5% recover
reasonably, but hypothermia may improve the prognosis
somewhat (104).
CLONIC AND TONIC STATUS EPILEPTICUS
Clonic SE is relatively rare in adults and is seen primarily in
infants and children. EEG patterns are variable; some show
rhythmic bilateral bursts of high-amplitude delta activity with
intermixed spikes or polyspikes (11) (Fig. 28.14), often occur-
ring synchronously with clonic jerking.
Generalized tonic status epilepticus (TSE) occurs predom-
inantly in young patients with major neurologic and cogni-
tive deficits or other severe encephalopathies from birth or
early childhood (198). TSE is rare in adults in the absence of
other seizure types. Infrequently, TSE arises in patients with
IGEs (199).
TSE consists of maintenance of a tonic posture, particularly
of axial and often, proximal limb, musculature rather than con-
vulsions (198). In clinically obvious cases, there are usually brief
(10 to 15 seconds) tonic contractions of the face and axial mus-
culature, in the extremities or in combinations of these. Motor
activity may be very subtle, showing only mild tonic contrac-
tions of paraspinal musculature or upward deviation of the
eyes. Tonic seizures usually last less than 10 seconds but can
recur hundreds of times in a day. TSE can be difficult to inter-
rupt with AEDs (200), and BDZs may precipitate or worsen
TSE in some cases (201).
EEG may be necessary to make the diagnosis of TSE, given
the frequently subtle presentation. It often shows widespread
symmetric fast activity or very rapid spikes but may also include
periods of background suppression or attenuation, generalized
electrodecremental activity (voltage suppression), or brief runs
of low-voltage generalized fast activity (Fig. 28.18). The
amplitude of the ictal discharge gradually increases and the
frequency decreases, sometimes followed by rhythmic general-
ized spikes (47).
 Chapter 28 ■ Convulsive Status Epilepticus 589

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Ecg k2-K1
100 uv 1 sec

Figure 28.18 Tonic status epilepticus in a 48-year-old man with Lennox–Gastaut syndrome. Forty-eight hours after initi-
ation of rufinamide, he developed recurrent tonic seizures every 5 minutes, without recovery to baseline between seizures.
EEG showed a high-voltage generalized slow wave, followed by a generalized electrodecremental period with superim-
posed EMG activity.

REFERENCES
1. Shorvon S. The concept of status epilepticus and its history.
Chapter 1. In: Shorvon S, ed. Status Epilepticus: Its Clinical Features
and Treatment in Children and Adults. Cambridge: Cambridge
University Press; 1994:1–20.
2. XXV-XXVIth tablet of the Sakkiku cuneiform, 718-612 BC. British
Museum. Kinnier Wilson, JV, Reynolds EH. Translation and analy-
sis of a cuneirform text forming part of a Babylonian treatise on
epilepsy. Medical History. 1990;34,185–198.
3. Temkin O, ed. The Falling Sickness. Baltimore: The Johns Hopkins
Press; 1971:149.
4. Willis T. Pathologiae cerebri et nervosi generis specimen. In: quo
agitur de morbis convulsivis et de scorbuto, 1667, translated by S.
Pordage, 1681. London: Dring. Cited in: Shorvon S. ref 1 above;
Chapter 1.
5. Calmeil LF. De l’épilepsie, étudiée sous le rapport de son siege et
de son influence sur la production de l’aliénation mentale. Paris:
Thèse de Université de Paris, 1824.
6. Trousseau A. Lectures on clinical medicine delivered at the Hôtel
Dieu, Paris, 1868 vol. 1, translated by Bazire, PV. London: New
Sydenham Society.
7. Clark LP, Prout TP. Status epilepticus: a clinical and pathological
study in epilepsy [An article in 3 parts.] Am J Insanity. 1903/4;60:
291–306; 60:645–675; 61:81–108.
8. Treiman DM, Walton NY, Kendrick C. A progressive sequence of
electroencephalographic changes during generalized convulsive
status epilepticus. Epilepsy Res. 1990;5:49–60.
9. Shorvon S. Definition, classification and frequency of status epilep-
ticus. Chapter 2. In: Shorvon S, ed. Status Epilepticus: Its Clinical
Features and Treatment in Children and Adults. Cambridge:
Cambridge University Press; 1994:21–33.
10. Gloor P. The work of Hans Berger. In: Cobb WA, ed. Appraisal and
Perspective of the Functional Explorations of the Nervous System.
Handbook of Electroencephalography and Clinical Neurophysiology.
Amsterdam: Elsevier Publishing Co.; 1971:IA11–IA24.
11. Gastaut H. Classification of status epilepticus. Adv Neurol.
1983;34:15–35.
12. Commission on Epidemiology and Prognosis, International
League Against Epilepsy. Guidelines for Epidemiologic Studies on
Epilepsy. Epilepsia. 1993;34:592–596.
13. DeLorenzo RJ, Hauser WA, Towne AR, et al. A prospective, popula-
tion-based epidemiologic study of status epilepticus in Richmond,
Virginia. Neurology. 1996;46:1029–1035.
590 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
14. Treiman DM, Meyers PD, Walton NY, et al. A comparison of four
treatments for generalized convulsive status epilepticus. N Engl J
Med. 1998;339:792–798.
15. Lowenstein DH, Bleck T, Macdonald RL. It’s time to revise the def-
inition of status epilepticus. Epilepsia. 1999;40:120–122.
16. DeLorenzo RJ, Garnett LK, Towne AR, et al. Comparison of status
epilepticus with prolonged seizure episodes lasting from 10 to 29
minutes. Epilepsia. 1999;40:164–169.
17. Theodore WH, Porter RJ, Albert P, et al. The secondarily general-
ized tonic-clonic seizure: a videotape analysis. Neurology. 1994;44:
1403–1407.
18. Chen JWY, Wasterlain CG. Status epilepticus: pathophysiology
and management in adults. Lancet Neurol. 2006;5:246–256.
19. Kapur J, Macdonald RL. Rapid seizure-induced reduction of ben-
zodiazepine and Zn 2 sensitivity of hippocampal dentate gran-
ule cell GABA-A receptors. J Neurosci. 1997;17:7532–7540.
20. Kapur J. Pathophysiology of nonconvulsive status epilepticus.
Chapter 7. In: Kaplan PW, Drislane FW, eds. Nonconvulsive Status
Epilepticus. New York: Demos Medical Publishing; 2009:81–94.
21. Mazarati A, Lu X. Regulation of limbic status epilepticus by hip-
pocampal galanin type 1 and type 2 receptors. Neuropeptides. 2005;
39:277–280.
22. Naylor DE, Liu H, Wasterlain CG. Trafficking of GABA(A) recep-
tors, loss of inhibition, and a mechanism for pharamacoresistance
in status epilepticus. J Neurosci. 2005;25:7724–7733.
23. Goodkin HP, Yeh JL, Kapur J. Status epilepticus increases the intra-
cellular accumulation of GABA receptors. J Neurosci. 2005;25:
5511–5520.
24. Lowenstein DH, Alldredge BK. Status epilepticus at an urban pub-
lic hospital in the 1980’s. Neurology. 1993;43:483–488.
25. Bertram EH, Lothman EW. NMDA receptor antagonists and lim-
bic status epilepticus: A comparison with standard anticonvul-
sants. Epilepsy Res. 1990;5:177–184.
26. Wasterlain CG, Liu H, Mazarati AM, et al. Self-sustaining status
epilepticus: a condition maintained by potentiation of glutamate
receptors and by plastic changes in substance P and other peptides
neuromodulators. Epilepsia. 2000;41:134–143.
27. Fountain NB. Cellular damage and the neuropathology of status
epilepticus. Chapter 9. In: Drislane FW, ed. Status Epilepticus, a
Clnical Perspective. Totowa, NJ: Humana Press; 2005:181–193.
28. Williamson JM, Lothman EW. The effect of MK-801 on kindled
seizures: implications for use and limitations as an antiepileptic
drug. Ann Neurol. 1989;26:85–90.
29. Mazarati AM, Wasterlain CG. N-methyl-D-aspartate receptor
antagonists abolish the maintenance phase of self-sustaining sta-
tus epilepticus in rat. Neurosci Lett. 1999;265:187–190.
30. Lowenstein DH, Shimosaka S, So YT, et al. The relationship between
electrographic seizure activity and neuronal injury. Epilepsy Res.
1991;10:49–54.
31. Meldrum BS, Horton RW. Physiology of status epilepticus in pri-
mates. Arch Neurol. 1973;28:1–9.
32. Meldrum BS, Brierley JB. Prolonged epileptic seizures in primates.
Ischemic cell change and its relation to ictal physiological events.
Arch Neurol. 1973;28:10–17.
33. Meldrum BS, Vigouroux RA, Brierley JB. Systemic factors and
epileptic brain damage. Arch Neurol. 1973;29:82–87.
34. Lothman EW, Bertram EH, Bekenstein JW, et al. Self-sustaining
limbic status epilepticus induced by ‘continuous’ hippocampal
stimulation: electrographic and behavioral characteristics. Epilepsy
Res. 1989;3:107–119.
35. DiGiorgio CM, Tomiyasu U, Gott PS, et al. Hippocampal pyrami-
dal cell loss in human status epilepticus. Epilepsia. 1992;33:23–27.
36. Knake S, Rosenow F, Vescovi M, et al. Incidence of status epilepti-
cus in adults in Germany: a prospective, population-based study.
Epilepsia. 2001;42:714–718.
37. Sadarangani M, Seaton C, Scott JAG, et al. Incidence and outcome
of convulsive status epilepticus in Kenyan children: a cohort study.
Lancet Neurol. 2008;7:145–150.
38. DeLorenzo RJ, Pellock JM, Towne AR, et al. Epidemiology of sta-
tus epilepticus. J Clin Neurophysiol. 1995;12:316–325.
39. Logroscino G, Hesdorffer DC, Cascino G, et al. Short-term mor-
tality after a first episode of status epilepticus. Epilepsia. 1997;38:
1344–1349.
40. Hesdorffer DC, Logroscino G, Cascino GD, et al. Recurrence of
afebrile status epilepticus in a population-based study in Rochester,
Minnesota. Neurology. 2007;69:73–78.
41. Shorvon S, Walker M. Status epilepticus in idiopathic generalized
epilepsy. Epilepsia. 2005;46:73–79.
42. Aminoff MJ, Simon RP. Status epilepticus: causes, clinical features,
and consequences in 98 patients. Am J Med. 1980;69:657–666.
43. Barry E, Hauser W. Status epilepticus: the interaction of epilepsy
and acute brain disease. Neurology. 1993;43:1473–1478.
44. Treiman DM. Electroclinical features of status epilepticus. J Clin
Neurophysiol. 1995;12:343–362.
45. Gastaut H, Broughton R. Epileptic seizures. Clinical and electro-
graphic features, diagnosis and treatment. Charles C Thomas.
Springfield, IL, 1972.
46. Niedermeyer E. Epileptic seizure disorders. Chapter 27. In:
Niedermeyer E, Lopes da Silva F, eds. Electroencephalography.
Basic Principles, Clinical Applications, and Related Fields. 5th ed.
Philadelphia, PA: Lippincott, Williams & Wilkins; 2005:
505–619.
47. Roger J, Lob H, Tassinari CA. Status epilepticus. In: Vinken P, Bruyn
G, eds. Handbook of Clinical Neurology, Vol 15, The Epilepsies.
Amsterdam: North Holland Publishing Co.; 1974:145–188.
48. Bragin A, Engel J Jr, Wilson CL, et al. Hippocampal and entorhinal
cortex high-frequency oscillations (100–500 Hz) in human epilep-
tic brain and in kainic acid-treated rats with chronic seizures.
Epilepsia. 1999;40:127–137.
49. Drislane FW, Schomer DL. Clinical implications of generalized
electrographic status epilepticus. Epilepsy Res. 1994;19:111–121.
50. Lowenstein D, Aminoff M. Clinical and EEG features of status
epilepticus in comatose patients. Neurology. 1992;42:100–104.
51. DeLorenzo RJ, Waterhouse EJ, Towne AR, et al. Persistent noncon-
vulsive status epilepticus after the control of convulsive status
epilepticus. Epilepsia. 1998;39:833–840.
52. Nei M, Lee JM, Shanker VL, et al. The EEG and prognosis in status
epilepticus. Epilepsia. 1999;40:157–163.
53. Drislane FW, Blum AS, Schomer DL. Focal status epilepticus: clin-
ical features and significance of different EEG patterns. Epilepsia.
1999;40:1254–1260.
54. Waterhouse EJ, Garnett LK, Towne AR, et al. Prospective popula-
tion-based study of intermittent and continuous convulsive status
epilepticus in Richmond, Virginia. Epilepsia. 1999;40:752–758.
55. Boggs JG, Painter JA, DeLorenzo RJ. Analysis of electrocardio-
graphic changes in status epilepticus. Epilepsy Res. 1993;14:87–94.
56. Lothman E. The biochemical basis and pathophysiology of status
epilepticus. Neurology. 1990;40(suppl 2):13–23.
57. Brown JK, Husain IHMI. Status epilepticus. I: pathogenesis. Dev
Med Child Neurol. 1991;35:3–17.
58. Cole AJ. Status epilepticus and periictal imaging. Epilepsia.
2004;45(suppl 4):72–77.
59. Pohlmann-Eden B, Gass A, Peters CAN, et al. Evolution of MRI
changes and development of bilateral hippocampal sclerosis during

long lasting generalised status epilepticus. J Neurol Neurosurg
Psychiatry. 2004;75:898–900.
60. Maytal J, Shinnar S, Moshe SL, et al. Low morbidity and mortality
of status epilepticus in children. Pediatrics. 1989;83:323–331.
61. Dodrill CB, Wilensky AJ. Intellectual impairment as an outcome of
status epilepticus. Neurology. 1990;40:(suppl 2):23–27.
62. Adachi N, Kanemoto K, Muramatsu R, et al. Intellectual prognosis
of status epilepticus in adult patients: analysis with Wechsler Adult
Intelligence Scale-Revised. Epilepsia. 2005;46:1502–1509.
63. Shinnar S, Babb TL, Moshé SL, et al. Long-term sequelae of status
epilepticus. Chapter 65. In: Engel J, Pedley TA, eds. Epilepsy: A
Comprehensive Textbook. 2nd ed. Philadelphia: Lippincott-Raven
Publishers; 2008:751–759.
64. Hesdorffer D, Logroscino G, Cascino G, et al. Incidence of status
epilepticus in Rochester, Minnesota, 1965–1984. Neurology. 1998;
50:735–741.
65. Eriksson KJ, Koivikko MJ. Status epilepticus in children: aetiology,
treatment, and outcome. Dev Med Child Neurol. 1997;39:652–658.
66. Berg AT, Shinnar S. Do seizures beget seizures? An assessment of the
clinical evidence in humans. J Clin Neurophysiol. 1997;14:102–110.
67. Hauser WA, Rich SS, Annegers JF, et al. Seizure recurrence after a
1st unprovoked seizure: an extended follow-up. Neurology. 1990;
40:1163–1170.
68. DeLorenzo RJ, Towne AR, Pellock JM, et al. Status epilepticus in
children, adults and the elderly. Epilepsia. 1992;33(suppl 4):15–25.
69. Oxbury JM, Whitty CWM. Causes and consequences of status
epilepticus in adults. A study of 86 cases. Brain. 1971;94:733–744.
70. Hauser WA. Status epilepticus: epidemiologic considerations.
Neurology. 1990;40(suppl 2):9–13.
71. Waterhouse EJ, Vaughan JK, Barnes TY, et al. Synergistic effect of
status epilepticus and ischemic brain injury on mortality. Epilepsy
Res. 1998;29:175–183.
72. Rossetti AO, Hurwitz S, Logroscino G, et al. Prognosis of status
epilepticus: role of aetiology, age, and consciousness impairment
at presentation. J Neuro Neurosurg Psychiatry. 2006;77:611–615.
73. DeLorenzo RJ. Clinical and epidemiologic study of status epilepticus
in the elderly. Chapter 15. In: Rowan AJ, Ramsay RE, eds. Seizures and
Epilepsy in the Elderly. Boston: Butterworth-Heinemann; 1997:
191–205.
74. Logroscino G, Hesdorffer DC, Cascino GD, et al. Long-term mor-
tality after a first episode of status epilepticus. Neurology. 2002;58:
537–541.
75. Mayer SA, Claassen J, Lokin J, et al. Refractory status epilepticus.
Frequency, risk factors, and impact on outcome. Arch Neurol.
2002;59:205–210.
76. Scholtes FB, Renier WO, Meinardi H. Generalized convulsive sta-
tus epilepticus: causes, therapy, and outcome in 346 patients.
Epilepsia. 1994;35:1104–1112.
77. Towne AR, Pellock JM, Ko D, et al. Determinants of mortality in
status epilepticus. Epilepsia. 1994;35:27–34.
78. Drislane FW, Blum AS, Lopez MR, et al. Duration of refractory
status epilepticus and outcome: loss of prognostic utility after sev-
eral hours. Epilepsia. 2009;50:1566–1571.
79. Vignatelli L, Rinaldo R, Baldin E, et al. Impact of treatment on the
short-term prognosis of status epilepticus in two population-
based cohorts. J Neurol. 2008;255:197–204.
80. Cascino GD, Hesdorffer D, Logroscino G, et al. Treatment of non-
febrile status epilepticus in Rochester, Minn, from 1965 through
1984. Mayo Clin Proc. 2001;76:39–41.
81. Logroscino G, Hesdorffer DC, Cascino G, et al. Time trends in
incidence, mortality, and case-fatality after first episode of status
epilepticus. Epilepsia. 2001;42:1031–1035.
Chapter 28 ■ Convulsive Status Epilepticus 591
82. Shorvon S, Baulac M, Cross H, et al. For the TaskForce on Status
Epilepticus of the ILAE Commission for European Affairs.
Epilepsia. 2008;49:1277–1284.
83. Leppik IE, Derivan AT, Homan RW, et al. Double-blind study of
lorazepam and diazepam in status epilepticus. JAMA. 1983;249:
1452–1454.
84. Lowenstein DH, Alldredge BK. Status epilepticus. N Engl J Med.
1998;338:970–976.
85. Shaner DM, McCurdy SA, Herring MO, et al. Treatment of status
epilepticus: a prospective comparison of diazepam and phenytoin
versus phenobarbital and optional phenytoin. Neurology. 1988;38:
202–207.
86. Alldredge BK, Gelb AM, Isaacs SM, et al. A comparison of
lorazepam, diazepam, and placebo for the treatment of out-of-
hospital status epilepticus. N Engl J Med. 2001;345:631–637.
87. Knake S, Hamer HM, Rosenow F. Status epilepticus: a critical
review. Epilepsy Behav. 2009;15:1–14.
88. Bleck TP. Critical care of the patient in status epilepticus. In:
Wasterlain C, Treiman DT, eds. Status Epilepticus. Boston: MIT
Press; 2006:607–613.
89. Krishnamurthy KB, Drislane FW. Relapse and survival after bar-
biturate anesthetic treatment of refractory status epilepticus.
Epilepsia. 1996;37:863–867.
90. Claassen J, Hirsch LJ, Emerson RG, et al. Treatment of refractory
status epilepticus with pentobarbital, propofol, or midazolam: a
systematic review. Epilepsia. 2002;43:146–153.
91. Jagoda A, Riggio S. Refractory status epilepticus in adults. Ann
Emerg Med. 1993;22:1337–1348.
92. Prasad A, Worrall BB, Bertram EH, et al. Propofol and midazo-
lam in the treatment of refractory status epilepticus. Epilepsia.
2001;42:380–386.
93. Sahin M, Menache CC, Holmes GL, et al. Outcome of severe refrac-
tory status epilepticus in children. Epilepsia. 2001;42:1461–1467.
94. Lambrechtsen FACP, Buchhalter JR. Aborted and refractory
status epilepticus in children: a comparative analysis. Epilepsia.
2008;49:615–625.
95. Legriel S, Mourvillier B, Bele N, et al. Outcomes in 140 critically
ill patients with status epilepticus. Intensive Care Med. 2008;34:
476–480.
96. Sahin M, Menache CC, Holmes GL, et al. Prolonged treatment
for acute symptomatic refractory status epilepticus: outcome in
children. Neurology. 2003;61:398–401.
97. Holtkamp M, Othman J, Buchheim K, et al. A “malignant” vari-
ant of status epilepticus. Arch Neurol. 2005;62:1428–1431.
98. Kalita J, Nair PP, Misra UK. Status epilepticus and encephalitis. A
study of clinical findings, magnetic resonance imaging and
response to antiepileptic drugs. J Neurovirol. 2008;14:412–417.
99. Claassen J, Hirsch LJ, Emerson RG, et al. Continuous EEG mon-
itoring and midazolam infusion for refractory nonconvulsive sta-
tus epilepticus. Neurology. 2001;57:1036–1042.
100. Yaffe K, Lowenstein DH. Prognostic factors for pentobarbital
therapy for refractory generalized status epilepticus. Neurology.
1993;43:895–900.
101. Osorio I, Reed RC, Peltzer JN. Refractory idiopathic status
epilepticus: a probable paradoxical effect of phenytoin and car-
bamazepine. Epilepsia. 2000;41:887–894.
102. Krishnamurthy KB, Drislane FW. Phenobarbital and benzodi-
azepine assisted withdrawal of prolonged pentobarbital treatment
for refractory status epilepticus. J Epilepsy. 1997;10:211–214.
103. Mirski MA, Williams MA, Hanley DF. Prolonged pentobarbital
and phenobarbital coma for refractory generalized status epilep-
ticus. Crit Care Med. 1995;23:400–404.
592 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
104. Rossetti AO, Oddo M, Liaudet L, et al. Predictors of awakening
from postanoxic status epilepticus after therapeutic hypother-
mia. Neurology. 2009;72:744–749.
105. Cooper AD, Britton JW, Rabinstein AA. Functional and cognitive
outcome in prolonged refractory status epilepticus. Arch Neurol.
2009;66:1505–1509.
106. Claassen J, Hirsch LJ, Mayer SA. Treatment of status epilepticus:
a survey of neurologists. J Neurol Sci. 2003;211:37–41.
107. Holtkamp M. The anaesthetic and intensive care of status epilep-
ticus. Curr Opin Neurol. 2007;20:188–193.
108. Mirsattari SM, Sharpe MD, Young GB. Treatment of refractory
status epilepticus with inhalational anesthetic agents isoflurane
and desflurane. Arch Neurol. 2004;61:1254–1259.
109. Sheth RD, Gidal BE. Refractory status epilepticus: response to
ketamine. Neurology. 1998;51:1765–1766.
110. Ng Y, Kim HL, Wheless JW. Successful neurosurgical treatment of
childhood complex partial status epilepticus with focal resection.
Epilepsia. 2003;44:468–471.
111. Desbiens R, Berkovic SF, Dubeau F, et al. Life-threatening focal
status epilepticus due to occult cortical dysplasia. Arch Neurol.
1993;50:695–700.
112. Alexopoulos A, Lachhwani DK, Gupta A, et al. Resective surgery
to treat refractory status epilepticus in children with focal epilep-
togenesis. Neurology. 2005;64:567–570.
113. Lhatoo SD, Alexopoulos AV. The surgical treatment of status
epilepticus. Epilepsia. 2007;48(suppl 8):61–65.
114. D’Giano CH, Garcia MDC, Pomata H, et al. Treatment of refrac-
tory partial status epilepticus with multiple subpial transaction:
case report. Seizure. 2001;10:382–385.
115. Spencer SS, Schramm J, Wyler A, et al. Multiple subpial transac-
tion for intractable partial epilepsy: an international meta-analy-
sis. Epilepsia. 2002;43:141–145.
116. Krishnamurthy KB, Drislane FW. Depth of EEG suppression and
outcome in barbiturate anesthetic treatment for refractory status
epilepticus. Epilepsia. 1999;40:759–762.
117. Kaplan PW. Assessing the outcomes in patients with nonconvul-
sive status epilepticus: nonconvulsive status epilepticus is under-
diagnosed, potentially overtreated, and confounded by
comorbidity. J Clin Neurophysiol. 1999;16:341–352.
118. Rossetti AO, Logroscino G, Bromfield EB. Refractory status
epilepticus: effect of treatment aggressiveness on prognosis. Arch
Neurol. 2005;62:1698–1702.
119. Rashkin MC, Youngs C, Penovich P. Pentobarbital treatment of
refractory status epilepticus. Neurology. 1987;37:500–503.
120. Knake S, Rochon J, Fleischer S, et al. Status epilepticus after stroke
is associated with increased long-term case fatality. Epilepsia.
2006;47:2020–2026.
121. Jaitly R, Sgro JA, Towne AR, et al. Prognostic value of EEG mon-
itoring after status epilepticus: a prospective adult study. J Clin
Neurophysiol. 1997;14:326–334.
122. Husain AM, Mebust KA, Radtke RA. Generalized periodic
epileptiform discharges: etiologies, relationship to status epilep-
ticus, and prognosis. J Clin Neurophysiol. 1999;16:51–58.
123. Claassen J, Mayer SA, Kowalski RG, et al. Detection of electro-
graphic seizures with continuous EEG monitoring in critically ill
patients. Neurology. 2004;62:1743–1748.
124. Jette N, Claassen J, Emerson RG, et al. Frequency and predictors
of nonconvulsive seizures during continuous electroencephalo-
graphic monitoring in critically ill children. Arch Neurol. 2006;
63:1750–1755.
125. Hirsch LJ. Continuous EEG monitoring in the intensive care
unit: an overview. J Clin Neurophysiol. 2004;21:332–340.
126. Ferrie CD, Caraballo R, Covanis A, et al. Autonomic status epilep-
ticus in Panayiotopoulos syndrome and other childhood and
adult epilepsies: a consensus view. Epilepsia. 2007;48:1165–1172.
127. Wieser HG, Chauvel P. Simple partial status epilepticus and
epilepsia partialis continua of Kozhevnikov. Chapter 61. In: Engel
J, Pedley TA, eds. Epilepsy: A Comprehensive Textbook. 2nd ed.
Philadelphia: Lippincott-Raven Publishers; 2008:705–723.
128. Schomer DL. Focal status epilepticus and epilepsia partialis
continua in adults and children. Epilepsia. 1993;34(suppl 1):
S29–S36.
129. Fejerman N, DiBlasi AM. Status epilepticus of benign partial
epilepsies in children: Report of two cases. Epilepsia. 1987;28:
351–355.
130. Rumbach L, Sablot D, Berger E, et al. Status epilepticus and
stroke. Report on a hospital-based stroke cohort. Neurology. 2000;
54:350–354.
131. Scholtes FB, Renier WO, Meinardi H. Simple partial status
epilepticus: causes, treatment, and outcome in 47 patients. J
Neurol Neurosurg Psychiatry. 1996;61:90–92.
132. Devinsky O, Kelley K, Porter RJ, et al. Clinical and electroen-
cephalographic features of simple partial seizures. Neurology. 1988;
38:1347–1352.
133. Grand’Maison F, Reiher J, Leduc CP. Retrospective inventory of
EEG abnormalities in partial status epilepticus. Electroencephalogr
Clin Neurophysiol. 1991;79:264–270.
134. Juul-Jensen P, Denny-Brown D. Epilepsia partialis continua. A
clinical, electroencephalographic, and neuropathological study
of nine cases. Arch Neurol. 1966;15:563–578.
135. Thomas JE, Reagan TJ, Klass DW. Epilepsia partialis continua. A
review of 32 cases. Arch Neurol. 1977;34:266–275.
136. Cockerell OC, Rothwell J, Thompson PD, et al. Clinical and phys-
iological features of epilepsia partialis continua. Cases ascer-
tained in the UK. Brain. 1996;119(Pt 2):393–407.
137. Kozhevnikov A. Eine besondere Form von corticaler Epilepsie.
Neurol Centralbl. 1895;14:47–48.
138. Botez MI, Brossard L. Epilepsia partialis continua with well-
delimited subcortical frontal tumor. Epilepsia. 1974;15:39–43.
139. Feddersen B, Bender A, Arnold S, et al. Aggressive confusional
state as a clinical manifestation of status epilepticus in MELAS.
Neurology. 2003;61:1149–1150.
140. Shavit YB, Graus F, Probst A, et al. Epilepsia partialis continua: a
new manifestation of anti-Hu-associated paraneoplastic
encephalomyelitis. Ann Neurol. 1999;45:255–258.
141. Ruggieri PM, Najm I, Bronen R, et al. Neuroimaging of the cor-
tical dysplasias. Neurology. 2004;62(suppl 3):S27–S29.
142. Singh BM, Strobos RJ. Epilepsia partialis continua associated
with nonketotic hyperglycemia: clinical and biochemical profile
of 21 patients. Ann Neurol. 1980;8:155–160.
143. Kuroiwa Y, Tohgi H, Takahashi A, et al. Epilepsia partialis con-
tinua: active cortical spike discharges and high blood flow in the
motor cortex and enhanced transcortical long loop reflex. J
Neurol. 1985;232:162–166.
144. Wieser HG, Graf HP, Bernoulli C, et al. Quantitative analysis of
intracerebral recordings in epilepsia partialis continua.
Electroencephalogr Clin Neurophysiol. 1978;44:14–22.
145. Watanabe K, KuroiwaY, Toyokura A. Epilepsia partialis continua.
Epileptogenic focus in motor cortex and its participation in
transcortical reflexes. Arch Neurol. 1984;41:1040–1044.
146. Cowan JMA, Rothwell JC, Wise RJS, et al. Electrophysiological
and positron emission studies in a patient with cortical
myoclonus, epilepsia partialis continua and motor epilepsy. J
Neurol Neurosurg Psychiatr. 1986;49:796–807.

147. Kristiansen K, Kaada BR, Henriksen GF. Epilepsia partialis con-
tinua. Epilepsia. 1971;12:263–267.
148. Bancaud J, Bonis A, Trottier S, et al. L’épilepsie partielle continue:
syndrome et maladie. Rev Neurol (Paris). 1982;138:803–814.
149. Morrell F, Whisler WW, Bleck TP. Multiple subpial transection: a
new approach to the surgical treatment of focal epilepsy. J
Neurosurg. 1989;32:553–559.
150. Morrell F, Whisler WW, Smith MC. Multiple subpial transaction
in Rasmussen’s encephalitis. Chapter 20. In: Andermann F, ed.
Chronic Encephalitis and Epilepsy. Rasmusen’s Syndrome. Boston:
Butterworth-Heinemann; 1991:219–233.
151. Chatrian GE, Shaw CM, Leffman H. The significance of periodic
lateralized epileptiform discharges in EEG: An electrographic,
clinical and pathological study. Electroenceph clin Neurophysiol.
1964;17:177–193.
152. Rasmussen T, Olszewski J, Lloyd-Smith DL. Focal seizures due to
chronic localized encephalitis. Neurology. 1958;8:435–445.
153. Oguni H, Andermann F, Rasmussen TB. The natural history of
the syndrome of chronic encephalitis and epilepsy. A study of the
MNI series of forty-eight cases. In: Andermann F, ed., Chronic
Encephalitis and Epilepsy. Boston: Butterworth-Heinemann;
1991:7–35.
154. Aguilar MJ, Rasmussen T. Role of encephalitis in pathogenesis of
epilepsy. Arch Neurol. 1960;2:663–676.
155. Rasmussen T. Further observations on the syndrome of chronic
encephalitis and epilepsy. Appl Neurophysiol. 1978;41:1–12.
156. So NK, Gloor P. Electroencephalographic and electrocortico-
graphic findings in chronic encephalitis of the Rasmussen type.
In: Andermann F, ed. Chronic Encephalitis and Epilepsy. Boston:
Butterworth-Heinemann; 1991:37–46.
157. Rogers SW, Andrews PI, Gahring LC, et al. Autoantibodies to
glutamate receptor GluR3 in Rasmussen’s encephalitis. Science.
1996;46:242–246.
158. Hart YM, Cortez M, Andermann F, et al. Medical treatment of
Rasmussen’s syndrome (chronic encephalitis and epilepsy): effect
of high-dose steroids or immunoglobulins in 19 patients.
Neurology. 1994;44:1030–1036.
159. Villemure JG, Vernnet O, Delalande O. Hemispheric disconnec-
tion: callosotomy and hemispherectomy. Adv Tech Stand
Neurosurg. 2000;26:25–78.
160. Westmoreland BF, Klass DW, Sharbrough FW. Chronic periodic
lateralized epileptiform discharges. Arch Neurol. 1986;43:
494–496.
161. Snodgrass SM, Tsuburaya K, Ajmone-Marsan C. Clinical signifi-
cance of periodic lateralized epileptiform discharges: Relationship
with status epilepticus. J Clin Neurophysiol. 1989;6:159–172.
162. Baykan B, Kinay D, Gokyigit A, et al. Periodic lateralized epilepti-
form discharges: association with seizures. Seizure. 2000;9:402–406.
163. Schraeder, PL, Singh N. Seizure disorders following periodic lat-
eralized epileptiform discharges. Epilepsia. 1980;21:647–653.
164. Schwartz MS, Prior PF, Scott DR. The occurrence and evolution
in the EEG of a lateralized periodic phenomenon. Brain. 1973;
96:613–622.
165. Chong DJ, Hirsch LJ. Which EEG patterns warrant treatment in
the critically ill? Reviewing the evidence for treatment of periodic
epileptiform discharges and related patterns. J Clin Neurophysiol.
2005;22:79–91.
166. Jirsch J, Hirsch LJ. Nonconvulsive seizures: developing a rational
approach to the diagnosis and management in the critically ill
population. Clin Neurophysiol. 2007;118:1660–1670.
167. Reiher J, Rivest J, Grand'Maison F, et al. Periodic lateralized
epileptiform discharges with transitional rhythmic discharges:
Chapter 28 ■ Convulsive Status Epilepticus 593
association with seizures. Electroencephalogr clin Neurophysiol.
1991;78:12–17.
168. Ali II, Pirzada NA, Vaughn BV. Periodic lateralized epileptiform
discharges after complex partial status epilepticus associated with
increased focal cerebral blood flow. J Clin Neurophysiol.
2001;18:565–569.
169. Assal F, Papazyan JP, Slosman DO, et al. SPECT in periodic later-
alized epileptiform discharges (PLEDs): a form of partial status
epilepticus? Seizure. 2001;10:260–265.
170. Handforth A, Cheng JT, Mandelkern MA, et al. Markedly
increased mesiotemporal lobe metabolism in a case with PLEDs:
further evidence that PLEDs are a manifestation of partial status
epilepticus. Epilepsia. 1994;35:876–881.
171. Terzano MG, Parrino L, Mazzucchi A, et al. Confusional states
with periodic lateralized epileptiform discharges (PLEDs): a
peculiar epileptic syndrome in the elderly. Epilepsia. 1986;27:
446–457.
172. Garzon E, Fernandes RMF, Sakamoto AC. Serial EEG during
human status epilepticus. Evidence for PLED as an ictal pattern.
Neurology. 2001;57:1175–1183.
173. de la Paz D, Brenner RP. Bilateral independent periodic lateral-
ized epileptiform discharges. Clinical significance. Arch Neurol.
1981;38:713–715.
174. Celesia GG, Grigg MM, Ross E. Generalized status myoclonicus in
acute anoxic and toxic-metabolic encephalopathies. Arch Neurol.
1988;45:781–784.
175. Jumao-as A, Brenner RP. Myoclonic status epilepticus: a clinical and
electroencephalographic study. Neurology. 1990;40:1199–1202.
176. Hallett M. Myoclonus: relation to epilepsy. Epilepsia. 1985;
26(suppl 1):67–77.
177. Wakamoto H, Nagao H, Manabe K, et al. Nonconvulsive status
epilepticus in eyelid myoclonia with absences—evidence of
provocation unrelated to photosensitivity. Neuropediatrics.
1999;30:149–150.
178. Ohtahara S, Ohtsuka Y. Myoclonic status epilepticus. Chapter 63.
In: Engel J Jr, Pedley TA, eds. Epilepsy: A Comprehensive Textbook.
2nd ed. Philadelphia: Lippincott-Raven Publishers; 2008:725–729.
179. Asconape J, Penry JK. Some clinical and EEG aspects of benign
juvenile myoclonic epilepsy. Epilepsia. 1984;25:108–114.
180. Kimura S, Kobayashi T. Two patients with juvenile myoclonic
epilepsy and nonconvulsive status epilepticus. Epilepsia. 1996;
37:275–279.
181. Adachi M, Inoue T, Tsuneishi S, et al. Eyelid myoclonia with
absences in monozygotic twins. Pediatr Int. 2005;47:343–347.
182. Wheless JW. Acute management of seizures in the syndromes
of idiopathic generalized epilepsies. Epilepsia. 2003;44(suppl 2):
22–26.
183. Genton P, Gelisse P, Thomas P, et al. Do carbamazepine and
phenytoin aggravate juvenile myoclonic epilepsy? Neurology.
2000;55:1106–1109.
184. Berkovic SF, Andermann F, Carpenter S, et al. Progressive
myoclonus epilepsies: specific causes and diagnosis. N Engl J
Med. 1986;315:296–305.
185. Dravet C, Bureau M, Oguni H, et al. Severe myoclonic epilepsy in
infancy: Dravet syndrome. Adv Neurol. 2005;95:71–102.
186. Korff CM, Laux L, Kelley Kr, et al. Dravet syndrome (severe
myoclonic epilepsy in infancy): a retrospective study of 16 patients.
J Child Neurol. 2007;22:185–194.
187. Simon RP, Aminoff MJ. Electrographic status epilepticus in fatal
anoxic coma. Ann Neurol. 1986;20:351–355.
188. Wijdicks EF, Young GB. Myoclonus status in comatose patients
after cardiac arrest. Lancet. 1994;343:1642–1643.
594 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
189. Krumholz A, Stern BJ, Weiss HD. Outcome from coma after car-
diopulmonary resuscitation. Neurology. 1988;38:401–405.
190. Young GB, Gilbert JJ, Zochodne DW. The significance of
myoclonic status epilepticus in postanoxic coma. Neurology.
1990;40:1843–1848.
191. Brown P, Thompson PD, Rothwell JC, et al. A case of postanoxic
encephalopathy with cortical action and brainstem reticular
reflex myoclonus. Mov Disord. 1991;6:139–144.
192. Sheth RD, Gidal BE. Intravenous valproic acid for myoclonic sta-
tus epilepticus. Neurology. 2000;54:1201.
193. Genton P. When antiepileptic drugs aggravate epilepsy. Brain
Dev. 2000;22:75–80.
194. Perucca E, Gram L, Avanzini G, et al. Antiepileptic drugs as a
cause of worsening seizures. Epilepsia. 1998;39:5–17.
195. Lance JW, Adams RD. The syndrome of intention or action
myoclonus as a sequel to hypoxic encephalopathy. Brain. 1963;
86:111–136.
196. Arnoldus EPJ, Lammers GJ. Postanoxic coma: good recovery
despite myoclonus status. Ann Neurol. 1995;38:697.
197. Morris HR, Howard RS, Brown P. Early myoclonic status and out-
come after cardiorespiratory arrest. J Neurol Neurosurg Psychiatry.
1998;64:267–268.
198. Gastaut H, Roger J, Ouahchi S, et al. An electroclinical study of
generalized epileptic seizures of tonic expression. Epilepsia. 1963;
4:15–44.
199. Kobayashi E, Thomas P, Andermann F. Tonic status epilepticus in
patients with idiopathic generalized epilepsy. Epileptic Disord.
2005;7:327–331.
200. Somerville ER, Bruni J. Tonic status epilepticus presenting as a
confusional state. Ann Neurol. 1983;13:549–551.
201. Tassinari CA, Dravet C, Roger J, et al. Tonic status epilepticus pre-
cipitated by intravenous benzodiazepine in five patients with
Lennox-Gastaut syndrome. Epilepsia. 1972;13:421–435.

Nonconvulsive Status Epilepticus
FRANK W. DRISLANE, PETER W. KAPLAN, AND SUSAN T. HERMAN
HISTORY
Although generalized convulsive status epilepticus has been
recognized for millennia, nonconvulsive status epilepticus
(NCSE) is a much more recent concept. Perhaps by association
with other forms of epilepsy with altered consciousness, some
speculation about the possible existence of NCSE emerged by
1800 or so. Trousseau, at Hotel Dieu in Paris, noted that petit
mal seizures might recur with sufficient frequency “that one
seizure would become confused with the next, simulating a
continuous seizure which might persist for two or three days
millennia” thus anticipating a diagnosis of absence status
epilepticus (1–3). Nonconvulsive spells also attracted the atten-
tion of Jules Falret (1824 to 1902) who said that a patient with
“petit mal intellectuel … might leave home or work, with
clouded mind … [and] … complete lapses of memory (2,4).”
Samuel Wilks in London in the late 1800s described a patient
who was: “… in the condition which is popularly called ‘lost;’
he is scarcely conscious of acts and conversation going on
around him, yet he may continue walking in a given direction,
showing that his movements must still, in a measure, be guided
by his senses … in much the same state as a somnambulist. This
condition … is called the status epilepticus, although the term
is more usually applied to [a] patient … who, after a succession
of fits, lay for hours in a state of lethargy. In the milder forms it
is one of great interest from a physiology point-of-view and
seems to point to the possibility of a subconscious state, in
which the brain is sufficiently active to control the spinal system
and yet not awake [for] consciousness (1,5).”
Also in the late 1800s, Charcot and others described cases
suggestive of NCSE. Charcot postulated that a state of somnam-
bulism derived from ongoing seizures. He presented a patient
with a nonconvulsive form of long-lasting seizures, several
times in his “Leçons du Mardi” at the Saltpêtrière—a delivery
man who wandered about Paris and even to the coast, at Brest,
being arrested on one occasion and released only on the cog-
nizance of Charcot—who treated him with bromides (6). In
Britain, Gowers speculated that similar states rather than being
ictal, occurred after the seizure: “After epileptic fits of moderate
severity, the patient may pass into a condition of mental
automatism, in which various acts are performed in an appar-
ently conscious manner, but of which no recollection is after-
ward retained (7).”
Charcot’s observations fostered speculation about the exis-
tence of NCSE, but without an appreciation of the role of electric-
ity in the nervous system [starting to emerge in the 17th century
(8)], it would be hard to conceive a modern understanding of
CHAPTER
29
NCSE. Clinical descriptions of confusional states (even with
signs of epilepsy such as twitching of facial or limb muscles)
were insufficient to prove that NCSE was a proper diagnosis.
Encephalopathy, psychogenic unresponsiveness, or postictal
behavior might appear very similar.
Berger’s 1929 invention and development of the electroen-
cephalogram (EEG) made it possible to record the previously
suspected electrical activity of the brain (9). He also began
studies correlating EEG activity, as a measure of brain electrical
function, with psychologic and behavioral abnormalities, and
eventually, epileptic seizures (10). Thus, earlier historical
descriptions of what might have been NCSE were thrust into a
provable domain. Just over a decade later, clinical-EEG correla-
tions showed beyond doubt that NCSE was part of epilepsy and
not, as some had suspected, from hysterical or nonepileptic
fugue states. The stage was set for modern electroclinical corre-
lation, including video-EEG monitoring, various forms of
imaging, and refinement of treatment.
Lennox detailed one of the first cases of “petit mal status”
with an electrographic correlate in 1945 (11). In 1954, Penfield
and Jasper described simple partial status epilepticus (SPSE) or
aura continua in the form of recurrent sensory phenomena,
commenting that they were “at least as common as continuing
circumscribed movements (12).” In the first thorough descrip-
tion of complex partial status epilepticus (CPSE) in 1956,
Gastaut and Roger described a nurse whose seizures may have
lasted for months (13). In the past half century, there have been
numerous clinical series detailing the electroclinical correlates
and nature of NCSE.
DEFINITION
It is difficult to state a definition of NCSE, primarily because
the clinical manifestations are protean and because the EEG
findings, even while crucial for diagnosis, are often ambiguous
or controversial. The significance of epileptiform waveforms
and other abnormalities on the EEG is often unclear, and even
the role of clearly epileptiform EEG abnormalities in causing
clinical deficits can be uncertain or impossible to determine.
Because NCSE is considered an epileptic process without con-
vulsions (or no more than minimal motor manifestations),
almost all definitions insist on EEG evidence of continuing or
very frequent epileptiform activity.
The definition of generalized convulsive status epilepticus
(GCSE) can be complicated and controversial (Chapter 28), but
that for NCSE is much more so. It has been defined as a “range
of conditions in which electrographic seizure activity is prolonged
595
596 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
and results in nonconvulsive clinical symptoms. [It is … ] prima-
rily as a form of epileptic cerebral response which is dependant
largely on the level of cerebral development and integrity, the pres-
ence or absence of encephalopathy, the type of epilepsy syndrome,
and the anatomical location of the epileptic activity (14).” This
definition reflects the pleomorphic character of NCSE and
alludes to important insights into its biology, but diagnosis of
individual cases can remain challenging. While there have been
several proposals to change a temporal definition of GCSE from
30 to 5 or 10 minutes (15), it is probably most reasonable to use
a 30-minute definition for NCSE (16).
CLINICAL PRESENTATION
NCSE has remarkably varied and often very subtle presenta-
tions (17,18). At times, it follows generalized seizures and
GCSE, that is, GCSE continues in a more subtle form; the
epileptiform discharges of a seizure continue on the EEG after
the motor manifestations (convulsions) have ceased. In this
case, it is best to consider NCSE as a later stage of GCSE, in
terms of its pathophysiology, clinical implications, and man-
date for treatment. Any patient who has not recovered from a
seizure within 20 to 30 minutes or so should be considered as
possibly in NCSE.
NCSE can also occur without clinically evident seizures, that
is, purely nonconvulsive events. Even on its own (not following
GCSE), NCSE has pleomorphic presentations. Its behavioral
correlates in different types of NCSE arise from—or are at least
associated with—different areas of maximal involvement of
seizure activity in the brain.
NCSE typically encompasses an ictal impairment of mental
status, cognition, or behavior compared with a baseline state;
automatisms; or change in sensory perception (e.g., auditory,
visual, somatosensory, or psychic) (19). Anterograde and retro-
grade memory may be affected; affects can be altered (agitation,
sadness, weeping); speech and language may be impaired (with
mutism or paraphasic errors); and confusional states may
occur. There can be psychomotor retardation or behavioral
changes, beyond the patient’s baseline. Sometimes, there are
admixed, but generally minimal, motor manifestations, such as
subtle facial or limb jerking, blinking, eye deviation with nys-
tagmus (20), or perioral and eyelid myoclonus—highly sugges-
tive of NCSE in the appropriate clinical setting (21).
Most epidemiologic studies have been focused on the inci-
dence of GCSE. All forms of NCSE may constitute one-fourth
of all cases of SE (22), and thus have an incidence of about
10/100,000 persons per year in the general population (16).
About half of all cases in a German study of SE were noncon-
vulsive, implying about 8.5 incident cases per 100,000 popula-
tion per year (23). Given the difficulties in diagnosis, these
estimates of the incidence of NCSE may be far too low.
DIFFERENTIAL DIAGNOSIS
NCSE often occurs in patients with underlying chronic medical
and neurologic illnesses, and also with acute and subacute med-
ical problems such as infections, vascular disease, or metabolic
abnormalities. The other illnesses may obscure the diagnosis.
Many patients with NCSE have had their compromised cogni-
tion attributed to problems such as electrolyte imbalance,
hyperglycemia, pneumonia, or earlier seizures (24). Many
patients have dementia, mental retardation, or psychiatric dis-
orders, making it difficult for unfamiliar observers, including
physicians, to notice a new perturbation in cognition or behav-
ior (19,24–28). It is often difficult to determine if the patient is
“more confused than usual,” “confused beyond what might be
attributable to an infection or metabolic disturbance,” or “inap-
propriately lethargic after a convulsion.”
Altered alertness or behavior may be ascribed to an
encephalopathy or a postictal state, delaying diagnosis (19,24).
Patients with chronic psychiatric illness are at risk of delayed
diagnosis of NCSE because of their neuroleptic burden and
propensity for starting and stopping benzodiazepines
(BDZs)—triggers for de novo NCSE (29). Especially in the eld-
erly, NCSE may present with confusion, suggesting medication
side effects or dementia (30,31). Many sick, hospitalized
patients have altered mental status due to medications, infec-
tions, and metabolic encephalopathies from their underlying
illnesses, and the same illness may cause or precipitate NCSE
(32,33). Diagnostic clues of NCSE include an earlier history of
epilepsy or risk factors for epilepsy.
NCSE is typically discovered in the emergency department,
intensive care units (ICUs), and on neurology and psychiatry
services, although it can occur during other hospital services
and even in ambulatory patients. In the emergency department,
the diagnosis of NCSE may be missed when there is lethargy or
confusion attributed to a postictal state; ictal confusion mis-
taken for a metabolic encephalopathy; unresponsiveness and
catalepsy presumed to be psychogenic; obtundation thought to
be due to alcohol or drug intoxication; hallucinations and agi-
tation mistaken for psychosis or delirium; lethargy presumed
due to hyperglycemia; mutism attributed to aphasia; and laugh-
ing and crying ascribed to emotional lability (24). In hospital-
ized or institutionalized patients, NCSE can present primarily
as an impaired level of consciousness or responsiveness.
Drugs precipitating some cases of NCSE include neuroleptics,
cephalosporins, radiologic contrast agents (34), and gamma
amino butyric acid (GABA) agonists, including baclofen (35),
tiagabine, and vigabatrin (36,37). Occasionally, refractory NCSE
can be caused by mitochondrial disorders, including mitochon-
drial encephalomyopathy with lactic acidosis and stroke-like
episodes (MELAS) (38). Unusual forms of NCSE may result from
paraneoplastic syndromes, especially in those likely caused by cell
surface antibodies (39). In the anti-NMDA-receptor-associated
encephalitis, patients are particularly likely to be young women
with ovarian teratomas, presenting with psychiatric or memory
disorders, transferred to ICUs with stupor, refractory NCSE, and
bizarre orofacial dyskinesias; many respond well to immunosup-
pressive treatment (and tumor removal) but may take many
months to return to normal function (40).
Mimics of NCSE include migraine, transient global amnesia,
or sleep disorders, including cataplexy. Others include psychi-
atric disorders, metabolic derangements, paroxysmal cardiovas-
cular or autonomic dysfunction, endocrine dysfunction, and

limbic encephalitis—whether paraneoplastic or related to volt-
age-gated potassium channel antibodies (39), and confusional
states induced by many drugs, such as lithium, other psy-
chotropic medications, or alcohol (24,41).
OVERVIEW OF DIAGNOSIS
While generalized convulsions are usually apparent clinically,
nonconvulsive seizures (probably the majority in adults) are
harder to recognize. Paradoxically, NCSE may be even harder to
recognize than individual seizures because there may be no sud-
den behavioral change at onset; the patient’s condition may fluc-
tuate little over the day; the significance of many EEG patterns is
controversial; and the response to antiepileptic drugs (AEDs)
may be subtle or delayed.
The diagnosis of NCSE derives from two major elements: an
alteration in baseline cognition, behavior, or other neurologic
function, and concurrent epileptiform seizure patterns on the
EEG (42,43). Satisfying these criteria is often complicated and
difficult. To recognize the clinical features, one must maintain a
high index of suspicion. The sign of NCSE is not merely the
abnormal behavior, but its change from a baseline state.
To diagnose NCSE, epileptologists have required impaired
consciousness for 30 to 60 minutes and an EEG with continuous
(44), or at least some form of, seizure activity (42). EEG wave-
form morphologies may include rhythmic slowing, sharp waves,
spikes, and mixtures of these features (45). Discharges may be
continuous, persistent with brief pauses of a few seconds, or
intermittent. Many EEGs are more ambiguous (see section “EEG
Diagnosis of Nonconvulsive Status Epilepticus”), with more
blunted waveforms slower than 1.5 Hz, sometimes resembling
triphasic waves (TWs) (43,46). Over time, NCSE patterns tend to
evolve in morphology, amplitude and frequency, and wax or
wane. EEG ictal discharges should be continuous or recurrent for
more than 30 minutes, without the patient’s return to a normal
clinical state or resumption of the preictal EEG pattern between
seizures. Specific EEG features of individual types of NCSE are
detailed in the sections below describing those types of NCSE.
Although the EEG is the most reliable diagnostic test for
NCSE, an AED challenge may also be useful. Some patients
with less “classic” EEG findings show a clinical and electro-
graphic response to BDZs or other AEDs (42). In the appropri-
ate clinical setting, a rapid response to AEDs can be persuasive
for a diagnosis of NCSE. Lorazepam may be given in small
sequential doses, while monitoring blood pressure, respiratory
rate, and oxygenation. To be diagnostic of ongoing seizure
activity, there should be prompt improvement in both the
patient’s clinical state and the EEG, or complete cessation of
electrographic ictal activity with return of normal EEG back-
ground activity (43,46) (see Figs. 29.6, 29.14, 29.15, 29.19, and
29.25). Intravenous BDZs may abolish electrographic seizures,
but they can also suppress nonepileptic EEG patterns, such as
TWs (47), so improvement in the EEG alone upon AED admin-
istration does not prove that a particular EEG pattern was a
seizure. A controversial EEG pattern is confirmed as ictal only
when marked clinical improvement accompanies the EEG
improvement—but often such an improvement does not occur.
Chapter 29 ■ Nonconvulsive Status Epilepticus 597
Frequently, the response to AEDs is inconclusive, and the diag-
nosis must be made on clinical and EEG grounds alone.
Importantly, lack of a prompt improvement with AEDs does
not refute a diagnosis of NCSE. Even in patients with definite
NCSE, a rapid clinical response is uncommon, especially in
obtunded or comatose patients (48). In many, the response is
equivocal or substantially delayed (24,49), and sedating BDZs
may also impair a clinical response. Following a clinical seizure,
if the patient improves on EEG, but not clinically, continuous
EEG (C-EEG) monitoring should be considered to look for evi-
dence of subsequent recurrent seizures.
There have also been a few cases of focal NCSE diagnosed
clearly by neuroimaging, for example, with 18F-fluo-
rodeoxyglucose (FDG)-positron emission tomography (PET)
scans (50,51). Some patients in these reports had clear areas of
persistent focal hypermetabolism during clinical episodes of
definite aphasia or apparent epileptic amnesia (with an unre-
markable EEG, or focal slowing) and an appropriate and quick
response to intravenous BDZs. Occasionally, hyperperfusion on
ictal single-photon emission computed tomography (SPECT)
may help to make, or confirm, the diagnosis (52).
PATHOPHYSIOLOGY
The pathophysiology of convulsive status epilepticus is dis-
cussed in Chapter 28. That of NCSE is substantially less well
understood, both in the generation of neural activity that pro-
duces and sustains NCSE, and in the possibility of its inducing
subsequent neuronal damage and longer-lasting epilepsy on its
own.
Most NCSE, at least in adults, is either CPSE or secondary
generalized NCSE—both with underlying focal onset, whether
evident clinically or not. This NCSE usually begins at the same
focus of epileptogenesis as for focal seizures that do not become
prolonged to the point of SE. How do these focal seizures reach
a self-sustaining state?
In a rodent model of limbic SE produced by repetitive, rapid
direct electrical stimulation of the hippocampus for 30 to 90
minutes, seizures progress from self-limited to much more pro-
longed, and eventually self-sustaining SE, persisting for 12 to 24
hours after the stimulation has ceased (53,54). The induced
seizures appear similar to those of humans with CPSE—
although it is still unclear whether an analogous process occurs
in humans. Seizures beginning in the hippocampus appear to
promote the propagation of yet more seizure activity, which is
maintained in a circuit of structures including the entorhinal
cortex, dentate gyrus, and parts of the hippocampus and
subiculum (55,56). The initiation of SE in these models was
similar to that for individual seizures, but inhibitory or termi-
nating mechanisms failed (see Chapter 28). Generalized
seizures, or seizures with onset in other areas, can also precipi-
tate focal temporolimbic seizures and focal SE by way of sec-
ondary involvement of the hippocampus or other mesial
temporal structures (57).
The activity of certain neurotransmitters can facilitate or
inhibit the activation of complex circuits of limbic neuronal
activity, prolonging or interrupting nonconvulsive seizures in a
598 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
(somewhat similar) model of self-sustaining SE produced by
stimulation of the perforant pathway in rats (58). For example,
injection of galanin (a peptide that inhibits glutamate release
presynaptically) into the hippocampus attenuates seizure activ-
ity (59). Galanin production can decline during SE, while sub-
stance P (which may enhance SE by promoting glutamate
release) may increase (60), facilitating the prolongation of
seizures. As with GCSE, the balance of excitatory and inhibitory
function seems to shift toward excitation as NCSE progresses.
The pathophysiology of absence SE and other forms of pri-
marily generalized SE, however, is starkly different from that of
focal-onset NCSE (61). These generalized forms of NCSE
involve excessively prolonged synchronization in widespread
thalamocortical circuits as part of the generation, maintenance,
and reinforcement of aberrant excitatory electrical rhythms in
a neuronal system involving both cortical and subcortical struc-
tures (55,62). These abnormal electrical rhythms are mani-
fested by rhythmic generalized, anteriorly predominant, spike
and polyspike and slow-wave discharges on the EEG.
Experimentally, these 3-Hz generalized cortical spike-and
slow-wave discharges can be produced by electrical stimulation
of midline intralaminar nuclei of the thalamus (63), and the
same thalamic discharges occur during absence seizures in
humans (64). The discharges are dependent on low-threshold
T-type calcium channel activity (55). Thalamic injections of
GABA-B agonists increase the spike-and slow-wave discharges,
and this activity can be suppressed by specific AEDs (55).
Little experimental evidence is available on long-term conse-
quences of NCSE. Models of repetitive limbic seizures and SE
have been produced by local tissue damage caused by injections
of kainic acid, but chemical and electrical methods of inducing
SE may damage neurologic tissue independent of the subse-
quent seizures, that is, it may be the local electrical or chemical
precipitant or toxin that damages neurons.
There are a few clinical reports of patients with hippocam-
pal edema during episodes of NCSE, with subsequent hip-
pocampal atrophy in the same area (65,66), but not all studies
show this progression to atrophy, possibly depending in part on
how prolonged the episodes of NCSE were. These results also
contrast with the very minimal clinical morbidity attributable
to NCSE directly. Lothman has summarized the many physio-
logic changes that occur during SE, but most apply primarily to
GCSE; some may apply to NCSE (67).
It is very difficult to prove that any neurologic damage
accrues from NCSE. Pathologic studies of the effects of GCSE in
humans are relatively few, in part because fatal SE cases are often
associated with acute, severe brain-injuring illnesses that may
cause damage independently. Episodes of NCSE are even less
often fatal unless they occur in association with GCSE or other
acute, severe neurologic illness—in which case it is difficult to
sort out the cause of any neuronal damage. For the most part,
pathologic studies of patients with pure forms of NCSE remain
unavailable.
The experimental models described in the chapter on GCSE
included SE durations of well over 30 minutes (typically 1 to 2
hours of convulsions and electrical discharging activity) to
produce neuronal damage in animals. In one rodent model of
SE, neurologic damage in the amygdala and pyriform cortex
correlated strongly with prolonged “high-frequency” (10 Hz)
discharges, but there was no damage following discharges
slower than 1 Hz (68). Damage was directly related to the dura-
tion and intensity of electrographic seizure activity. It is not
clear that such prolonged intense stimulation, or the resultant
10-Hz spike discharges, are an accurate model of human NCSE.
Also, in the NCSE model with continuous hippocampal stimu-
lation, neuronal loss did not occur when the interval between
nonconvulsive seizures was increased (69). Currently, little is
known about the pathologic effects of NCSE in humans.
Clinical consequences of NCSE are discussed in later sections.
CLASSIFICATION
No classification system for NCSE will be satisfactory to all
basic investigators, electroencephalographers (EEG-ers), clini-
cians, and scholars. Gastaut stated that there were “as many
types of status as there are types of epileptic seizures” (70), but
the correspondence is not strict. For example, different focal
seizures emanate from different brain areas, with different signs
and symptoms, but many can go on to generalized convulsions
and GCSE, and to subsequent NCSE after convulsions—with
all later stages appearing similar despite the different origins.
Others reject a classification of NCSE according to seizure
types, pointing out that epilepsy syndromes such as juvenile
myoclonic epilepsy (JME) consist of more than seizures types
alone, and also that NCSE is not always simply the prolongation
of individual seizures (71). Different NCSE syndromes include
EEG, clinical, and genetic bases; developmental aspects; age-
specific presentations; varied concomitant structural brain abnor-
malities or encephalopathies; quite varied responses to treatment;
and different longer-term outcome. Thus, NCSE can also be clas-
sified based primarily on age and the state of cerebral develop-
ment or maturation, presence or absence of an encephalopathy,
occurrence within an epilepsy syndrome, and, for NCSE of focal
origin, understanding by anatomic localization (Table 29.1) (14).
This scheme, organized primarily by age of onset, aids in under-
standing NCSE in the context of the brain’s development and
prior condition. Many of the syndromes included occur in
neonates, infants, and young children; some continue into adoles-
cence in individuals, or occur in adults.
Boundary syndromes are mostly those with markedly abnor-
mal baseline neurologic conditions and epileptiform abnormali-
ties on EEG—in which it is difficult or impossible to determine
what role the epileptiform discharges have in causation of the
clinical manifestations. Most are pediatric NCSE syndromes
that occur within a broad range of illnesses such as the
Lennox–Gastaut syndrome (LGS), and include electrical status
epilepticus in sleep (ESES), described briefly here but more in
Chapters 25 and 26. Boundary syndromes may also include
epileptiform EEG patterns of uncertain significance such as
“triphasic” waves or periodic discharges (lateralized or general-
ized), where there is substantial uncertainty in differentiating
whether this represents seizure activity or not (see next section)
(41,72). “Subtle” SE often includes prominent myoclonus, usually
occurring in seriously ill patients in stupor or coma.
 Chapter 29 ■ Nonconvulsive Status Epilepticus 599

Tabl e 2 9 . 1

Classification of Nonconvulsive Status Epilepticus Syndromes

1. Neonatal and infantile syndromes (including):

• West syndrome, Ohtahara syndrome
• Severe myoclonic epilepsy of infancy (SMEI; i.e., Dravet syndrome)
2. Childhood syndromes (including):
• Genetic-based illnesses such as ring chromosome 20 syndrome; Angelman syndrome, Rett syndrome
• Benign childhood epilepsies without a discrete genetic origin, for example, Panayiotopoulos syndrome
• Childhood epileptic encephalopathies such as Landau–Kleffner syndrome and electrical status epilepticus in sleep (ESES)
Febrile SE
3. Childhood and adult forms of NCSE:
(a) With epileptic encephalopathy:
(i) For example, within the Lennox–Gastaut syndrome, with atypical absence SE or tonic SE, and the continuation of
LGS into adulthood
(ii) Other epileptic encephalopathies
(b) Without epileptic encephalopathy:
• Typical absence SE of idiopathic generalized epilepsies
• Focal NCSE, sometimes referred to as “aura continua” with sensory, autonomic, or cognitive symptoms
• Complex partial SE: limbic or nonlimbic
• NCSE that follows the convulsions of generalized seizures or generalized convulsive SE
• “Subtle” SE with myoclonic SE following convulsions
4. Later adult NCSE: including de novo absence SE of late onset
5. “Boundary syndromes”:
• Some “atypical absence” SE of childhood
• Epileptic behavioral disturbances or psychosis
• Other forms of epileptic encephalopathy and coma due to brain injury (e.g., anoxia) with epileptiform discharges,
including “subtle” status.
• Drug or metabolic confusional states with epileptiform discharges (generally not considered true NCSE)

Modified from Shorvon (14).

EARLY LIFE, AGE-RELATED
NONCONVULSIVE STATUS
EPILEPTICUS SYNDROMES
Many neonatal seizures and NCSE have minimal clinical mani-
festations and remain underdiagnosed if EEG monitoring is not
utilized (73,74) (see also Chapter 25), especially in neonates with
severe brain injury (75) and in pharmacologically paralyzed
infants. Electrographic patterns of neonatal seizures also differ
from those of older patients, often remaining localized to rela-
tively small brain areas, likely due to incomplete myelination and
neuronal migration (76). Multifocal seizures (more than three
foci) are common, especially with multifocal or diffuse brain
injury. Some seizures show evolution in frequency, amplitude,
morphology, or spatial distribution, but many show little or no
such evolution (76). Morphologies include sharply contoured,
sinusoidal, or rounded waveforms in the alpha, theta, and delta
frequency range, but the frequency and morphology may vary
within a single seizure and from seizure to seizure (Fig. 29.1A–C).
Many neonatal seizures last 2 to 3 minutes but recur frequently;
prolonged continuous seizures are less frequent (77).
Conversely, not all abnormal-appearing neonatal movements
are seizures (74). Proper use of EEG and EEG monitoring can
help to avoid inappropriate AED treatment of movements such
as some swimming or pedaling motions, and stimulus-sensitive
clonus or myoclonus, when they are not epileptic in origin. In
infants, particular attention must be paid to exclude artifacts
mimicking electrographic seizures.
600 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

A
Fp1-F3

F3-C3

C3-P3

P3-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

ECG
50 uV 2 sec

B
Fp1-F3

F3-C3

C3-P3

P3-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

ECG
50 uV 2 sec
 Chapter 29 ■ Nonconvulsive Status Epilepticus 601

C
Fp1-F3

F3-C3

C3-P3

P3-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

ECG
50 uV 2 sec

Figure 29.1 Neonatal status epilepticus in a 36-week conceptional age neonate with multifocal infarcts due to a placen-
tal abruption following a maternal motor vehicle accident. Multifocal seizures had various electrographic patterns. A:
Monomorphic sharply contoured delta activity. B: Continuation of the same seizure, showing evolution into rhythmic
spike-and slow-waves at 2.5 Hz, maximal in the right posterior quadrant. C: Later in same seizure, now with rhythmic
spike-and-wave over the left hemisphere and independent lower voltage ictal delta activity in the right hemisphere.

Several neonatal seizures and SE syndromes have had genetic
characterization. Dravet syndrome (severe myoclonic epilepsy of
infancy) was originally described as being manifested clinically
by prolonged obtundation, referred to as “obtundation status”
(78,79). Infantile epileptic encephalopathy, or Ohtahara syn-
drome, typically presents with greater flexor, extensor, or tonic
spasms and prominent suppression bursts on the EEG (80).
Many of the most prominent pediatric NCSE syndromes
associated with underlying developmental delay or regression
demonstrate activation of epileptiform abnormalities during
sleep (81,82). ESES implies activation by sleep of persistent
epileptiform activity suggestive of SE, although not all children
with these EEGs have clinical seizures. The waking EEG usually
shows infrequent focal or occasionally generalized epileptiform
discharges, with the spike component usually more prominent
than the slow wave. During rapid eye movement (REM) sleep,
epileptiform discharges are similar to those on the waking EEG.
Some children have an associated ESES syndrome, with
ESES on the EEG along with progressive cognitive dysfunction,
often referred to as an epileptic encephalopathy (83). One such
syndrome, continuous spikes and waves during slow sleep
(CSWS), is an age-related pediatric epilepsy syndrome, usually
starting late in the first decade of life in children who were neu-
rologically normal before seizures began (84). Seizures include
generalized convulsions and often evolve to become atypical
absence and atonic seizures with dangerous falls (81,85). EEG
during wakefulness and REM sleep shows infrequent focal or
generalized epileptiform discharges. Non-REM sleep prompts
continuous generalized or bilaterally synchronous (if some-
times asymmetric) spike-and-wave complexes at 1.5 to 3.5 Hz
(Fig. 29.2). Continuous spike-wave patterns occupying at least
85% of slow-wave sleep are required for the diagnosis of CSWS
(Fig. 29.2) (85,86). Most children with CSWS have focal or gen-
eralized seizures, global behavioral problems, and regression of
cognitive function, more than in language alone (85,87). The
epilepsy is often severe and can include many seizure types;
some children have prolonged atypical absence SE (81).
CSWS may last for years, but the typical EEG pattern usually
occurs between ages 5 and 15 years. The EEG eventually nor-
malizes, including during sleep, and almost all seizures resolve,
602 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

FP1-F3

F3-C3

C3-P3

P3-O1

FP2-F4

F4-C4

C4-P4

P4-O2

FP1-F7

F7-T3

T3-T5

T5-O1

FP2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
75 uV 1 sec

Figure 29.2 Continuous spike waves during slow-wave sleep in a 9-year-old girl with mild developmental delay.

but cognitive and behavioral problems persist in most cases
(81,85). In LGS, the CSWS syndrome may include a prominent
component of negative myoclonus or asterixis, at times explain-
ing apparent atonic seizures and dangerous falls (27). These
syndromes are covered more extensively in Chapter 26.
The Landau–Kleffner syndrome (LKS) can be considered
another “boundary” syndrome and may overlap with CSWS and
other “benign” focal epilepsy syndromes. LKS is characterized by
acquired aphasia, seizures, and a behavioral disorder presenting
in children with previously normal language development,
between the ages of 2 and 4 years (88). A gradual deterioration,
specifically in language and speech output, may begin with
apparent word deafness or a form of auditory agnosia (89), but
there can also be hyperkinetic behavior disorders. LKS exhibits
clinical seizures in most cases, though seldom overt SE. It is
unclear whether the deterioration in language function is the
result of epileptiform activity or if the language deterioration
and epileptic activity (both clinical and EEG) result from a com-
mon underlying pathologic condition. Most LKS cases are of
unknown origin; a few cases are due to symptomatic lesions
(90). Corticosteroids are the most common treatment for LKS
(91), but most children have persistent language deficits (89,92).
The waking EEG in LKS may be normal or show unilateral
or bilateral temporal spike-and-wave discharges. There are
sleep-activated bursts of diffuse or multifocal spike-and-wave
discharges at 1.5 to 3.5 Hz (ESES), most commonly in anterior
and middle temporal regions, but also in temporoparietooccip-
ital regions, some with a very broad, nearly generalized field
(Fig. 29.3) (86,93). In contrast to CSWS, epileptiform discharges
are present in less than 85% of slow-wave sleep, and may occur
during REM sleep (94). The EEG manifestations usually appear
between 3 and 5 years of age and resolve at about 15. The EEG
may improve over time, with or without treatment (90).
LKS and other ESES syndromes are often treated with AEDs,
especially BDZs, with a goal of eliminating epileptiform dis-
charges and the idea (not entirely proved) that the epileptic
process contributes to the cognitive deterioration. Although
episodes of ESES may recur frequently, the prognosis for seizure
control is good, but CSWS has a poor prognosis for cognitive
function. LKS and CSWS are rare syndromes, each accounting
for about 0.2% of all pediatric epilepsy patients (95).
In ring chromosome 20 syndrome, particularly in patients
from Japan between the ages of 13 and 31 years, NCSE consists
of prolonged confusional states, with or without motor phe-
nomena, and prominent bilateral or generalized high-voltage
slow waves with intermixed spikes on the EEG (96). The
epilepsy is characterized by intermittent motor seizures with
confusion, staring, head turning, mutism and meaningless
 Chapter 29 ■ Nonconvulsive Status Epilepticus 603

Fp1-F3
F3-C3

C3-P3
P3-O1
Fp2-F4
F4-C4

C4-P4
P4-O2
Fp1-F7
F7-T3

T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6

T6-O2
Fz-Cz
Cz-Pz
LUC
RLC
EKG
75 uV 1 sec

Figure 29.3 A 4-year-old boy with language regression, Landau–Kleffner syndrome. EEG during sleep shows bilateral
1-Hz centrotemporal sharp waves, more prominent over the right hemisphere.

utterances, facial flushing, and limb shaking. Impulsive behav-
ior, inappropriate responses, and myoclonus occur in episodes
up to several times daily.
Although NCSE in children is perhaps best considered in the
context of age and brain development, the descriptions of NCSE
in adults in this chapter follow more familiar International
League Against Epilepsy (ILAE) seizure types and syndromes,
focusing on the type of seizure apparently persisting and caus-
ing the patient’s clinical deficit. They are organized into focal
and generalized forms, acknowledging that seizures can progress
from focal onset to generalized clinical and EEG manifestations.
Myoclonic, tonic, and clonic SE are detailed further in the con-
vulsive SE chapter because of their motor manifestations.
FOCAL NONCONVULSIVE STATUS
EPILEPTICUS
Focal status epilepticus may be either SPSE with no impairment
of consciousness, or CPSE with consciousness impaired. SPSE
presents with ictal symptoms reflecting the involvement of dis-
crete regions of brain, such as motor, sensory, special sensory,
psychic, or autonomic symptoms (97). Focal SE with motor
components is described in Chapter 28. Nonmotor (noncon-
vulsive) forms of focal SE include persistent sensory distur-
bances (e.g., visual hallucinations) and autonomic and
cognitive deficits, including affective, amnestic, and language
disturbances such as aphasia (42).
Overall, focal onset of symptoms and electrographic abnor-
malities is more common than SE with a generalized onset
(98). Focal NCSE is seen frequently after strokes or other acute
brain injuries, and should be suspected when patients do not
stabilize or improve as expected. In some critically ill patients,
NCSE consists of brief cyclical electrographic seizures occur-
ring every 5 to 10 minutes over several hours (99). Focal lesions
from vascular causes, trauma, developmental abnormalities
(such as heterotopias and other migration abnormalities),
and mitochondrial disorders are among many possible causes
(100,101).
Focal Sensory Nonconvulsive Status Epilepticus
Persistent sensory seizures (without motor accompaniment) have
been called “aura continua” (102,103). Infrequently, they may last
for hours. Simple partial sensory status epilepticus is identical to
the prolonged continuation of an isolated aura continua. In these
cases, local inhibitory function is sufficient to prevent spread of
604 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Fp1-F3
F3-C3
C3-P3

P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2

Fp1-F7
F7-T3
T3-T5
T5-O1

Fp2-F8
F8-T4
T4-T6

T6-O2
Fz-Cz
Cz-Pz

LUC
RLC
EKG 75 uV
1 sec

Figure 29.4 An 8-year-old boy with simple partial status epilepticus arising in the right occipital region. Seizures con-
sisted of seeing pulsating green and red circles in the left visual field.

the seizure more broadly in the brain. If, however, these seizures
spread to involve areas of the brain necessary for responsiveness
(for a long period), this becomes CPSE. Occasionally, benign
rolandic epilepsy, in which there is speech arrest, drooling, dys-
phagia, facial weakness, head deviation, and mild confusion, may
be prolonged enough to constitute NCSE (104,105).
Pure sensory SE is generally considered rare, but this may
reflect a detection bias as an EEG may show little at the time of
SE (97), and persistent sensory disturbances are often attrib-
uted to transient ischemia (106) or other nonepileptic causes.
In one report, four patients had focal sensory NCSE lasting up
to several days, without motor or behavioral symptoms or alter-
ation in responsiveness (107). Some were frontal or temporal in
origin. They appeared to be less common than auras progress-
ing to more obvious seizures, and far less common than CPSE.
Focal lesions in temporal, parietal, and occipital cortices may
produce SE with persistent cognitive or sensory symptoms, with
hallucinations of any sensation, including visual, auditory, olfac-
tory, gustatory, somatosensory, or vertiginous perception (103).
Olfactory hallucinations from temporal epilepsy may be the
most common. Some SPSE consist of auditory sensations alone,
with epileptiform discharges confined to Heschl’s gyrus (108).
Transient cortical blindness or visual field loss (sometimes
referred to as status epilepticus amauroticus) can be a manifesta-
tion of occipital SPSE (109), although simple visual hallucina-
tions and ictal blindness are usually brief (110). Occipital seizures
can also include macropsia, micropsia, misperception of spatial
orientation, hallucinations of faces or animals, or simple patterns
of color and light (110) (Figs. 29.4 and 29.5). There may be sco-
tomata or simple visual hallucinations predominating in the cen-
tral visual field, but complex hallucinations are also possible
(111). Slightly more anteriorly, involvement of the temporopari-
etooccipital junction may produce nystagmus with contraversive
eye deviation (20).
The “positive” sensory symptom of pain during SPSE appears
even more rare, although shorter painful seizures have been
reported (103,112). Some are parietal in origin, in the somatosen-
sory area. Parietal NCSE can also be manifested solely by persist-
ent paresthesias (51). Other unpleasant epileptic sensory seizures
may include nausea, ictal fear, anorexia, poorly described visceral
sensations, and the distress of “abdominal epilepsy” originating
in mesial temporal areas, especially involving the amygdala
(112–114). Autonomic symptoms are reported increasingly fre-
quently, particularly in children (115).
Some focal SE involve a larger brain area than a typical simple
focus, and could be called “regional.” The most common (usually
pediatric) form of regional SE is the Panayiotopoulos syndrome,
in which an exact focus is often difficult to determine (116). This
syndrome is an age-dependent epileptic susceptibility that is usu-
ally idiopathic and benign, although perhaps 15% are due to
focal lesions. It affects about 13% of all children aged 3 to 6 with
nonfebrile seizures. Manifestations are primarily autonomic,
with ictal emesis and occasional syncope, usually arising from
sleep. Half of the episodes last longer than 30 minutes and thus
constitute autonomic SE. Most children have few episodes and
typically do better without treatment, and the long-term risk of
 Chapter 29 ■ Nonconvulsive Status Epilepticus 605

A
Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
75 uV 1 sec

B
Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
75 uV 1 sec

Figure 29.5 Focal status epilepticus in a 14-year-old girl with right occipitoparietal cortical dysplasia. A: Seizure begins
with rhythmic beta activity over both occipital regions, more prominent on the right. B: Seizure evolves into high-voltage
3- to 4-Hz spikes in the left occipital region, and spikes with slower frequencies in the right occipital region. Seizures
recurred every 2 to 3 minutes for several hours before presentation to the ER.
606 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

A
Fp1-F3

F3-C3

C3-P3

P3-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
50uV 1 sec

B
Fp1-F3

F3-C3

C3-P3

P3-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
50 uV 1 sec

Figure 29.6 A: A 49-year-old woman with mild developmental delay who presented with an anterior aphasia. She was
able to follow commands. EEG during aphasia shows repetitive sharp waves in the left temporal region at 2 Hz. B: Same
patient after 2 mg IV lorazepam. EEG shows resolution of ictal activity and a left midtemporal spike. The aphasia resolved.

epilepsy is small (116). Earlier described in correlation with
occipital spikes, this syndrome may have discharges elsewhere or
multifocally and is probably better considered a regional (but not
quite generalized) form of seizures and SE.
Amnestic and Aphasic Status Epilepticus
Persistent memory dysfunction often follows complex partial
seizures in the postictal period, and it may follow, and be due to,
an episode of status epilepticus (117,118). Episodes of pure
amnesia (with preserved cognitive function in other realms) can
also be due to individual epileptic seizures, referred to as “tran-
sient epileptic amnesia (TEA)” (119). TEA almost always occurs
in patients with earlier epilepsy and is often associated with
additional persistent memory deficits. Seizures preferentially
affecting memory may require involvement of medial temporal
structures bilaterally, as shown by depth electrode recording
(120). TEA typically occurs in older patients, thus overlapping
clinically, and often difficult to distinguish from (nonepileptic)
transient global amnesia. Many episodes of TEA are prolonged
beyond 30 minutes, constituting NCSE with amnesia as the sole
clinical manifestation (119). Other amnestic syndromes can be
caused by frontal or generalized NCSE (121); the frontal dis-
charges suggest an inability to access extant memory during the
seizures. Various “psychic” phenomena, including apraxia and
acalculia, as well as prolonged mood changes, such as depression
and panic attacks, can also occur in NCSE (122).
Aphasic SE, with preserved responsiveness and a clinical
deficit restricted to language dysfunction alone, is a focal or
regional NCSE (while CPSE implies a disturbance in respon-
siveness) (123–127). Careful language testing is necessary to
determine that the deficit is truly an aphasia (125,128).
Speech arrest alone can result from seizures in many areas. A
focal neurologic deficit in speech production may be mistaken
for other syndromes such as transient ischemia or psychiatric
illness.
A wide variety of aphasic NCSE types has been characterized
by appropriate and thorough language testing—with preserved
responsiveness and other cognition except for the aphasia
(128). Often, the ongoing epileptiform discharges on EEG cor-
respond anatomically to the area associated with a particular
aphasia (Fig. 29.6), such as posterotemporal focal discharges
in patients with clinically evident Wernicke’s-type aphasia
(124,126). Those with more anterior aphasias (some with fron-
totemporal discharges on EEG) are often unable to speak but
may retain verbal memory and respond appropriately to verbal
commands, giving evidence of relatively preserved comprehen-
sion (122,123). Some have alexia and reduction in spontaneous
speech. Still, the correlation to classical (usually stroke-defined)
areas of clinical involvement for different types of aphasia is far
from precise (128).
Diagnosis of Focal Nonconvulsive
Status Epilepticus
Focal NCSE is usually more subtle in presentation than either
focal motor SE or generalized forms of SE. SPSE manifested by
sensory or cognitive symptoms alone can be very difficult to
Chapter 29 ■ Nonconvulsive Status Epilepticus 607
diagnose and can be confused with many other illnesses (see
Differential Diagnosis above). The EEG is extremely useful in
establishing a diagnosis, but not all patients have diagnostic
findings. In one report of six patients with prolonged neu-
rocognitive deficits and evidence of focal-onset NCSE, all had
rapid rhythmic epileptiform discharges on EEG and eventual
resolution of all symptoms with treatment (122).
EEG
Surface EEGs, even when recorded during the symptoms, have
limited sensitivity for detection of ictal discharges in SPSE, at
least in part because the seizures may involve a small volume of
brain tissue. The EEG may be normal or show nonspecific
changes (129). One report described 87 well-characterized sim-
ple partial seizures in 14 patients, and just 20% of seizures had
identifiable epileptiform abnormalities on scalp EEG (97). Two
thirds of seizures had primarily sensory symptoms, and these
showed ictal EEG changes in just 15%. In another study, only
20% to 35% of simple partial seizures had ictal correlates on
surface EEG (130). Sometimes, SPSE can remain undiagnosed
until invasive EEG monitoring is performed—usually carried
out because of (other) refractory seizures (131,132).
When SPSE is visible on EEG, it usually shows ictal onset
with a single topographic EEG maximum, followed by evolution
in frequency, voltage, and distribution. Focal fast-frequency dis-
charges, rhythmic waveforms with evolving morphology, repeti-
tive epileptiform discharges, or regional or lateralized alterations
in EEG background activities may be seen (129,130). If SE is dis-
continuous, the background activity between seizures is typically
slow and disorganized, with occasional focal interictal epilepti-
form discharges.
Electrographic focal SE is seen in some stuporous or coma-
tose patients with no clear clinical signs of seizure activity
(133). Seizures may be either continuous or repetitive, and EEG
patterns are similar to those in SPSE and CPSE. Such partial SE
is not rare after strokes or other acute brain injuries and should
be suspected when patients in those settings do not stabilize or
improve as expected. EEG should be performed for any ICU
patient in whom mental status changes are unexplained or
appear out of proportion to the degree of acute neurologic
injury. Patients with suspected SPSE also warrant an MRI scan
to look for an underlying focal lesion.
Complex Partial Status Epilepticus
The first definite case of CPSE was reported by Gastaut and Roger
in 1956 (13). In the 1970s, CPSE was considered rare and was the
subject of isolated case reports (117,118), and by 1985 only 17
clearly identified cases had been published (134). With the realiza-
tion that absence SE is a separate category rather than simply any
NCSE with generalized discharges, CPSE was eventually recog-
nized as more common (17,18,135), although it remains under-
diagnosed (136). Other terms for CPSE include temporal lobe SE
(often not an accurate localization), prolonged epileptic fugue
state, and psychomotor SE (134,137–140).
The definition of SE as a seizure or series of seizures without
recovery for 30 minutes (141) allows for the possibility that
608 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
75 uV 1 sec

Figure 29.7 “Psychic” status epilepticus in a 41-year-old woman with right mesial temporal sclerosis. She presented with
mild confusion but could follow most commands. She did not answer questions and had bizarre speech, repeating “you
are my Norm, you are my one true love” continuously.

CPSE can consist of a single very long complex partial seizure
or a series of recurrent complex partial seizures without clinical
normalization between seizures (Fig. 29.10) (118). CPSE is
probably the most common form of NCSE in adults who are
ambulatory or not critically ill.
CPSE begins with a focal-onset seizure, progressing to
involve more of the brain such that responsiveness is impaired,
and then with prolongation or recurrence. CPSE may last
hours, days, or even months (17,142). The most common
causes are new acute vascular disease, or old strokes with a
more recent precipitant such as infection or metabolic derange-
ment (44,143). Other infections (e.g., encephalitis), tumors,
congenital developmental abnormalities, and vascular malfor-
mations are possible. The diagnosis of CPSE should be consid-
ered in patients who are partially or completely unresponsive,
especially those with earlier epilepsy or vascular disease.
While most patients have earlier epilepsy (144), CPSE may
arise in patients without prior seizures. CPSE is seldom
reported in children, but it may be significantly underrecog-
nized and occur in those with substantial developmental delay
and neurologic deficits, and even in those with other seizure
types (145). From the incidence of SE in different populations
with epilepsy, it has been estimated that CPSE occurs in 35 cases
per million population per year, but possibly five times as often
in developmentally delayed patients (144).
CPSE may present as abnormal behavior or diminished
responsiveness associated with lateralized epileptiform seizure
activity. It may begin with a simple partial seizure, or conscious-
ness may be impaired at the onset. Impairment of consciousness,
which may cycle or fluctuate, ranges from almost nondiscernible
clouding of higher cortical function to coma (44,99,118,145).
CPSE may include an “epileptic twilight state” with a lack of
responsiveness or confusion, and bizarre, and particularly fluctu-
ating, behavior or paranoia (17,117,118,134,140) (Fig. 29.7). The
extensive variety of clinical presentations of CPSE makes EEG
essential to distinguish CPSE from other causes of altered mental
status.
Lip-smacking, other oroalimentary automatisms, lateralized
limb automatisms and dystonic posturing, eye deviation, and
nystagmus are typical of CPSE; myoclonus is rare (44,140,146).
Fear, aggressivity, irritability and anxiety, and stereotyped, com-
plex automatisms are more frequent with CPSE than with
absence SE (146), but when the automatisms or movements are
minimal, CPSE can be difficult to distinguish from absence SE.
EEG
In contrast to SPSE, most CPSE shows discernible surface EEG
changes. The EEG patterns of CPSE are variable, reflecting
differences in the location of ictal onset zones and propaga-
tion pathways, but they are often very similar to those seen in
 Chapter 29 ■ Nonconvulsive Status Epilepticus 609

Figure 29.8 Quantitative EEG showing complex partial status epilepticus with recurrent seizures originating independ-
ently from both hemispheres, in a 45-year-old woman with multiple sclerosis and malfunction of an intrathecal baclofen
pump. The bottom two lines of the compressed spectral array (CSA) show peaks (see arrows) of rhythmic activity of alpha
and theta frequencies indicative of seizures. Expanded “usual” EEG below shows, (A) right hemisphere seizures, corre-
sponding to the thick arrows on the CSA, and (B) left hemisphere seizures corresponding the thin arrows on the CSA above.
(See color insert)

isolated complex partial seizures. CPSE may be characterized
by either repetitive complex partial seizures, each showing
focal onset, with background slowing between seizures, or by
continuous ictal rhythms, corresponding to cycling or contin-
uous clinical manifestations (70,99,134,147,148). Rarely,
ictal patterns arise independently from both hemispheres
(Fig. 29.8). Waveform morphologies include repetitive epilep-
tiform spikes; spike-and-slow wave discharges; rhythmic theta,
delta, or alpha frequencies; or rhythmic low-voltage fast activ-
ity (45). Occasionally, the EEG can be normal or obscured by
artifact (149).
Often, early EEG abnormalities are focal or lateralized
throughout the episode (130). Focal onset is usually followed by
evolution in distribution, voltage, and frequency, in commonly
recognized patterns of electrographic seizures (110,150,151).
Epileptiform discharges may merge gradually to produce con-
tinuous focal seizure activity. Many have subsequent spread to
involve a greater area of the cortex (Fig. 29.9) and result in dif-
fuse or generalized epileptiform discharges difficult to distin-
guish from those of absence SE (18,44,45). As CPSE progresses,
the epileptiform discharges often increase in voltage, slow in
frequency, and broaden in distribution. Later, seizures may
become fragmentary, interrupted by brief periods of back-
ground attenuation (152). Between discrete seizures, the back-
ground shows focal slowing, focal epileptiform discharges, or
periodic lateralized epileptiform discharges (PLEDs) (152).
Temporal Versus Frontal Complex
Partial Status Epilepticus
CPSE may arise from many brain regions. Many episodes rep-
resent prolonged temporal lobe seizures, as demonstrated
clearly by depth electrode recording (153), but in another study
of eight patients with depth electrode-recorded episodes of
CPSE, all had frontal origins for the seizures (145). At least on
610 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz
Cz-Pz

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz
Cz-Pz

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz 400 uV
Cz-Pz 2 sec

Figure 29.9 Left temporal seizure evolution in a 21-year-old woman with intractable epilepsy admitted with confusion
and aphasia. She had complex partial status epilepticus consisting of frequent discrete electrographic and clinical seizures.
EEG in three horizontal sections, of 18 channels each. Top section: one of frequent discrete electrographic seizures arises
in the left frontotemporal region, evolving with higher voltage, slightly lower frequency sharp waves in the left temporal
region; with spread to the right side in the middle section; eventual slowing of discharges and interruption by brief nondis-
charging periods in the bottom section. Just after this, there was generalized, 1.5 Hz slowing for 5 seconds and then a very
suppressed background in all areas as the seizure ended.

the scalp EEG, parietal CPSE is usually nonlocalizing or falsely
localizing (154). Overall, extratemporal onset of CPSE may be
more common, including in children (115).
Extratemporal foci may produce visual, auditory, somes-
thetic, or vestibular hallucinations; a perception of warmth;
changes in facial color; nausea; or unilateral arm automatisms
or posturing. When seizures spread to involve the amygdala and
hippocampus, there may be chewing movements, lip-smacking,
and gesticulatory automatisms (118).
Frontal CPSE generally has less impairment of conscious-
ness and fewer fluctuations than does temporal lobe CPSE
(146). Frontal CPSE may recur frequently, but it is misdiag-
nosed often. Temporal CPSE includes more fear, anxiety, anger,
irritability, aggressiveness, agitation, and simple and complex
automatisms (146). Psychomotor slowing is more frequent in
frontal than in temporal lobe CPSE, but about the same as in
absence SE, and CPSE with a frontal origin may be very difficult
to distinguish from absence SE.
Temporal Complex Partial Status Epilepticus
The clinical manifestations of temporolimbic CPSE are exten-
sive. “Experiential” sensations involving memories of percep-
tions or experiences, often with an emotional overtone, indicate
limbic structure involvement of seizures (155) (Table 29.2). For
the seizure to be “experienced,” or remembered, there should
also be involvement of both medial temporal and neocortical
structures (102). Seizures involving limbic areas of the hip-
pocampus and amygdala interfere with responsiveness, becom-
ing “dyscognitive” seizures and CPSE (102). In temporal CPSE,
dyscognitive focal SE may overlap with aura continua or evolve
from it (153).
Temporal CPSE may have a sudden or gradual onset (102).
Hippocampal, amygdalar, and amygdalohippocampal CPSE
often present with cyclical changes in behavior (99,102,118). An
early report described a “twilight state” with partial responsive-
ness and intermittent speech arrest (118). Some patients were
completely unresponsive, with motionless staring. Others had
 Chapter 29 ■ Nonconvulsive Status Epilepticus 611

Tabl e 2 9 . 2 CPSE originating in the temporal lobes may have different

Symptoms of Temporal Lobe Complex Partial
Status Epilepticus
“Experiential” seizures combining perceptual and memory
abnormalities, and mood or affect disturbances, including
depression and anxiety.
Memory abnormalities, including déjà vu, jamais vu, and so
on or gaps in memory.
“Psychic” auras, including forced thinking, and behavior
mimicking psychiatric disorders.
Emotional symptoms, including fear (if involving the
amygdala), sadness, and distress.
Alterations in perception of reality, including
depersonalization or out of body experience.
Adapted from articles by Wieser (102,153) and Gloor (155).
oroalimentary or more complex automatisms, vocalizations,
and perseverative gesticulations. A more anterior seizure origin
or spread may induce bizarre limb automatisms or wandering,
as in a fugue state (poriomania) (118). Behavior may fluctuate
and become very strange. Memory for the event is often oblit-
erated or impaired.
manifestations from seizures starting elsewhere and spreading
to temporal structures. Opercular, frontal, and occipital origin
seizures can all spread secondarily to involve mesial temporal
regions. Some seizures with temporal lobe-related manifesta-
tions actually originate in frontal lobes, with the initial behav-
ioral features representing frontal dysfunction.
EEG
CPSE arising from mesial temporal regions usually begins with
unilateral temporal 5 to 7-Hz rhythmic activity, which appears
within 30 seconds of the initial clinical symptoms or signs (150)
(Figs. 29.9 and 29.10). Interictally, there is focal slowing or
attenuation on the side of seizure onset. In neocortical tempo-
ral seizures, discharges are more widely distributed, sometimes
over the entire hemisphere. As compared to the EEG findings in
medial temporal CPSE, the seizures of neocortical temporal
CPSE often begin with slower frequencies, show less stable fre-
quency and voltage, and appear later after clinical signs (151).
Frontal Complex Partial Status Epilepticus
Thomas and colleagues described two types of frontal lobe (FL)
NCSE (139). Type 1 had continuous mood and behavioral dis-
turbances, with mild hypomania, affective disinhibition (or
conversely, indifference), and subtle impairment of cognition,
but no overt confusion. Patients could exhibit a blank facial
expression, lack of spontaneous emotion, and decreased verbal

FP1-F3
F3-C3
C3-P3

P3-O1
FP2-F4

F4-C4
C4-P4
P4-O2
FP1-F7
F7-T3
T3-T5
T5-O1
FP2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz

Cz-Pz
LUC
RLC
EKG 75 uV 1 sec

Figure 29.10 Right temporal complex partial seizure in a 32-year-old man with mesial temporal sclerosis, presenting with
status epilepticus consisting of repetitive seizures. The early EEG before this seizure shows widespread postictal slowing
following the previous seizure.
612 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 29.11 Frontal lobe complex partial status epilepticus in a 56-year-old woman with a history of traumatic brain
injury. EEG shows irregular 5- to –8-Hz spikes and polyspikes over the left frontal region.
fluency. Some could recall events that occurred during the spell
and could carry out routine activities, such as eating. There
were simple gestural automatisms such as scratching, rubbing,
or picking at clothes, but complex bipedal or bimanual automa-
tisms were not observed. Type 1 FL-CPSE led to no morbidity
in later cognitive function.
The rarer type 2 FL-CPSE had greater impairment of con-
sciousness, often with cyclic fluctuations. There was marked
behavioral disturbance, temporospatial disorientation, confu-
sion, prominent perseveration, and occasional catatonic stupor.
Some patients required restraints because of aggressiveness.
There was occasional head and eye deviation, and perioral
myoclonia. Patients were universally amnestic for the episode.
EEG
While ictal EEG is nonlocalizing or normal in many patients
with frontal lobe complex partial seizures (149), scalp EEGs
eventually show bilateral sharp-and-slow-wave activity in most
(145) (Fig. 29.11). Some patients have focal rhythmic alpha or
beta activity at the onset of seizures originating near scalp elec-
trodes (149), but the scalp EEG often shows no changes until the
seizure spreads to medial temporal structures, producing ictal
patterns typical for medial temporal seizures. In type 1 FL-CPSE
(above), the EEG showed a unilateral frontal focus; in type 2, there
were bilateral asymmetric frontal foci with an abnormal back-
ground (139). Spread of epileptiform activity from a unilateral
frontal focus may lead to bilaterally synchronous or generalized
discharges on EEG, indistinguishable from those of absence SE,
and often with a very similar clinical appearance (138,156).
Complex Partial Status Epilepticus:
Treatment and Outcome
Usually by the time of diagnosis of CPSE, seizures have been
present for at least 30 minutes, so treatment should begin
promptly with intravenous AEDs. Possibly because of the asso-
ciation with vascular disease and other lesions, and prior focal
epilepsy, CPSE is often harder to treat than primary generalized
SE and is more likely to recur (17). Fortunately, patients with
CPSE in the setting of earlier epilepsy are very likely to survive
and generally warrant a less aggressive approach than does
GCSE (157). CPSE is usually treated in the same way as focal-
onset GCSE, but somewhat less aggressively if there is an earlier
history of epilepsy, and CPSE is usually more easily treated than
is GCSE (158,159).
The outcome from episodes of CPSE depends primarily on
the etiology, and morbidity often derives from the unresponsive
state and from complications of the underlying illness causing
the SE (particularly encephalitis and strokes), rather than from
the CPSE itself. In one study, there was no difference in mortal-
ity between patients with CPSE (with focal discharges on EEG)
and those with GCSE, but the CPSE included patients with sec-
ondary generalization (160). With CPSE, mostly of frontal ori-
gin, episodes of longer duration had no worse an outcome than
those with shorter durations (145).

Early reports on CPSE included very few patients. Almost all
returned to normal or “baseline cognitive function” (134,139,
161), with resolution of all neurologic deficits over time, but
this may take weeks or months (117). In one CPSE series, none
of the 20 patients had lasting cognitive deterioration, and 5 had
meticulous neuropsychologic assessment (17). There are
reports of prolonged memory and other cognitive deficits, for
weeks or months after NCSE (51,117,118,122,143), but many
resolve eventually. In some cases, both clinical and imaging
abnormalities can resolve completely over a period as long as a
year (17,139).
Long-term clinical effects of CPSE could also include wors-
ened cognitive function or increased subsequent seizure fre-
quency. Most outcome studies, however, are of GCSE, and it
correlates with worsened seizure control, but this might be
attributed to the underlying illness rather than to the episode of
SE itself (162).
GENERALIZED STATUS EPILEPTICUS
Absence Status Epilepticus
Although “petit mal intellectual” was postulated in the 19th cen-
tury, Lennox demonstrated in 1945 that altered mental status
could be caused by absence seizures. He suspected that a series
of individual absence seizures and EEG discharges could persist
for much longer than was generally recognized (11). Absence
SE was first described as “status epilepticus in petit mal” by
Schwab in 1953 (163), recognizing the same EEG patterns seen
in petit mal epilepsy by Lennox earlier. The prolonged confu-
sion and slowed behavior were labeled “spike-wave stupor” by
Niedermeyer and Khalifeh (164). Typical absence SE has also
been called petit mal status, minor epileptic status, absence con-
tinue, epilepsia minoris continua, and status pyknoepilepticus.
Early reports and classification of NCSE (i.e., in the 1970s
through the 1990s [!]) generally divided it into two types:
“absence SE” (all NCSE with generalized epileptiform discharges
on the EEG) and “complex partial SE” (all others felt to have a
focal seizure onset). CPSE could have focal EEG discharges or a
clear focal onset clinically (44). The oversimplification of the
generalized forms is no longer tenable.
Absence SE is a pure form of primary generalized epilepsy
with clinical episodes lasting over 30 minutes. In its classic
appearance, it occurs in patients with earlier absence epilepsy or
other idiopathic generalized epilepsies (IGEs) (165–167).
Absence epilepsy (see Chapter 26) begins anywhere from age 3
years until early adulthood, but usually in childhood or adoles-
cence; seizures usually decrease markedly in the third decade. The
incidence of absence SE is about one case per million population
per year (144), and about 10% of adults who still have absence
epilepsy will have at least one episode of absence SE (168).
Absence SE occurs in nearly all IGE syndromes, including juve-
nile absence epilepsy (Fig. 29.12), JME, perioral myoclonia with
absences, eyelid myoclonus and absence, and IGEs with phantom
absences (169)—even when those syndromes are usually mani-
fested by other types of seizures, for example, myoclonus in JME
(167) (Table 29.3). NCSE is uncommon in JME (170).
Chapter 29 ■ Nonconvulsive Status Epilepticus 613
Precipitants include sleep deprivation, alcohol ingestion,
metabolic abnormalities, and intercurrent illnesses. Many med-
ications have been implicated in the precipitation of absence
SE, especially neuroleptics, tricyclic antidepressants, lithium,
and BDZ withdrawal. AEDs such as phenytoin and carba-
mazepine may contribute (171). A generalized convulsion may
set off an episode of absence SE or end it. In the evaluation of
patients with absence SE, MRI scans are almost invariably nor-
mal; any focal abnormalities are likely incidental.
Typical absence SE starts abruptly, without warning, and
consists of individual seizures in rapid succession without
recovery over a long period or, more commonly, a single pro-
longed episode with continuous epileptiform discharges and
impaired consciousness (172). The clinical presentation can
be very subtle, with cognitive deficits discernible only on for-
mal neuropsychologic testing. There may be just a relatively
hard-to-diagnose change in responsiveness, with preserved
alertness (165). Cognition may fail when more complex
planning is needed. Patients may remain reactive to external
stimuli, and some can walk or carry on relatively complex
activities without interruption, but usually with slowness and
diminished insight and complexity of activities (164,172).
There is often a fluctuating diminution of responsiveness or
overt confusion. Language use may be relatively preserved, but
with slowness, paucity of speech, and monosyllabic answers
(156), and often an inability to discuss more complex con-
cepts. Language may be reduced to simple phrases or less, and
patients may follow only the simplest of instructions, or none.
Infrequently, absence SE progresses to complete unrespon-
siveness (164).
Motor abnormalities are minimal. There may be brief spo-
radic myoclonic eyelid jerks and eye blinking, especially in the
IGE syndromes (146). Limb myoclonus is common. Gestural and
ambulatory automatisms occur (166,173), but automatisms are
usually less complex than in CPSE. Gait tends to be preserved if
the patient can stand. Sometimes there are bradykinesia, delayed
reactions, and withdrawal (166). Patients may appear to have
Tabl e 2 9 . 3
Status Epilepticus in the Idiopathic Generalized
Epilepsies
Generalized convulsive tonic–clonic SE, especially upon
awakening
Myoclonic SE
Nonconvulsive forms:
• Typical absence SE
• Atypical absence SE
• “De novo” absence SE
• Autonomic SE (perhaps better characterized as regional)
Adapted from Shorvon and Walker (167).
614 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Fp1-F3

F3-C3
C3-P3
P3-O1

Fp2-F4
F4-C4
C4-P4

P4-O2

Fp1-F7
F7-T3
T3-T5

T5-O1
Fp2-F8
F8-T4

T4-T6
T6-O2
Fz-Cz

Cz-Pz
LUC
RLC

EKG
100 uV1 sec

Figure 29.12 Absence status epilepticus with generalized, frontally predominant, 3-Hz spike-and slow-wave discharges.

diminished initiative or motivation. Perseveration in motor
activity or on performance tests is typical.
Patients’ experiential descriptions are vivid. There may be
visual hallucinations, a dreamy state, and inappropriate man-
ner. Some patients are aware, themselves, of slowing or diffi-
culty with thought and speech, sometimes noticeable only to
those closest to them. Occasionally, patients show hostility and
aggression, agitation, and impulsive behavior (13,172).
Behavior mistaken for psychosis is more likely with CPSE, and
confabulation was seen in CPSE only (146). Absence SE often
induces relatively little amnesia or postictal confusion (172).
Absence SE may last anywhere from 30 minutes (by defini-
tion) to months (24). It may cease abruptly—spontaneously
or upon treatment with BDZs, with rapid return to normal
mentation. Some episodes end with a generalized convulsion.
Most reported cases have been isolated, but some recur fre-
quently (172).
Recently described is a syndrome in which episodes of
absence SE are the primary manifestation of epilepsy, with few
other seizures (169). All patients (adults) had normal neuro-
logic examinations and imaging and no family history of
epilepsy. Episodes of absence SE lasted from 30 minutes up to a
few weeks, and many recurred several times a year.
Clinically, absence SE may appear remarkably similar to CPSE
of frontal or temporal origin; and all of these can be difficult to
differentiate from a postictal confusion; from nonepileptic causes,
including psychiatric conditions such as dissociative states,
depression, or hysterical behavior (166,172); or from posttrau-
matic amnesia or metabolic encephalopathies (24). Medication
side effects can appear very similar to absence SE clinically.
Myoclonus is not rare in absence SE, but it is rare in CPSE (146).
EEG
EEG is often required for the accurate diagnosis of absence SE.
The characteristic EEG of absence SE is very similar to those of
absence epilepsy: frequently recurrent or prolonged uninter-
rupted seizures for over 30 minutes, with extremely rhythmic,
bilaterally symmetric runs of spike-and-slow-wave discharges;
polyspikes may be admixed (Figs. 29.12 and 29.13) (174,175).
Other EEG patterns include generalized rhythmic slowing with
intermixed spike-and-slow-wave complexes, irregular sharp-
and-slow-wave discharges, or diffuse background slowing with
superimposed bursts of fast activity (172). The epileptiform
discharges are bilaterally synchronous and usually of maximal
voltage over frontal and central regions. Characteristic-appearing
generalized epileptiform discharges may have some asymmetry,
but this should not be overinterpreted (138).
Discharge frequencies may reach up to 4 Hz at the onset of
the episode but typically slow quickly to the usual 3-Hz pattern
within seconds. Discharges are characteristically regular, at 3 to
3.5 Hz. If seizures are prolonged, the discharges may slow to
less than 3 Hz and may become irregular (176). More rapid dis-
charges and polyspikes suggest other IGE syndromes. Discharges
are occasionally interrupted by brief bursts of normal EEG
 Chapter 29 ■ Nonconvulsive Status Epilepticus 615

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
200 uV
2 sec

Figure 29.13 Absence SE in a 17-year-old boy with juvenile absence epilepsy. Brief absence seizures (during an episode
of SE) demonstrate generalized frontally maximal spike-wave activity at 3 Hz. Frequent 5- to –10-second absence seizures
recur every 5 to 10 seconds over 20 minutes until administration of IV lorazepam. Clinical manifestations were staring,
unresponsiveness, eyelid fluttering, and subtle limb automatisms.

background activity. Interictal EEGs may show the same dis-
charges in brief bursts, almost always on a normal background.
Treatment and Outcome
Typical absence SE that occurs within absence epilepsy or other
IGE syndromes is almost always readily interrupted (clinically
and on EEG) by relatively low doses of BDZs or other AEDs,
preferably intravenously, but sometimes effective even orally
(Fig. 29.14). Interruption by intravenous BDZs is so often effec-
tive that it is a useful diagnostic test. Lorazepam doses under 0.1
mg/kg will often suffice (45). Intravenous valproic acid at 20
mg/kg is a reasonable alternative (177) and is also effective in
patients with recurrent absence SE, particularly in the IGE syn-
dromes (178). Lamotrigine is also helpful. Typically, the EEG
will resume a normal background quickly, and the patient may
emerge from the confusion within minutes. Phenytoin and car-
bamazepine are not ideal treatments as they may exacerbate or
even precipitate absence SE (171).
For typical absence SE, the outcome and consequences
depend on the etiology—in this case, the epilepsy syndrome,
which is benign. Absence SE has produced no mortality and
minimal morbidity in reported series, and no significant long-
term morbidity can be attributed to it (29,172) or to any other
NCSE in IGE syndromes. Nevertheless, this conclusion rests on
clinical impressions in moderate-sized series rather than on
confirmation by neuropsychologic testing. Still, the prognosis
of absence SE is superb.
De novo Absence Status Epilepticus
Typical absence SE is unlikely in older patients without an ear-
lier history of an IGE, but occasional episodes have been
reported in patients older than 50 years with no prior epilepsy
(de novo absence SE of late onset) (29). Most have moderate
impairment of consciousness. The most common setting is that
withdrawal of BDZs used for sleep or psychiatric reasons. Other
triggers include toxins or metabolic dysfunction and the use of
neuroleptic and other psychotropic medications (179). Some
“de novo” patients may have had primary generalized epilepsies
earlier in life without diagnosis, and others may have been mis-
diagnosed. It is unclear that all of these cases are truly “de novo.”
Fortunately, episodes of de novo SE are relatively uncommon
and usually easy to interrupt quickly (167). Especially when
precipitated by some identifiable cause such as BDZ with-
drawal, most patients do not need maintenance AEDs.
The EEG in de novo absence SE tends to be characteristic of
absence SE, showing generalized spike-and-wave activity at
616 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

A
FP1-F3

F3-C3

C3-P3

P3-O1

FP2-F4

F4-C4

C4-P4

P4-O2

FP1-F7

F7-T3

T3-T5

T5-O1

FP2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
75 uV 1 sec

B
FP1-F3

F3-C3

C3-P3

P3-O1

FP2-F4

F4-C4

C4-P4

P4-O2

FP1-F7

F7-T3

T3-T5

T5-O1

FP2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
75 uV 1 sec

Figure 29.14 A: A 75-year-old woman with a history of generalized tonic–clonic seizures (probably a long-standing IGE),
now with absence status epilepticus causing confusion. EEG shows semirhythmic high-voltage 3-Hz sharp-and-slow-wave
discharges, alternating with periods of diffuse delta activity and slower epileptiform discharges. B: Same patient with com-
plete cessation of ictal activity after IV lorazepam, likely causing the widespread faster beta activity.
 Chapter 29 ■ Nonconvulsive Status Epilepticus 617

FP1-F3

F3-C3

C3-P3

P3-O1

FP2-F4

F4-C4

C4-P4

P4-O2

FP1-F7

F7-T3

T3-T5

T5-O1

FP2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz
75 uV 1 sec

Figure 29.15 A 58-year-old woman with no history of epilepsy, admitted with confusion. EEG shows generalized 2.5-Hz
spike-and slow-wave activity. A few minutes after 2 mg IV lorazapam, the EEG showed mild diffuse slowing with plenti-
ful beta activity. The confusion improved concurrently with the EEG improvement, about 4 minutes after the lorazepam
administration.

frequencies from 0.5 to 4 Hz (180) (Fig. 29.15). The spike-and-
wave component can be less prominent and more irregular
than in typical absence SE but it still shows a similar frontal or
centrally maximal distribution (29).
Atypical Absence Status Epilepticus
Atypical absence status epilepticus (AASE) is a relatively rare
type of SE that occurs primarily in children with cryptogenic
and secondarily generalized epilepsy, such as in the LGS with
substantial neurologic deficits, developmental delay, and severe
encephalopathies (70,181,182). These patients often have
mixed epilepsies, with multiple seizure types and episodes of
tonic, atonic, and myoclonic seizures (Fig. 29.16). The inci-
dence of atypical absence SE is anywhere from 100 to 200
episodes per million per year (144).
Atypical absence seizures resemble typical absence seizures
in their occurrence primarily in young patients, with general-
ized spike-and-slow-wave discharges, and with the same favor-
able (and unfavorable) responses to medications, for example,
exacerbation with AEDs such as carbamazepine, and treatment
with ethosuximide and BDZs (183). Nevertheless, AASE is
more complicated than typical absence SE. It presents at an ear-
lier age, typically before adolescence, with significant back-
ground encephalopathies, both clinically and on EEG (181).
The clinical manifestation of AASE is often a further slowing
or reduction in cognitive activity—from a poor baseline.
Patients are more confused or less responsive than before. They
may become mute or immobile, but some children are able to
follow simple commands (181). Consciousness may fluctuate
and become markedly impaired. Patients exhibit poor control
of movements and frequent falls (pseudoataxia) and cognitive
difficulties with near stupor (pseudodementia) (27). AASE can
go on for days, without a precise onset or ending. There is typ-
ically a gradual onset and resolution of altered behavior in pro-
longed episodes, but automatisms and other motor activities
(such as walking) may continue, even while responsiveness is
reduced (184). Often, there are no overt clinical seizures, and it
can be very difficult to distinguish AASE from the encephalo-
pathic, developmental, or psychiatric behavioral baseline of the
(often institutionalized) patients.
LGS is the most common setting for AASE. With LGS and
“myoclonic-astatic” seizures, there may be myoclonus or stupor
(182). When the AASE of LGS continues into adulthood, psy-
chiatric features may predominate, including “regressive behav-
ior” or stubbornness; episodes of obtundation may point to
NCSE (28). Other etiologies include several genetic syndromes
associated with mental retardation (181). Angelman syndrome
can cause atypical absence seizures, with high-voltage bursts of
618 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz
Cz-Pz

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz
Cz-Pz

Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fz-Cz 300 uV
Cz-Pz
2 sec

Figure 29.16 Atonic status epilepticus in a 12-year-old girl with new-onset epilepsy. Generalized high-voltage slow waves
with irregular generalized spikes correlated with atonic episodes of the limbs and trunk. Ninety seconds of EEG show
recurrent atonic seizures; a similar pattern persisted for more than 30 minutes.

slowing on the EEG, evolving to spike and wave discharges later
in childhood (185).
Diagnosis of AASE is difficult, primarily because most patients
have severe underlying encephalopathies and neurologic deficits,
along with a markedly abnormal mental status, even when not in
SE. AASE includes impaired responsiveness, prominent epilepti-
form discharges, and electrographic seizures. To qualify as SE,
there should be a clear change from baseline, both clinically and
electrographically, but because patients often progress in and out
of AASE, electrographic seizures may not correlate well with clin-
ical signs, and it is unclear what role the epileptiform discharges
play in the clinical syndrome. It could thus be considered another
“boundary syndrome.” Radiologic scans are often unremarkable,
but some show evidence of cortical dysplasia or hematomas (181).
Episodes of AASE can recur frequently and remain refractory
to treatment. Intravenous BDZs may interrupt the EEG dis-
charges without changing behavior substantially, but BDZs may
also precipitate tonic seizures or tonic SE in these patients (186).
The underlying mental retardation and the difficulty of differen-
tiating AASE from an interictal state has often led to less vigorous
treatment. When episodes end, the baseline neurologic func-
tion is often still significantly impaired. Occasionally, convulsions
herald or terminate AASE. Long-term morbidity is essentially
always considered due to the underlying neurologic condition.
EEG
The EEG of AASE is characterized by generalized spike or poly-
spike and slow-wave discharges with a dominant frequency
below 3 Hz (slow-spike-and-wave), with somewhat less rhyth-
micity and less perfect symmetry than seen in absence seizures
(Figs. 29.17 and 29.18) (184). Discharge frequencies usually
range from 1 to 2.5 Hz and are often irregular, usually superim-
posed on a slow background. Interictally, the background is
usually quite abnormal, including generalized or focal slowing
with generalized spike and sharp waves, in isolation or in brief
bursts, and admixed low-voltage generalized paroxysmal alpha
or beta frequency activity (70,182).
In a model of AASE with the same response to AEDs as in
human AASE, there was a gradual onset of reduced movement,
and epileptiform discharges were slower (183). Whereas in typi-
cal absence seizures the epileptiform spike and wave discharges
appear more restricted to thalamocortical circuitry, in experi-
mental models of atypical absence, epileptiform discharges also
involve the hippocampus and other deep limbic structures (187).
 Chapter 29 ■ Nonconvulsive Status Epilepticus 619

Figure 29.17 Atypical absence status epilepticus (AASE) in a 49-year-old man with Lennox–Gastaut syndrome who pre-
sented with drooling and lethargy. Interictal EEG shows slow-spike-and-wave-activity at 1 Hz. EEG during AASE showed
2- to 2.5-Hz generalized spike-and-wave activity.

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2

Fz-Cz

Cz-Pz

EKG
100 uV1 sec

Figure 29.18 Atypical absence status in a 26-year-old woman with tuberous sclerosis and Lennox–Gastaut syndrome.
Clinically, there were frequent episodes of staring and blinking lasting 20 to 30 seconds followed by confusion.
620 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Correspondingly, clinical AASE in humans appears to involve
more of deeper structures, correlating with the more persistent
memory and other cognitive problems in AASE, as opposed to an
apparent lack of residua following typical absence SE. The epilep-
tiform discharges of AASE also appear more likely to engender
(pathologic) hippocampal synchronization than that occurs in
typical absence subjects, perhaps because of earlier cerebral dam-
age that may have occurred in patients with LGS or other causes
of severe childhood epilepsies (187).
Secondarily Generalized Nonconvulsive
(Electrographic) Status Epilepticus
Most generalized NCSE in adults without IGE syndromes are not
truly absence SE, but rather secondary from a focus (16,49,139),
even though the EEG and clinical signs are often generalized at the
time of diagnosis and management. Although secondarily gener-
alized, these episodes of SE may appear very similar clinically (and
on EEG) to cases of primary generalized NCSE and it may be
impossible to tell the difference at the time of presentation.
Secondarily generalized NCSE is the true diagnosis of most cases
described earlier in the literature as “absence SE,” but they have lit-
tle to do with typical absence epilepsy. In the IGE syndromes,
absence SE is relatively uncommon, typically easy to treat, and has
a superb prognosis. Secondarily generalized NCSE usually has
underlying (and sometimes severe) lesions, is much harder to
treat, and has the prognosis of the underlying illness. Secondarily
generalized NCSE is the most common type of SE in ICUs.
EEG
Often, generalized NCSE cases cannot be separated easily into
those of a primarily generalized nature versus those with focal
onset spreading to a broader or even generalized distribution.
[Experimentally, generalized epileptiform discharges can result
from focal electrical stimulation in frontal regions (188)]. Rarely
does one capture the onset of SE on EEG except during continu-
ous monitoring, and an EEG begun after the clinical onset of
NCSE may not allow distinction between focal and generalized
SE. Nevertheless, the EEG can still be helpful in deducing whether
there was a focal onset to the seizures. Focal epileptiform dis-
charges, especially with a consistent location, help point to a focal
onset. Some episodes of apparent absence SE include prominent
focal features on the EEG and clinically (138), possibly because the
generalized epileptiform discharges, regular as they appear, are
secondarily generalized (Figs. 29.19 and 29.20). Some of these
EEGs include discharges with more irregularity or asymmetry
than are found on EEGs of typical absence SE. Also, some appar-
ently generalized epileptiform activity in NCSE evolves toward
unilateral discharges with treatment, suggesting a focal onset (45).
Finally, some seizures have minimal or no clinical or EEG features
suggesting a focal onset but are associated with known unilateral
structural lesions and are very likely to have focal origins.
In one review of EEG patterns during NCSE, there were three
main groups: (1) epileptiform discharges generalized at the
onset; (2) discharges that begin focally, with or without second-
ary generalization; and (3) patterns with both focal and general-
ized features or otherwise poorly classified (45). In that series of
85 episodes of NCSE in 78 patients, 69% showed a generalized
pattern, 13% a focal pattern, and 18% a generalized pattern with
focal predominance. Among pediatric ICU patients, focal or
multifocal NCSE (65%) was more common than generalized SE
(35%), often reflecting underlying acute focal brain lesions (189).
This ongoing electrographic (nonconvulsive, and usually sec-
ondarily generalized) seizure activity is very common among
critically ill ICU patients. Many have had earlier generalized
convulsions or GCSE, often in the setting of serious medical ill-
ness with cerebrovascular disease, sepsis, hypoxia, or severe
metabolic derangements. Those with minimal movement (typi-
cally nystagmus or myoclonus) after GCSE have been called
“subtle” SE (190). Some are referred to as electrographic status
epilepticus (ESE) (Fig. 29.21) (33). Others refer to “status epilep-
ticus in coma,” but not all patients with ongoing electrographic
seizures are comatose. Some label these patients with severe
medical illness as having “epileptic encephalopathies,” indicating
that the underlying disease causing the discharges is key, and
that the epileptic component is not primary and may not
respond to AEDs. This term, however, is probably best reserved
for childhood conditions such as those associated with ESES
(see section on Early Life, Age-related Nonconvulsive Status
Epilepticus Syndromes).
The term “subtle” SE often describes the later stages of
GCSE, in which initially overt motor seizures become subtle
gradually, often in comatose patients (Chapter 28). The EEG
may show discontinuous ictal activity with brief bursts of gen-
eralized polyspikes or generalized periodic discharges (33,190).
Some consider these EEG patterns to indicate ongoing SE only
if the patient had previous clinical seizures or SE (32). Many
consider this NCSE as a natural progression from GCSE (190).
Most agree that the urgency of treatment continues during the
persistent electrographic seizures, at least when preceded by
generalized convulsions. Nevertheless, abolition of electro-
graphic seizures may not lead to a clinical improvement, and
these patients have high morbidity and mortality (32,190–192).
Despite the prominent epileptiform activity on EEG, the diag-
nosis of NCSE is not always clear (see sections below on EEG
Diagnosis of Nonconvulsive Status Epilepticus, and Controversial
EEG Patterns). In clinical practice, however, ESE (rhythmic, rela-
tively rapid epileptiform discharges, typical for SE on the EEG)
without obvious clinical manifestations should be considered
SE rather than simply an encephalopathy with epileptiform dis-
charges—even when treatment is unsuccessful in effecting a clin-
ical improvement. The EEGs are very similar to those in many
published NCSE studies. Also, patients with ESE upon emergence
from pentobarbital or midazolam infusion treatment usually go
on to have clinically evident seizures (193,194), indicating that
ESE is not just an EEG aberration. Finally, some patients with
these EEGs do respond well to AEDs (195).
NONCONVULSIVE STATUS EPILEPTICUS
IN THE INTENSIVE CARE UNIT
The diagnosis of NCSE in acutely ill patients with abnormal
mental status is often complicated and difficult, and many other
conditions mimic NCSE. Among the many reasons for altered
behavior and impaired consciousness are severe toxic, metabolic,
and infectious encephalopathies; sedating medications; postictal
or postoperative encephalopathies (many due to medications);
 Chapter 29 ■ Nonconvulsive Status Epilepticus 621

Figure 29.19 A: A 31-year-old woman with metastatic colon cancer, admitted with catatonia. EEG initially showed rhyth-
mic 1.5-Hz delta activity over the right hemisphere. B: Several hours later, the EEG shows generalized sharp-and slow-wave
discharges at 1.5 Hz, mimicking generalized status epilepticus. (Ten minutes after 4 mg IV lorazepam, the EEG showed
resolution of all rhythmic and sharp activity, and the patient became more responsive clinically.)

BDZ withdrawal states or toxicity; neuroleptic malignant and
serotonin syndromes; and other systemic illnesses. Several of
these conditions are accompanied by abnormal EEGs, sometimes
even with epileptiform features such as sharp waves.
There are also patients treated for SE with the wrong diag-
nosis! When clinical manifestations are atypical and appear
refractory to first-line AEDs, some diagnoses of SE should be
questioned and should be confirmed by EEG prior to more
aggressive treatment. It is sometimes necessary to exclude psy-
chogenic nonepileptic SE (a conversion disorder), which does
not usually respond to AEDs (196,197). Pseudostatus can mimic
NCSE, with prolonged staring, blinking, and unresponsiveness,
and even expert clinicians may be unable to distinguish it from
SE. Often, psychogenic SE can be diagnosed definitively by EEG
622 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Figure 29.20 A: A 46-year-old man with a right frontal cavernous malformation, admitted with bizarre behavior. Ictal
EEG during an episode of status epilepticus shows initial bilateral frontal (right left) polyspike and wave activity. B:
Within 10 seconds of onset, the EEG evolves to show generalized symmetric polyspike and wave activity at 4 Hz, sugges-
tive of primary generalized epilepsy. He had 32 seizures in a 4-hour period.

alone. The correct diagnosis can prevent inappropriate, aggressive
treatment with IV AEDs, general anesthesia, and intubation—
and their complications (198,199).
In pseudostatus, the EEG background activity is usually nor-
mal (Fig. 29.22), but muscle and movement artifact may obscure
the EEG (197). Following true epileptic seizures, postictal slow-
ing is the rule, and the presence of a normal alpha rhythm
immediately following apparent convulsions (or during brief
pauses in motor activity) provides strong evidence for pseu-
dostatus (Fig. 29.23).
 Chapter 29 ■ Nonconvulsive Status Epilepticus 623

Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2

EogL-A1

EogR-A2

EKG
50 uV 1 sec

B
Fp1-F3

F3-C3

C3-P3

P3-O1

Fp2-F4

F4-C4

C4-P4

P4-O2

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp2-F8

F8-T4

T4-T6

T6-O2

EogL-A1

EogR-A2

EKG
75 uV 1 sec

Figure 29.21 A: Electrographic status epilepticus in a 50-year-old man with multiorgan dysfunction after anoxic insult.
The patient was comatose, but there were no other clinical manifestations. B: Same patient after treatment with IV leve-
tiracetam, now showing subtle generalized electrographic status epilepticus.
624 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 29.22 Nonepileptic psychogenic status epilepticus. A 51-year-old woman with history of epilepsy and left tempo-
ral lobectomy, admitted with more than 30 episodes per hour characterized by speech arrest, unresponsiveness, and closed
eyelids during the episode. EEG shows left hemisphere breach artifact from temporal lobe surgery, but no ictal changes.
On the other hand, ICU patients with impaired responsive-
ness are not rarely in NCSE (98,140,200,201), especially when
they have had earlier seizures or clinically evident SE (33,49,
195,202). Of 164 patients who had EEGs after apparent control
of clinical SE in one series, 42% had continued seizure dis-
charges, and 14% were considered in NCSE (98). Similarly, 8% of
patients with coma of all causes and similar monitoring (without
clinical signs of seizures) had NCSE (200). With C-EEG monitor-
ing for the question of seizures, or to evaluate coma, another
group found that 19% of patients monitored in a neurologic ICU
over a 6-year period had seizures recorded, and 92% of these
were strictly nonconvulsive (203). Coma, prior epilepsy, age below
18 years, and earlier convulsive seizures were risks for nonconvul-
sive seizures.
There are many causes of NCSE in the ICU. Patients with
acute neurologic injuries are at particular risk, especially those
with intracerebral or subarachnoid hemorrhage, CNS infection,
brain tumors, severe head trauma, or following neurosurgery—
in each of which 20% to 30% of monitored patients have had
nonconvulsive seizures in different studies (201,203–206). With
C-EEG monitoring of patients with intracerebral hemorrhage,
about one-third had seizures (although patients were selected in
part due to prior seizures or a suspicion of seizures) (206). Half
of seizures detected were purely nonconvulsive.
Coma is a particular risk for NCSE in the ICU, especially if
the cause is unclear. In one study, more than half of 97 coma-
tose patients undergoing C-EEG had electrographic seizures
(203). Other risks include recent generalized convulsions, a
history of epilepsy, and remote symptomatic CNS insults (e.g.,
stroke, neurosurgical intervention, tumor, etc.), especially when
the patient is encephalopathic (21). Other risks for NCSE are
intercurrent electrolyte abnormalities, infections such as pneu-
monia or urinary tract infections, or glucose dysregulation.
The patients at highest risk for unrecognized NCSE are those
who were in GCSE and were thought to have been treated suc-
cessfully—but were not! A prolonged “postictal” state after a
convulsion should raise concern for the possibility of ongoing
NCSE, and EEG is usually the only way to make the diagnosis
(33,49). After generalized convulsions, most patients begin to
recover responsiveness within 20 to 30 minutes, although there
is a broad range. All patients with seizures or SE who do not
return to a normal level of consciousness should have EEGs to
see if treatment was adequate or the patient is still seizing.
One group determined that nonconvulsive seizures and
NCSE should be suspected in the ICU when (1) a prolonged
encephalopathy follows generalized convulsions, an operation,
or a neurologic insult; (2) there is acutely impaired or fluctuat-
ing consciousness interrupted by normal alertness; (3) there is
impaired consciousness with facial myoclonus or nystagmus;
(4) staring, aphasia, or automatisms (e.g., limb or facial) occur
episodically; or (5) other acutely altered behavior has no obvi-
ous cause (191).
How long should C-EEG be continued in the search for non-
convulsive seizures and NCSE in patients with reduced respon-
siveness? In one study, a routine 30-minute EEG identified
electrographic seizures in 11% of 105 critically ill patients, while

Figure 29.23 Psychogenic generalized status epilepticus. This 22-year-old woman had prolonged episodes of generalized
on–off tremulous limb movements. The EEG shows repetitive EMG artifact, with normal background EEG activity evident
in the areas between the tremor artifact.
the subsequent C-EEG (for a median of 2.9 days) found seizures
in 27% (207). In another series, only half of 110 patients with
seizures seen eventually on EEG had them during the first hour
of recording (208). In that series, 95% of noncomatose patients
had their seizures found in the first 24 hours, versus 80% for
comatose patients. After 48 hours, this increased to 98% and
87%, respectively. Among children with seizures, half had the
first seizure detected in the first hour of EEG recording, and 80%
within 24 hours (209). The authors concluded that C-EEG mon-
itoring was appropriate for 24 hours in patients without coma
and for 48 hours in comatose patients, in those with periodic
discharges, or when AEDs are being withdrawn (209).
EEG DIAGNOSIS OF NONCONVULSIVE
STATUS EPILEPTICUS
Some unresponsive patients in ICUs (and elsewhere) are having
seizures or NCSE, and others are not. It is often impossible to tell
the difference solely by clinical examination. EEG is far superior
to any other technique in detecting nonconvulsive seizures and is
crucial in the assessment of unresponsive patients, particularly
those who have had earlier convulsions. Nevertheless, seizures
(including nonconvulsive) can be intermittent, and an individual
Chapter 29 ■ Nonconvulsive Status Epilepticus 625
EEG may miss them, so many modern ICUs utilize C-EEG to
detect seizures and other changes in neurophysiologic function.
The diagnosis of seizures and NCSE on an EEG relies on the
determination by a clinical neurophysiologist that certain pat-
terns are “ictal” in nature, that is, having to do with epileptic
seizures, but controversy exists regarding which EEG patterns are
diagnostic of, or consistent with, NCSE. Sometimes, seizures and
NCSE are readily apparent on the EEG, with characteristic or
even “classic” features. Other EEGs show disorganized slowing,
without rhythmicity or epileptiform abnormalities, more sugges-
tive of encephalopathies than seizures. Very similar EEG findings
can be seen in NCSE and in metabolic encephalopathies, and
many EEG patterns are neither pathognomonic nor easy to inter-
pret. Even after the diagnosis of NCSE, the clinical significance of
the epileptiform discharges can be unclear—that is, the electro-
graphic seizure activity may or may not be the cause of the clini-
cal abnormalities.
Also, the orderly temporal progression of EEG patterns
described for GCSE (Chapter 28) usually does not occur in NCSE
(32,210). In one study of 40 patients, 53% had discrete seizures
alone, without any other ictal patterns; 28% had discrete seizures
with periodic epileptiform discharges (PEDs) between seizures;
and 8% had continuous ictal activity (210). Occurrence of
626 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 29.24 Various EEG waveform morphologies seen in NCSE.
continuous seizures, discrete seizures, or PEDs could occur at any
time after SE, and no predictable pattern of evolution was iden-
tified. EEG evolution probably depends on the etiology and type
of SE, baseline condition of the patient, duration of seizure activ-
ity, and type and duration of treatment. The variability in evolu-
tion adds to the controversy regarding which patterns are “ictal.”
Seizures captured in their entirety typically show progression
from low-voltage, high-frequency spikes to high-voltage, low-fre-
quency spike-and-slow wave activity, before stopping abruptly
and being replaced by background slowing or suppression (Fig.
29.9). Usual morphologic features include typical rhythmic, gen-
eralized, symmetric spike-and-waves or polyspikes and waves at
2 to 3.5 Hz; atypical spike and wave with lower frequency and
less symmetry; multiple spike-and-wave (repetitive complexes of
two or more spikes followed by a slow wave); and high-voltage,
repetitive, rhythmic, focal or generalized delta activity with inter-
mixed spikes, sharp waves, or sharp components (Fig. 29.24)
(45). Diagnosis is more difficult when the seizure (or SE) pre-
cedes the beginning of the tracing and continues beyond its end.
In such cases, rhythmic sharp features, typically faster than 1 Hz,
may be seen, often with variability.
There is a lack of consensus on the terms and descriptors for
electrographic seizures and periodic and other abnormal EEG
Tabl e 2 9 . 4
Tentative Research Criteria for Nonconvulsive
Seizures [and if 30 minutes, Nonconvulsive
Status Epilepticus]
1. Repetitive generalized or focal spikes, polyspikes, sharp
waves, spike-and-wave or sharp-and-slow wave complexes,
or other rhythmic waveforms at 2.5/sec, lasting longer
than 10 seconds.
2. Same waveforms as above, with discharges 2.5/sec, with
clear clinical ictal phenomena, such as facial twitching,
nystagmus, or limb myoclonus.
3. Same waveforms as above, with discharges 2.5/sec, with
an unequivocal evolution of the rhythmic pattern,
including increase or decrease in frequency (by 1 Hz)
and change in discharge morphology or in location
(gradual spread of rhythmic activity into or out of a region
involving at least two electrodes). Changes in discharge
amplitude or “sharpness” alone are not sufficient.
4. Rhythmic theta or delta waves at 1/sec, with additional
criterion of unequivocal clinical improvement or
improvement on EEG (such as resolution of epileptiform
discharges and reappearance of previously absent normal
background rhythms and reactivity) or both, following
quickly after acute administration of rapidly acting AEDs,
typically benzodiazepines. (Resolution of discharges
leaving a slow background alone, without clinical
improvement, would not suffice.)
patterns, and on electrographic definitions of NCSE. A commit-
tee of the American Clinical Neurophysiology Society (ACNS)
has recommended standardization of terminology for periodic
and potentially epileptiform EEG findings in order to facilitate
multicenter clinical research (211). They recommend that EEG
findings be described by morphology, location, frequency, and
periodicity of waveforms—without using clinical connotations,
but with subsequent interpretation in the clinical context. The
committee also suggested tentative criteria for the EEG diagno-
sis of seizures and, if persistent enough, NCSE, to be tested
empirically, similar to those in Table 29.4 (72,191,204).
Seizures are generally indicated by (1) repetitive focal or gen-
eralized epileptiform activity or rhythmic waveforms faster than
2.5 Hz, lasting longer than 10 seconds; or (2) similar discharges
slower than 2.5 Hz with (a) unequivocal evolution in frequency,
morphology, or field; (b) clinical motor manifestations (e.g.,
limb or facial twitching); or (c) clear clinical and EEG improve-
ment after administration of rapid-acting intravenous AEDs,
typically BDZs. “Evolution” of the electrographic discharges of
seizures on the EEG over time includes an increase or decrease
in frequency by at least 1 Hz, significant change in waveform
morphology, or spread to at least two adjacent EEG electrodes
(72,191) (Fig. 29.9). Diagnostic criteria for invariant patterns
and discharges slower than 2.5 Hz are more controversial (212).
 Chapter 29 ■ Nonconvulsive Status Epilepticus 627

Figure 29.25 Generalized periodic epileptiform discharges (GPEDs) in a 62-year-old woman with intractable epilepsy.
Variable GPED repetition rate up to 3 Hz, with intervening periods including rhythmic fast activity and variable morphol-
ogy, which may represent ictal activity. After treatment with IV lorazepam, the background showed mild diffuse slowing
but no further ictal activity. The patient became less confused and was able to answer questions.

Slower discharges are more likely to signify epileptic seizures if
they evolve in frequency or follow clinical SE.
Straightforward examples of NCSE on EEG conform to the
patterns described above and continue for over 30 minutes.
Whatever the generalized or focal pattern, if the patient returns
rapidly to a baseline state (after AED treatment or sponta-
neously) and the EEG becomes normal, then this very likely was
NCSE (Fig. 29.25). Resolution of the epileptiform discharges on
EEG and resumption of a posterior alpha frequency back-
ground suggestive of wakefulness are also considered diagnos-
tic of an earlier seizure.
These criteria may provide for greater certainty in the diagno-
sis of electrographic seizures and NCSE, but they were designed
as guides for research purposes, and some patients whose EEGs
do not meet these strict criteria may also be having seizures or be
in NCSE (204). For example, in well-established clinical cases of
NCSE, stereotypic spike-and-wave activity at 3 Hz was seen in
only 7% of patients; discharge frequency was usually 1 to 2.5 Hz
(45). Similar rhythmic and sharp EEG patterns are still strongly
suggestive of seizure activity in the appropriate clinical setting,
especially following earlier generalized convulsions or GCSE,
or with severe medical illness. Many are associated with mini-
mal or no motor convulsive activity, as in the ESE described
above in the section on Secondarily Generalized Nonconvulsive
(Electrographic) Status Epilepticus.
Some epileptiform EEG findings generally agreed upon to
not represent seizures include midtemporal theta of drowsiness,
rhythmic midtemporal theta activity, subclinical rhythmic EEG
discharges of adults (SREDA), and wicket spikes (204) (see
also Chapter 14, on normal variants). Other patterns generally
not considered to represent NCSE include periodic discharges
in patients with acute cerebral injuries but without prior
clinical seizures, and discharges in patients with epileptic
encephalopathies in which the periodic discharges appear sim-
ilar to those on the baseline EEG (72,213).
Frequently, the EEG determination of seizures is not a
blinded exercise. The interpreter may be aware of the patient’s
staring, facial twitching, or subtle myoclonus, and be predis-
posed to diagnose ambiguous patterns as NCSE. For example,
the appearance of periodic discharges at 1 Hz in a patient with
focal twitching would favor a diagnosis of seizure over that of
an interictal periodic pattern (43,46). Often, EEG determina-
tion of seizures needs re-interpretation in the clinical setting.
CONTROVERSIAL EEG PATTERNS
Periodic Discharges
Sharp and rhythmic EEG features vary remarkably and range
from the nonictal “irritative” (post-, inter-, or preictal) to the
“actively seizing”—along an “ictal–interictal continuum”
(43,46,72,214). They include PEDs with various repetition rates
and morphologies, including PLEDs and bilateral independent
periodic lateralized epileptiform discharges (BiPLEDs) (see
Chapter 28), and generalized periodic epileptiform discharges
628 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 29.26 EEG of a 72-year-old man following cardiac arrest. On weaning of propofol, continuous electrographic
seizures began, with repetitive spike-and-slow-wave discharges at 2.5 to 3 Hz.
(GPEDs). Periodic discharges are more likely to be “ictal” when
they are rapid ( 2 Hz) (Fig. 29.26), or show clear evolution in
frequency (increase or decrease by at least 1 Hz), morphology, or
field (72,204). Periodic discharges with intervening fast and
admixed sharp components (e.g., “PLEDs-plus”) are also more
likely to be “ictal” than simple isolated discharges (see also
Chapter 28, Fig. 29.9A,B) (215). The presence of PEDs should
prompt consideration of C-EEG monitoring to look for electro-
graphic seizures arising out of that background.
GPEDs are continuous generalized spikes, polyspikes, sharp-
and-slow waves, or TWs, often with a repetition rate of about 1
Hz, typically arising from a diffusely slow or attenuated back-
ground (Figs. 29.25 through 29.28) (213). They often occur in
obtunded patients, typically after anoxia or other catastrophes,
metabolic insults, recent overt seizures, or in the late stages of
GCSE (190). GPEDs are highly associated with clinical seizures
but not necessarily indicative of seizures proceeding at the same
time (210,216). Some consider GPEDs to be reliably ictal and
recommend aggressive AED therapy (217), whereas others
believe that they are the sign of neuronal injury, but not actual
seizures and not requiring aggressive treatment (Figs. 29.27 and
29.28) (216,218). One study did not find any features that could
distinguish clearly between GPEDs after anoxia and GPEDs
after SE (219), even though those conditions have markedly dif-
ferent implications for treatment. GPEDs tend to persist even
with aggressive therapy, and it is not known whether patients
benefit from treatment of these EEG findings.
Focal or generalized periodic and quasi-periodic discharges
can also be elicited in stuporous or comatose patients upon
stimulation—“stimulus-induced rhythmic, periodic, or ictal
discharges” or SIRPIDs (Fig. 29.29) (220). SIRPIDs usually abate
after the stimulus recedes. Whether they represent seizures or an
abnormal arousal process is often unclear. Some are definitely
seizures (221). When elicited along with facial or limb jerking
(as in a minority of cases), they are more likely to be ictal phe-
nomena. Many appear to lie along the “ictal–interictal contin-
uum,” especially when there is no clinical motor correlate.
Triphasic Waves
Another controversial area in EEG diagnosis is that of triphasic
waves (TW)—bursts of moderate- to high-voltage (100 to 300
V) rhythmic complexes, often recurring at 1 to 2 Hz (Fig.
29.30A,B). TWs are named for their multiphasic morphology,
consisting of an initial blunted, low-voltage negative spike-like
phase; a second phase consisting of a slowly rising high-voltage
positive component; and third phase of a lower voltage, broad
negative slow wave. TWs are broadly distributed, usually with
a frontal voltage maximum. They frequently occur in clusters,
but may be continuous at 0.5 to 2 Hz (47). TWs are commonly
state responsive, increasing with stimulation (Fig. 29.30B) and
attenuating with deeper levels of stupor or coma, and may be
abolished or regress after administration of IV drugs or other
AEDs (47). TWs are often associated with diminished levels
of consciousness, and occasionally with myoclonus. There is
 Chapter 29 ■ Nonconvulsive Status Epilepticus 629

Figure 29.27 Generalized periodic epileptiform discharges (GPEDs) at the relatively slow repetition rate of 1 to 1.5 Hz
are usually not considered ictal activity.

Figure 29.28 Generalized periodic epileptiform discharges (GPEDs) in a 77-year-old man following cardiac arrest. Note
the absence of almost all EEG activity other than the discharges.
Figure 29.29 Stimulus-induced rhythmic, periodic, or ictal discharges (SIRPIDs) in a 17-year-old girl with encephalitis.
Despite high-dose thiopental, she continued to have frequent electrographic seizures, both spontaneously and induced by
stimulation as in this example. Some SIRPIDs were brief and seen only on EEG, but others evolved into clinical general-
ized convulsions.

Figure 29.30 A: A 65-year-old woman with uremia and intermittent episodes of unresponsiveness. There are occasional 1
Hz frontally maximal triphasic waves (TWs). The inset shows the characteristic anterior to posterior gradient of TWs—the
first negative component in channel 1 (frontal) precedes the negative component in channel 3 (occipital) by about 50 msec.

Figure 29.30 (Continued) B: With stimulation, the TWs increase in amplitude, field, and frequency, mimicking electro-
graphic seizure activity.
often an underlying toxic or metabolic state, including uremia,
hepatic insufficiency, hypothyroidism; toxicity from lithium,
baclofen, ifosfamide, tiagabine, cefapime, neuroleptics, and
other medications; or the neuroleptic malignant or serotonin
syndromes (41,47,222).
TWs have long been a source of confusion because of their
“epileptiform” appearance on EEG even while they occur in both
encephalopathies and NCSE. Typical TWs are generally not con-
sidered epileptic by most investigators, but they may be identical
to the discharges that occur in definite NCSE, especially when
they have a frequency exceeding 1 Hz (223). TWs are usually of a
broader, more blunted appearance than typical epileptiform dis-
charges. Epileptiform discharges in generalized NCSE usually
have higher frequencies (mean 2.4 Hz vs. 1.8 Hz), more poly-
spikes (69% vs. 0%), sharper morphologies, and less concomitant
background slowing than TWs associated with encephalopathies
(224). TWs also have a relative lack of phase reversal, longer
duration, and a lesser slope of incline in the third phase of the
waveform. A phase lag from anterior to posterior channels is
often seen with TWs (Fig. 29.30A), but not usually with NCSE.
Facial or limb myoclonus is more suggestive of SE. Table 29.5 lists
characteristics that can help to distinguish NCSE from TWs.
Resolution of TWs in response to administration of BDZs
may suggest a diagnosis of NCSE, but BDZs may also abolish
typical, even rhythmic, TWs from an underlying encephalopa-
thy, usually without clinical improvement (41,46,47).
Resolution of abnormalities and resumption of a normal alpha
Chapter 29 ■ Nonconvulsive Status Epilepticus 631
waking pattern are more suggestive of NCSE, whereas resolu-
tion of TWs leaving an encephalopathic slow background sug-
gests a nonepileptic encephalopathy. Still, some true NCSE do
not respond to AEDs or may also give way to an encephalo-
pathic pattern without clinical improvement.
TECHNOLOGY
Rapid detection of nonconvulsive seizures, in the ICU or else-
where, requires near “real-time” review of EEG. Linking contin-
uous digital EEG recordings to a hospital computer network
allows EEG technologists and EEG-ers to access ongoing EEG
from a variety of locations, including offsite over the Internet.
It also facilitates postrecording adjustment of montages, sensi-
tivity, and filters; acquisition and storage of large amounts of
data; quantitative data analysis; and automated spike and
seizure detection. For C-EEG to be useful for clinical decisions,
records should be reviewed and interpreted by trained person-
nel several times a day. In one study, only 40% of nonconvulsive
seizures were detected visually online at the time of the seizure,
with the remainder found at later review (225).
C-EEG produces more data than an individual clinical neu-
rophysiologist can review, at least in its raw form; for example,
following an episode of SE, long periods of C-EEG must be
reviewed to ensure that breakthrough or recurrent seizures have
not occurred. Several digital analysis methods and graphical
displays of quantitative EEG (QEEG) are commonly used to
632 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Tabl e 2 9 . 5

Triphasic Waves versus Periodic Epileptiform Discharges

Triphasic Waves:
Surface-negative, blunted triphasic complexes with (a) low-amplitude, blunted, negative first phase (often wide-based); (b)
dominant, steep positive second phase; and (c) slow rising third “slow-wave” component. No polyspike components.
Complex duration: 400–600 msec
Amplitude: 100–300 V on referential montage; smaller on bipolar.
Frequency: 1.0–2.5 Hz (typically 1.8 Hz).
Persistence: wax and wane, but 10% of a standard 20-minute recording.
Evolution/reactivity: decrease with sleep, drowsiness, or after benzodiazepines; increase and reappear with arousal or noxious
stimulation.
May exhibit phase lag, seen best on referential montage.
Periodic Epileptiform Discharges:
Surface-negative bi-, tri-, or polyphasic discharges with spike, sharp, or polyspike components; slow-wave complexes; or
combinations of these.
Complex duration: 60–600 msec (mean 200 msec)
Amplitude: 50–300 V (usually up to 150 V)
Frequency: 0.2–3 Hz (usually 0.5–2.0 Hz)
Persistence: minimum of 10 minutes in an EEG recording.
Evolution: static—with only minor variability in the above characteristics.

reduce the enormous quantity of data produced by C-EEG.
These methods allow rapid detection of even subtle changes in
EEG activity, obviating review of thousands of pages or screens
of raw EEG data (226).
Automated detection programs identify EEG segments that
include events of interest, such as epileptiform spikes or patterns
suggestive of seizures, flagging events for later review by EEG-ers
(Fig. 29.31). Several commercially available systems have algo-
rithms designed to identify paroxysmal rhythmic activity with
increased voltage or frequency compared to background activity
(227–229). The sensitivity of variables such as frequency (usu-
ally 3 to 20 Hz), field, and voltage can be adjusted to refine detec-
tion. Alternatively, artificial neural networks can be trained to
detect events by presenting many samples representative of
seizures and nonseizures, with the network trained to distin-
guish between the two patterns (227,230).
A practical concern in using detection algorithms is that
nonconvulsive seizures and NCSE exhibit very different elec-
trographic patterns in critically ill patients from those in
patients in standard epilepsy monitoring units (EMUs) being
evaluated for epilepsy surgery or for a question of seizure ver-
sus pseudoseizure (209). Seizures and NCSE in critically ill
patients often have poorer organization, slower frequencies,
longer duration, and indistinct onsets and endings. The differ-
ence in electrographic seizure morphology makes it harder for
standard algorithms to detect seizures in ICUs. Diffuse back-
ground slowing, focal slowing or attenuation, and breach arti-
fact may also alter the sensitivity of detection algorithms. Most
currently available software has been “trained” and validated
primarily on samples from EMUs and ambulatory epilepsy
patients (229,231) and is inadequate for detection of ictal pat-
terns in patients with critical neurologic illnesses. New soft-
ware trained on these patients, and better suited to C-EEG in
ICUs, should increase the utility of these algorithms in the ICU
(212,226,227,232).
One commonly employed technique involves creation of
graphical displays of QEEG data. QEEG uses fast Fourier trans-
formation of digital EEG signals into their component frequen-
cies and amplitudes. Compressed spectral array (CSA) is the
most commonly used graphical display for QEEG. EEG data for
single channels or combinations of channels over each hemi-
sphere, displayed as color graphs of power (e.g., total power;
power in certain frequency bands, usually theta and alpha; ratios
of power in certain bands to total power; etc.) versus time can be
used to detect electrographic seizures (Fig. 29.32) (212). QEEG
methods can help to select episodes for review by EEG-ers (209).
CSA displays can also assist with an initial, bedside EEG inter-
pretation by ICU personnel and physicians not trained in EEG
(212,233), who can then notify EEG technologists or clinical
neurophysiologists to review the EEG sections in question.
Still, automatic detection algorithms do not detect all seizures,
and they often select artifacts as possible seizures; the sensitivity
and specificity of these techniques for NCSE in the ICU have not
been determined well. Power changes can be caused by seizures,
periodic patterns, changes in sleep–wake states, arousal, or arti-
facts, and the specific cause cannot be determined by reviewing
 Chapter 29 ■ Nonconvulsive Status Epilepticus 633

FFT Power 6–14 Left

FFT Power 6–14 Right

FFT Spectrogram 1–20Hz Left

FFT Power 6-14 Left

FFT Spectrogram 1–20Hz Right

Figure 29.31 CSA shows total power in 6 to 14 Hz band in the left hemisphere (top line) and right hemisphere (second
line). Third and fourth lines show color density spectrogram in the right and left hemispheres, respectively. Peaks (in blue,
top two arrays; white, bottom two arrays) indicate seizures. Recurrent generalized seizures are eventually controlled by IV
thiopental (thick arrow), but recur soon as brief seizures (thin arrow), which gradually increase in duration (white arrow).
(See color insert)

the CSA display. The associated original, raw EEG data must be
available for review to exclude state changes and artifacts altering
the QEEG, and to confirm that electrographic seizures are not
missed (Fig. 29.32B). Computer analysis can aid in data compres-
sion, but it cannot currently replace expert clinical neurophysiol-
ogist review.
ARTIFACTS
C-EEG in the ICU is particularly susceptible to contamination
with a variety of artifacts, hindering accurate interpretation (234).
Electrode integrity can be difficult to maintain, necessitating use
of needle electrodes or disk electrodes affixed with collodion. ICU
equipment, bed machinery, hemoperfusion pumps, and ventila-
tors may produce 60-Hz line noise. Artifacts include ECG, ballis-
tocardiographic, and pacemaker effects. They can also result from
skull defects and scalp edema. Nursing care, stimulation, sponta-
neous arousals, chewing artifact, and chest percussion during
physical therapy can all produce artifact. Some artifacts are peri-
odic or rhythmic and mimic epileptiform discharges or electro-
graphic seizures (Figs. 29.33 and 29.34).
ICU patients who are restless or agitated produce substantial
movement artifact on EEG. Nonepileptic but rhythmic or parox-
ysmal movements such as dystonic posturing and (nonepileptic)
myoclonus can produce artifacts as well. Comatose ICU patients
may have clinical episodes of twitching, tremors, posturing, or
sudden changes of heart rate or blood pressure that may appear to
be seizures clinically, and by video. This is not diagnostic unless
correlated with simultaneous EEG evidence of seizure activity
(221). The EEG may show some such activity to be very unlikely
to be epileptic, and inappropriate to treat as seizures. Accurate
interpretation is also facilitated by staff assessment of the patient
and recording nursing interventions on the EEG, and by video and
audio recording obtained simultaneously with the EEG (which
also helps in diagnosing pseudoseizures). Findings that appear
“epileptiform” may be due to artifact, and others may be seizures
when they appear at first to be artifact (221,234). Video–EEG cor-
relation also aids in clarifying the clinical and behavioral correlates
of nonconvulsive seizure activity. Similar behavior might other-
wise be ascribed to postictal or interictal confusion or psychosis.
Many groups use video now even with routine EEG recordings, to
look for artifacts and evidence of their causes.
A
Seizure detections

Seizure probability

Rhythmic run detector, 1–24 Hz, left

Rhythmic run detector, 1–24 Hz, right

FFT spectrogram, 1–24 Hz, left

FFT spectrogram, 1–24 Hz, right

Figure 29.32 A: A 30-year-old woman with mild mental retardation and intractable epilepsy admitted with lethargy and
intermittent head and eye deviation to the right. Continuous EEG monitoring with automated seizure detection shows fre-
quent right hemisphere seizures over a 2-hour period. Seizures are apparent both on rhythmic run detection and display
(black arrows), and color spectrogram (white arrows). EEG showed frequent electrographic seizures beginning with rhyth-
mic beta activity in the right occipital region (seen on traditional EEG display in B). B: Individual episode of seizure activ-
ity (corresponding to large black arrow on rhythmic run detection and color spectrogram in Fig. 29.31A). EEG showed
frequent electrographic seizures beginning with rhythmic beta activity in the right occipital region, gradually increasing in
amplitude and spreading to the right posterior temporal region and left occipital region. (See color insert)
 Chapter 29 ■ Nonconvulsive Status Epilepticus 635

Figure 29.33 Rhythmic artifact from chest percussion obscures much of the EEG.

Fp1-F7
F7-T3
T3-T5
T5-O1

Fp2-F8
F8-T4
T4-T6
T6-02

Fp1-F3
F3-C3
C3-P3
P3-O1

Fp2-F4
F4-C4
C4-P4
P4-O2

Fz-Cz 75 uV 1 sec
Cz-Pz

Figure 29.34 Bed percussion artifact.

636 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
TREATMENT OF NONCONVULSIVE STATUS
EPILEPTICUS IN THE ICU
There have been no randomized controlled trials evaluating the
treatment of NCSE. Treatment decisions are extrapolated from
trials of GCSE treatment or tailored to the perceived urgency and
morbidity of the type of NCSE encountered (235,236). The like-
lihood of successful treatment is strongly related to the etiology
and type of NCSE. Treatment and outcome of absence, primarily
generalized NCSE (i.e., NCSE in the IGE syndromes), and CPSE
are detailed earlier in the discussions of the several individual
types of NCSE.
Absence and de novo absence SE are usually treated success-
fully and easily and seldom require aggressive treatment.
Atypical absence SE is much harder to diagnose and treat.
Except for occasional AASE and some cases of CPSE (also diffi-
cult to treat at times), they are seldom seen in ICUs. Other
forms of NCSE, particularly secondarily generalized NCSE, are
more often refractory to treatment (24).
Patients with NCSE in the setting of earlier epilepsy are usually
treated successfully and generally warrant a less aggressive
approach than does GCSE—unless the NCSE is the continuation
of GCSE (157). Treatment is usually similar to that for GCSE
(see the section “Treatment of Refractory Status Epilepticus” in
Chapter 28), but the most aggressive therapies, often those caus-
ing respiratory compromise, are usually avoided (158,159), unless
NCSE is refractory (236). Even with NCSE following GCSE, or in
the setting of acute medical illness with severe encephalopathy,
much of the diminished level of consciousness and consequent
neurologic morbidity stems from the underlying cause of seizures
rather than from the seizures themselves (195,237).
While risks of overtreatment exist, most patients with NCSE
should be treated quickly with AEDs because they are having
ongoing seizures with impaired consciousness and other neuro-
logic deficits that are potentially reversible. There is also the
attendant morbidity of incidental trauma, aspiration pneumo-
nia, and so on, and many episodes of NCSE begin, and may
end, with potentially harmful generalized convulsions.
EEG USE IN THE ICU MANAGEMENT
OF NONCONVULSIVE STATUS EPILEPTICUS
Once nonconvulsive seizures or NCSE have been diagnosed,
most patients still need EEG monitoring during prolonged
treatment in the ICU. Because clinical signs of NCSE are non-
specific, subtle, or nonexistent, C-EEG is usually necessary to
guide the use of intensive treatment and to assess the response to
therapy (Chapter 28). NCSE persists in 14% to 20% of patients
after cessation of clinically evident seizure activity (98). EEG also
helps to determine if NCSE recurs after treatment. Relapse of
NCSE is not rare, especially in the first 24 hours of treatment
(98) and also when AEDs are being tapered, but it portends a
worsened outcome and should be avoided if possible (238).
Individual seizures can be missed if EEG monitoring is inter-
mittent. A given EEG may show a “postictal” encephalopathy,
drug effect, isolated or periodic discharges, or a burst suppres-
sion tracing at one point, and nonconvulsive seizures or NCSE
may occur shortly thereafter (Figs. 29.35 and 29.36; see also Fig.
28.9A,B). Without C-EEG, the seizures will not be detected.
When NCSE recurs, it should be re-treated while following
the EEG, but the best method, intensity, and duration of treat-
ment have not yet been determined (see the section “EEG Use
in Refractory Status Epilepticus” of Chapter 28). For refractory
NCSE, many neurologists will use additional standard AEDs
and then a trial of aggressive treatment (often with midazolam,
pentobarbital, or propofol) to at least suppress all electro-
graphic and clinical seizures (239,240).
C-EEG aids in the recognition of nonconvulsive, electro-
graphic seizures and may also provide important prognostic
information (see Chapter 28). Earlier diagnosis and EEG-guided
treatment may improve the care of patients with SE, but its actual
effect on outcome has not yet been established (237,241). No
prospective studies have evaluated the use of C-EEG compared to
routine EEG or other clinical measures for SE in the ICU.
MORBIDITY OF NONCONVULSIVE STATUS
EPILEPTICUS
Outcome for specific NCSE types was detailed earlier along
with the clinical features of simple partial SE, absence SE, and
well-defined cases of CPSE, but many clinical series describe
NCSE without specifying the type. The prognosis for these
many different illnesses depends heavily on the type of NCSE,
as identified clinically and by EEG, and on their causes and pre-
cipitants, and may also be influenced by age, concomitant med-
ical illness, duration of SE, and the effectiveness and
complications of treatment. Etiology is by far the most impor-
tant prognostic factor. Some types of NCSE have an excellent
prognosis, without evidence of morbidity in any case, but oth-
ers have a high likelihood of a fatal outcome. For example, there
is essentially no mortality with simple partial sensory SE (107).
Longer-term risks of NCSE should consider both experi-
mental and clinical data—but most basic studies on the damage
due to SE have used models of GCSE. While GCSE poses a def-
inite threat to health and warrants vigorous treatment (Chapter
28), there is very little evidence of lasting harm due to NCSE
alone (242). The treatment imperative is less, but not negligible.
NCSE is potentially “underdiagnosed and overtreated” (136).
In non-ICU patients, clinical reports of NCSE show few
long-term sequelae of the NCSE itself (25,26,44,243), but most
had limited follow up. Cognitive deficits caused by NCSE,
including language dysfunction, can persist for months (51).
One study of 65 NCSE patients found good outcomes in all but
one (244). Another reported neuropsychologic testing on 15
epilepsy patients (without underlying lesions or “symptomatic”
causes) before and after episodes of SE (half nonconvulsive)
(245) demonstrating that no significant cognitive morbidity
accrued from the SE. Long-term cognitive complications of
NCSE appear uncommon, occurring primarily in the most pro-
longed and severe cases.
Outcome is not nearly as favorable for patients with second-
arily generalized NCSE in the setting of serious medical and neu-
rologic illness, especially in the ICU. The physiologic effects of
persistent NCSE in critically ill patients can include an increased
 Chapter 29 ■ Nonconvulsive Status Epilepticus 637

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Fp1-F7

F7-T3

T3-T5

T5-O1

Figure 29.35 Evolution of an electrographic seizure in a comatose 55-year-old woman with a left temporal intracerebral
hemorrhage. EEG shows left temporal periodic lateralized epileptiform discharges (PLEDs) evolving into faster frequency,
sharply contoured delta activity. Evolution of the seizure includes spread in the left temporal region, and then left parasagit-
tal region, and eventual end of the seizure. The patient remained comatose, without clinical manifestation of the seizure.

Fp1-F3

F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4

P4-O2
Fp1-F7

F7-T3
T3-T5
T5-O1

Fp2-F8
F8-T4

T4-T6
T6-O2
Fz-Cz

Cz-Pz
LUC
RLC
EKG
75 uV 1 sec

Figure 29.36 A focal seizure arises from the left hemisphere after an EEG showing PLEDs (see Fig. 28.7A) in the left pari-
etal and posterior temporal regions in a 26-year-old woman with adult onset Rasmussen encephalitis and earlier focal sta-
tus epilepticus.
638 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
cerebral blood flow, intracranial pressure, oxygen demands, and
the lactate–pyruvate ratio, even when clinical signs are absent
(246,247). NCSE might exacerbate neuronal damage in patients
with acute insults such as traumatic brain injury (TBI) or stroke
(201). Patients with both intracerebral hemorrhage and noncon-
vulsive electrographic seizures had a worse outcome in two
reports, but it was not possible to determine if this was due to
seizures or (more likely) the underlying illness (205,206). Among
94 consecutive patients with moderate-to-severe TBI, 22% had
electrographic seizures on EEG and 29% were in NCSE, with
higher mortality in these patients than in similar trauma patients
without seizures (201). It remains unclear whether treatment of
nonconvulsive seizures would have changed this (248). In one
series of 49 patients with nonconvulsive seizures, mortality cor-
related with etiology, age, and length of ICU stay, but only seizure
duration and delay-to-diagnosis were significant predictors upon
multivariate analysis (191).
One series of patients with NCSE found that patients with
minimal obtundation had a mortality of 7% versus 39% for
those with deep lethargy or coma (157). Patients with anoxia or
multiple medical problems, including sepsis, fare poorly, while
those with strokes, tumors, trauma, infection, alcohol abuse, or
other drugs have intermediate results (16). Generally, the most
favorable etiology is epilepsy itself with some relatively benign
precipitant, such as reduced AEDs.
In the setting of severe medical or neurologic disease, the
morbidity and mortality of NCSE are substantial (157),
although it is not possible to dissect out that portion of the
long-term harm done by epileptiform discharges, seizures, or
SE from the damage caused by the underlying illness
(32,33,195,202). Little of this morbidity is usually attributed to
the NCSE itself, but complications of treatment can also con-
tribute to morbidity (237). Overaggressive therapy can result in
hypotension or prolonged mechanical ventilation. C-EEG can
help guide treatment of SE and help to avoid undertreatment or
overtreatment.
REFERENCES
1. Temkin O, ed. The Falling Sickness. Baltimore: The Johns Hopkins
Press; 1971.
2. Shorvon S. The concept of status epilepticus and its history. In:
Shorvon S, ed. Status Epilepticus: Its Clinical Features and Treatment
in Children and Adults. Cambridge, UK: Cambridge University Press;
1994:1–20.
3. Trousseau A. Lectures on Clinical Medicine Delivered at the Hôtel
Dieu, Paris. Vol 1, translated by Bazire PV. London: New Sydenham
Society; 1868.
4. Falret J. De l’état mental des épileptiques. Arch Gén Méd, Fifth
Series. 1860;16:661–679; 1861;17:461–449.
5. Wilks S. Lectures on Diseases of the Nervous System. London:
Churchill; 1878.
6. Goetz CG. Charcot the Clinician: The Tuesday Lessons. New York,
NY: Raven Press; 1987.
7. Gowers WR. Epilepsy and Other Chronic Convulsive Disorders.
London: Churchill; 1901.
8. Zimmer C. Soul made Flesh. The Discovery of the Brain—and How
it Changed the World. New York, NY: Free Press, Simon & Schuster;
2004.
9. Berger H. Über das Elektrenkephalogramm des Menschen. Arch
Psychiatr Nervenkr. 1929;87:527.
10. Gloor P. The work of Hans Berger. In: Cobb WA, ed. Appraisal and
Perspective of the Functional Explorations of the Nervous System.
Handbook of Electroencephalography and Clinical Neurophysiology.
Amsterdam: Elsevier Publishing Co.; 1971:IA11–IA24.
11. Lennox WG. The treatment of epilepsy. Med Clin North Am. 1945;
29:1114–1128.
12. Penfield W, Jasper HH. Epilepsy and Functional Anatomy of the
Human Brain. Boston: Little Brown; 1954.
13. Gastaut H, Roger A. Sur la signification de certain fugues épilep-
tiques: état de mal temporal. Rev Neurol. 1956;94:298–301.
14. Shorvon SD. What is nonconvulsive status epilepticus, and what
are its subtypes? Epilepsia. 2007;48(Suppl 8):35–38.
15. Lowenstein DH, Bleck T, Macdonald RL. It’s time to revise the def-
inition of status epilepticus. Epilepsia. 1999;40:120–122.
16. DeLorenzo RJ, Hauser WA, Towne AR, et al. A prospective, popula-
tion-based epidemiologic study of status epilepticus in Richmond,
Virginia. Neurology. 1996;46:1029–1035.
17. Cockerell OC, Walker MC, Sander JWAS, et al. Complex partial sta-
tus epilepticus: a recurrent problem. J Neurol Neurosurg Psychiatry.
1994;57:835–837.
18. Tomson T, Svanborg E, Wedlund JE. Nonconvulsive status epilep-
ticus: high incidence of complex partial status. Epilepsia. 1986;
27:276–285.
19. Kaplan PW. Behavioral manifestations of nonconvulsive status
epilepticus. Epilepsy Behav. 2002;3:122–139.
20. Kaplan PW, Tusa RJ. Neurophysiologic and clinical correlations of
epileptic nystagmus. Neurology. 1993;43:2508–2514.
21. Husain AM, Horn GJ, Jacobson MP. Non-convulsive status epilep-
ticus: usefulness of clinical features in selecting patients for urgent
EEG. J Neurol Neurosurg Psychiatry. 2003;74:189–191.
22. Celesia GG, Grigg MM, Ross E. Generalized status myoclonicus in
acute anoxic and toxic-metabolic encephalopathies. Arch Neurol.
1988;45:781–784.
23. Knake S, Rosenow F, Vescovi M, et al. Incidence of status epilepti-
cus in adults in Germany: a prospective, population-based study.
Epilepsia. 2001;42:714–718.
24. Kaplan PW. Nonconvulsive status epilepticus in the emergency
room. Epilepsia. 1996;37:643–650.
25. Guberman A, Cantu-Reyna G, Stuss D, et al. Nonconvulsive
generalized status epilepticus: clinical features, neuropsychological
testing, and long-term follow-up. Neurology. 1986;36:1284–1291.
26. Dunne JW, Summers QA, Stewart-Wynne EG. Non-convulsive
status epilepticus: a prospective study in an adult general hospital.
Q J Med. 1987;62:117–126.
27. Stores G, Zaiwalla Z, Styles E, et al. Non-convulsive status epilep-
ticus. Arch Dis Child. 1995;73:106–111.
28. Brodtkorb E, Sand T, Kristiansen A, et al. Non-convulsive status
epilepticus in the adult mentally retarded. Seizure. 1993;2:115–123.
29. Thomas P, Beaumanoir A, Genton P, et al. “De novo” absence sta-
tus of late onset: report of 11 cases. Neurology. 1992;42:104–110.
30. Tatum WO, Ross J, Cole AJ. Epileptic pseudodementia. Neurology.
1998;50:1472–1475.
31. Chicarro AV, Kanner AM. Psychiatric manifestations of noncon-
vulsive status epilepticus. In: Kaplan PW, Drislane FW, eds.
Nonconvulsive Status Epilepticus. New York, NY: Demos Medical
Publishing; 2009:203–215.
32. Lowenstein D, Aminoff M. Clinical and EEG features of status
epilepticus in comatose patients. Neurology. 1992;42:100–104.
33. Drislane FW, Schomer DL. Clinical implications of generalized
electrographic status epilepticus. Epilepsy Res. 1994;19:111–121.

34. Pritchard PB III, O’Neal DB. Nonconvulsive status epilepticus fol-
lowing metrizamide myelography. Ann Neurol. 1984;16:252–254.
35. Zak R, Solomon G, Petito F, et al. Baclofen-induced generalized
nonconvulsive status epilepticus. Ann Neurol. 1994;36:113–114.
36. Panayiotopoulos CP, Agathonikou A, Sharoqi IA, et al. Vigabatrin
aggravates absences and absence status. Neurology. 1997;49:1467.
37. Shinnar S, Berg AT, Treiman DT, et al. Status epilepticus and
tiagabine therapy: review of safety data and epidemiologic com-
parisons. Epilepsia. 2001;42:372–379.
38. Feddersen B, Bender A, Arnold S, et al. Aggressive confusional state as
a clinical manifestation of status epilepticus in MELAS. Neurology.
2003;61:1149–1150.
39. Dalmau J, Rosenfeld MR. Paraneoplastic syndromes of the CNS.
Lancet Neurol. 2008;7:327–340.
40. Dalmau J, Gleichman AJ, Hughes EG, et al. Anti-NMDA-receptor
encephalitis: case series and analysis of the effects of antibodies.
Lancet Neurol. 2008;7:1091–1098.
41. Kaplan PW, Birbeck G. Lithium-induced confusional states: non-
convulsive status epilepticus or triphasic encephalopathy? Epilepsia.
2006;47:2071–2074.
42. Kaplan PK. Nonconvulsive status epilepticus. Semin Neurol. 1996;
16:33–40.
43. Kaplan PW. The EEG of status epilepticus. J Clin Neurophysiol.
2006;23:221–229.
44. Tomson T, Lindbom U, Nilsson BY. Nonconvulsive status epilepti-
cus in adults: thirty-two consecutive patients from a general hos-
pital population. Epilepsia. 1992;33:829–835.
45. Granner MA, Lee SI. Nonconvulsive status epilepticus: EEG analy-
sis in a large series. Epilepsia. 1994;35:42–47.
46. Kaplan PW. The EEG criteria for NCSE. Epilepsia. 2007;48:39–41.
47. Fountain NB, Waldman WA. Effects of benzodiazepines on tripha-
sic waves: implications for nonconvulsive status epilepticus. J Clin
Neurophysiol. 2001;18:345–352.
48. Treiman DM, Meyers PD, Walton NY, et al. A comparison of four
treatments for generalized convulsive status epilepticus. Veterans
Affairs Status Epilepticus Cooperative Study Group. N Engl J Med.
1998;339:792–798.
49. Fagan KJ, Lee SI. Prolonged confusion following convulsions due
to generalized non-convulsive status epilepticus. Neurology. 1990;
40:1689–1694.
50. Dong C, Sriram S, Delbeke D, et al. Aphasic or amnestic status epilep-
ticus detected on PET but not EEG. Epilpesia. 2009;50:251–255.
51. van Paesschen W, Porke K, Fannes K, et al. Cognitive deficits dur-
ing status epilepticus and time course of recovery: a case report.
Epilepsia. 2007;48:1979–1983.
52. Kutluay E, Beattie J, Passaro EA, et al. Diagnostic and localizing
value of ictal SPECT in patients with nonconvulsive status epilep-
ticus. Epilepsy Behav. 2005;6:212–217.
53. Lothman EW, Bertram EH, Bekenstein JW, et al. Self-sustaining
limbic status epilepticus induced by 'continuous’ hippocampal
stimulation: electrographic and behavioral characteristics. Epilepsy
Res. 1989;3:107–119.
54. van Landingham KE, Lothman EW. Self-sustaining limbic status
epilepticus. I. Acute and chronic cerebral metabolic studies: limbic
hypermetabolism and neocortical hypometabolism. Neurology.
1991;41:1942–1949.
55. Kapur J. Pathophysiology of nonconvulsive status epilepticus. In:
Kaplan PW, Drislane FW, eds. Nonconvulsive Status Epilepticus.
New York, NY: Demos Medical Publishing; 2009:81–94.
56. van Landingham KE, Lothman EW. Self-sustaining limbic status
epilepticus. II. Role of hippocampal commissures in metabolic
responses. Neurology. 1991;41:1950–1957.
Chapter 29 ■ Nonconvulsive Status Epilepticus 639
57. Lothman EW, Bertram EH. Epileptogenic effects of status epilep-
ticus. Epilepsia. 1993;34(Suppl 1):59–70.
58. Wasterlain CG, Mazarati AM, Naylor D, et al. Short-term plastic-
ity of hippocampal neuropeptides and neuronal circuitry in exper-
imental status epilepticus. Epilepsia. 2002;43(Suppl 5):20–29.
59. Mazarati AM, Halaszi E, Telegdy G. Anticonvulsive effects of
galanin administered into the central nervous system upon the
picrotoxin-kindled seizure syndrome in rats. Brain Res. 1992;
589:164–166.
60. Liu H, Mazarati AM, Katsumori H, et al. Substance P is expressed
in hippocampal principal neurons during status epilepticus and
plays a critical role in the maintenance of status epilepticus. Proc
Natl Acad Sci USA. 1999;96:5286–5291.
61. Hughes JR. Absence seizures: a review of recent reports with new
concepts. Epilepsy Behav. 2009;15:404–412.
62. Gloor P, Fariello RG. Generalized epilepsy: some of its cellular
mechanisms differ from those of focal epilepsy. Trends Neurosci.
1988;11:63–68.
63. Jasper HH, Droogleever-Fortuyn J. Experimental studies on the
functional anatomy of petit mal epilepsy. Res Publ Ass Nerv Ment
Dis. 1947;26:272–298.
64. Williams D. The thalamus and epilepsy. Brain. 1965;88:539–556.
65. Tien RD, Felsberg GJ. The hippocampus in status epilepticus:
demonstration of signal intensity and morphologic changes with
sequential fast spin–echo MR imaging. Radiology. 1995;194:249–256.
66. Wieshmann UC, Woermann FG, Lemieux L, et al. Development of
hippocampal atrophy: a serial magnetic resonance imaging study
in a patient who developed epilepsy after generalized status epilep-
ticus. Epilepsia. 1997;38:1238–1241.
67. Lothman E. The biochemical basis and pathophysiology of status
epilepticus. Neurology. 1990;40(Suppl 2):13–23.
68. Lowenstein DH, Shimosaka S, So YT, et al. The relationship between
electrographic seizure activity and neuronal injury. Epilepsy Res.
1991;10:49–54.
69. Bertram EH, Lothman EW, Lenn NJ. The hippocampus in experi-
mental chronic epilepsy: a morphometric analysis. Ann Neurol.
1990;27:43–48.
70. Gastaut H. Classification of status epilepticus. Adv Neurol. 1983;
34:15–35.
71. Shorvon S. The classification of nonconvulsive status epilepticus.
In: Kaplan PW, Drislane FW, eds. Nonconvulsive Status Epilepticus.
New York, NY: Demos Medical Publishing; 2009:11–22.
72. Chong DJ, Hirsch LJ. Which EEG patterns warrant treatment in
the critically ill? Reviewing the evidence for treatment of periodic
epileptiform discharges and related patterns. J Clin Neurophysiol.
2005;22:79–91.
73. Olmos-Garcia de Alba G, Udaeta Mora E, Malagon Valdez J, et al.
Neonatal status epilepticus II: electroencephalographic aspects.
Clin Electroencephalogr. 1984;15:197–201.
74. Mizrahi EM, Kellaway P. Characterization and classification of
neonatal seizures. Neurology. 1987;37:1837–1844.
75. Knauss TA, Carlson CB. Neonatal paroxysmal monorhythmic alpha
activity. Arch Neurol. 1978;35:104–107.
76. Scher MS. Controversies regarding neonatal seizure recognition.
Epileptic Disord. 2002;4:139–158.
77. Clancy RR, Legido A. The exact ictal and interictal duration of elec-
troencephalographic neonatal seizures. Epilepsia. 1987;28:537–541.
78. Dravet C, Bureau M, Oguni H, et al. Severe myoclonic epilepsy in
infancy: Dravet syndrome. Adv Neurol. 2005;95:71–102.
79. Korff CM, Laux L, Kelley KR, et al. Dravet syndrome (severe
myoclonic epilepsy in infancy): a retrospective study of 16 patients.
J Child Neurol. 2007;22:185–194.
640 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
80. Ohtahara S, Yamatogi Y. Ohtahara syndrome: with special refer-
ence to it developmental aspects for differentiating from early
myoclonic encephalopathy. Epilepsy Res. 2006;70(Suppl 1):58–67.
81. Bureau M. ‘Continuous spikes and waves during slow sleep’
(CSWS): definition of the syndrome. In: Beaumanoir A, Bureau
M, Deonna T, et al., eds. Continuous Spikes and Waves During Slow
Sleep. London: John Libbey and Co.; 1995:17–26.
82. Riviello JJ Jr. Status epilepticus in children. In: Drislane FW, ed.
Status Epilepticus: a Clinical Perspective. Totowa, NJ: Humana
Press; 2005:313–338.
83. Hadjiloizou S, Riviello JJ. Epileptic and epileptiform
encephalopathies. Neurology, eMedicine. WebMD. Available at:
http://www.emedicine.com/neuro/topic547.htm. Accessed on
May 15, 2006.
84. Patry G, Lyagoubi S, Tassinari CA. Subclinical “electrical status
epilepticus” induced by sleep in children. A clinical and elec-
troencephalographic study of six cases. Arch Neurol. 1971;24:
242–252.
85. Tassinari CA, Bureau M, Dravet C, et al. Epilepsy with continuous
spikes and waves during slow sleep—otherwise described as ESES
(epilepsy with electrical status epilepticus during slow sleep). In:
Roger J, Bureau M, Dravet C, et al., eds. Epileptic Syndromes in
Infancy, Childhood, and Adolescence. 2nd ed. London: John Libbey
and Co.; 1992:245–256.
86. De Negri M. Electrical status epilepticus during sleep (ESES).
Different clinical syndromes: towards a unifying view? Brain Dev.
1997;19:447–451.
87. Guilhoto LMFF, Morrell F. Electrophysiological differences between
Landau-Kleffner Syndrome and other conditions showing the
CSWS electrical pattern. Epilepsia. 1994;35(Suppl 8):126. Abstract.
88. Landau WM, Kleffner FR. Syndrome of acquired epileptic aphasia
with convulsive disorder in children. Neurology. 1957;7:523–530.
89. Bishop DVM. Age of onset and outcome in acquired aphasia with
convulsive disorder (Landau-Kleffner Syndrome). Dev Med Child
Neurol. 1985;27:705–712.
90. Riviello JJ Jr. Nonconvulsive status epilepticus in children. In:
Kaplan PW, Drislane FW, eds. Nonconvulsive Status Epilepticus.
New York, NY: Demos Medical Publishing; 2009:227–237.
91. Lerman P, Lerman-Sagie T, Kivity S. Effect of early corticosteroid
therapy for Landau-Kleffner Syndrome. Dev Med Child Neurol.
1991;33:257–260.
92. Shinnar S, Rapin I, Arnold S, et al. Language regression in child-
hood. Pediatr Neurol. 2001;24:183–189.
93. McVicar KA, Shinnar S. Landau-Kleffner syndrome, electrical sta-
tus epilepticus in slow wave sleep, and language regression in chil-
dren. Ment Retard Dev Disabil Res Rev. 2004;10:144–149.
94. Jayakar PB, Seshia SS. Electrical status epilepticus during slow-
wave sleep: a review. J Clin Neurophysiol. 1991;8:299–311.
95. Kramer U, Nevo Y, Neufeld MY, et al. Epidemiology of epilepsy in
childhood: a cohort of 440 consecutive patients. Pediatr Neurol.
1998;18:46–50.
96. Inoue Y, Fujiwara T, Matsuda K, et al. Ring chromosome 20 and
nonconvulsive status epilepticus. A new syndrome. Brain. 1997;120:
939–953.
97. Devinsky O, Kelley K, Porter RJ, et al. Clinical and electroen-
cephalographic features of simple partial seizures. Neurology. 1988;
38:1347–1352.
98. DeLorenzo RJ, Waterhouse EJ, Towne AR, et al. Persistent noncon-
vulsive status epilepticus after the control of convulsive status epilep-
ticus. Epilepsia. 1998;39:833–840.
99. Friedman DE, Schevon C, Emerson RG, et al. Cyclic electrographic
seizures in critically ill patients. Epilepsia. 2008;49:281–287.
100. Iizuka T, Sakai F, Suzuki N, et al. Neuronal hyperexcitability in
stroke-like episodes of MELAS syndrome. Neurology. 2002;59:
816–824.
101. Taylor I, Scheffer IE, Berkovic SF. Occipital epilepsies: identifica-
tion of specific and newly recognized syndromes. Brain. 2003;
126:753–769.
102. Wieser HG, Fischer M. Temporal lobe nonconvulsive status
epilepticus. In: Kaplan PW, Drislane FW, eds. Nonconvulsive
Status Epilepticus. New York, NY: Demos Medical Publishing; 2009:
119–137.
103. Seshia SS, McLachlan RS. Aura continua. Epilepsia. 2005;46:
454–455.
104. Fejerman N, DiBlasi AM. Status epilepticus of benign partial
epilepsies in children: report of two cases. Epilepsia. 1987;28:
351–355.
105. Colamaria V, Sgro V, Caraballo R, et al. Status epilepticus in
benign rolandic epilepsy manifesting as anterior operculum syn-
drome. Epilepsia. 1991;32:329–334.
106. Scott JS, Masland RL. Occurrence of “continuous symptoms” in
epilepsy patients. Neurology. 1953;3:297–301.
107. Manford M, Shorvon SD. Prolonged sensory or visceral symp-
toms: an under-diagnosed form of non-convulsive focal (simple
partial) status epilepticus. J Neurol Neurosurg Psychiatry. 1992;
55:714–716.
108. Blair HT. Temporal lobe or psychomotor status epilepticus. A case
report. Electroencephalogr Clin Neurophysiol. 1980;48:558–572.
109. Barry E, Sussman NM, Bosley TM, et al. Ictal blindness and sta-
tus epilepticus amauroticus. Epilepsia. 1985;26:577–584.
110. Salanova V, Andermann F, Olivier A, et al. Occipital lobe epilepsy:
electroclinical manifestations, electrocorticography, cortical
stimulation and outcome in 42 patients treated between 1930
and 1991. Brain. 1992;115:1655–1680.
111. Sowa MV, Pituck S. Prolonged spontaneous complex visual hal-
lucinations and illusions as ictal phenomena. Epilepsia. 1989;
30:524–526.
112. Young G, Blume WT. Painful epileptic seizures. Brain. 1983;106:
537–554.
113. Peppercorn M, Herzog AG. The spectrum of abdominal epilepsy
in adults. Am J Gastroenterol. 1989;84:1294–1296.
114. Eschle D, Siegel AM, Wieser H-G. Epilepsy with severe abdomi-
nal pain. Mayo Clin Proc. 2002;77:1358–1360.
115. Kikumoto K, Yoshinaga H, Kobayashi K, et al. Complex partial
status epilepticus in children with epilepsy. Brain Dev. 2009;31:
148–157.
116. Panayiotopoulos CP. Autonomic seizures and autonomic status
epilepticus peculiar to childhood: diagnosis and management.
Epilepsy Behav. 2004;5:286–295.
117. Engel J, Ludwig BI, Fetell M. Prolonged partial complex status
epilepticus: EEG and behavioral observations. Neurology. 1978;
28:863–869.
118. Treiman DM, Delgado-Escueta AV. Complex partial status epilep-
ticus. In: Delgado-Escueta AV, Wasterlain CG, Treiman DM, et al.,
eds. Status Epilepticus (Advances in Neurology, Vol 34). New York,
NY: Raven Press; 1983:69–81.
119. Butler CR, Graham KS, Hodges JR, et al. The syndrome of tran-
sient epileptic amnesia. Ann Neurol. 2007;61:587–598.
120. Palmini AL, Gloor P, Jones-Gotman M. Pure amnestic seizures in
temporal lobe epilepsy. Definition, clinical symptomatology and
functional anatomical considerations. Brain. 199;115:749–769.
121. Vuilleumier P, Despland PA, Regli F. Failure to recall (but not to
remember): pure transient amnesia during nonconvulsive status
epilepticus. Neurology. 1996;46:1036–1039.

122. Profitlich T, Hoppe C, Reuber M, et al. Ictal neuropsychological
findings in focal nonconvulsive status epilepticus. Epilepsy Behav.
2008;12:269–275.
123. Hamilton NG, Matthews T. Aphasia: the sole manifestation of
focal status epilepticus. Neurology. 1979;29:745–748.
124. Racy A, Osborn MA, Vern BA, et al. Epileptic aphasia. First onset
of prolonged monosymptomatic status epilepticus in adults.
Arch Neurol. 1980;37:419–422.
125. Gilmore RL, Heilman KM. Speech arrest in partial seizures:
evidence of an associated language disorder. Neurology. 1981;31:
1016–1019.
126. Knight RT, Cooper J. Status epilepticus manifesting as reversible
Wernicke’s aphasia. Epilepsia. 1986;27:301–304.
127. Kirshner HS, Hughes T, Fakhoury T, et al. Aphasia secondary to
partial status epilepticus of the basal temporal language area.
Neurology. 1995;45:1616–1618.
128. Benatar M. Ictal aphasia. Epilepsy Behav. 2002;3:413–419.
129. Schomer DL. Focal status epilepticus and epilepsia partialis
continua in adults and children. Epilepsia. 1993;34(Suppl 1):
S29–S36.
130. Grand’Maison F, Reiher J, Leduc CP. Retrospective inventory of
EEG abnormalities in partial status epilepticus. Electroencephalogr
Clin Neurophysiol. 1991;79:264–270.
131. Hirsch LJ, Emerson RG, Pedley TA. Prolonged “postictal” aphasia:
demonstration of persistent ictal activity with intracranial elec-
trodes. Neurology. 2001;56:134–136.
132. Jobst BC, Roberts DW, Williamson PD. Occipital lobe noncon-
vulsive status epilepticus. In: Kaplan PW, Drislane FW, eds.
Nonconvulsive Status Epilepticus. New York, NY: Demos Medical
Publishing; 2009:139–150.
133. Drislane FW, Blum AS, Schomer DL. Focal status epilepticus: clin-
ical features and significance of different EEG patterns. Epilepsia.
1999;40:1254–1260.
134. Ballenger CE, King DW, Gallagher BB. Partial complex status
epilepticus. Neurology. 1983;33:1545–1552.
135. Williamson PD. Complex partial status epilepticus. In: Engel JE Jr,
Pedley T, eds. Epilepsy. A Comprehensive Text. 2nd ed. Philadelphia,
PA: Lippincott Williams & Wilkins; 2008:677–692.
136. Kaplan PW. Assessing the outcomes in patients with nonconvulsive
status epilepticus: nonconvulsive status epilepticus is underdiag-
nosed, potentially overtreated, and confounded by comorbidity. J
Clin Neurophysiol. 1999;16:341–352.
137. Mikati MA, Lee WL, DeLong GR. Protracted epileptiform
encephalopathy: an unusual form of partial complex status epilep-
ticus. Epilepsia. 1985;25:563–571.
138. Niedermeyer E, Fineyre F, Riley T, et al. Absence status (petit mal
status) with focal characteristics. Arch Neurol. 1979;36:417–421.
139. Thomas P, Zifkin B, Migneco O, et al. Nonconvulsive status
epilepticus of frontal origin. Neurology. 1999;52:1174–1183.
140. Belafsky MA, Carwille S, Miller P, et al. Prolonged epileptic twi-
light states: continuous recordings with nasopharyngeal elec-
trodes and videotape analysis. Neurology. 1978;28:239–245.
141. Commission on Epidemiology and Prognosis, International
League Against Epilepsy. Guidelines for epidemiologic studies on
epilepsy. Epilepsia. 1993;34:592–596.
142. Shorvon S. Clinical forms of status epilepticus. In: Shorvon S, ed.
Status Epilepticus: Its Clinical Features and Treatment in Children
and Adults. Cambridge, UK: Cambridge University Press; 1994:
34–138.
143. Krumholz A, Sung GY, Fisher RS, et al. Complex partial status
epilepticus accompanied by serious morbidity and mortality.
Neurology. 1995;45:1499–1504.
Chapter 29 ■ Nonconvulsive Status Epilepticus 641
144. Shorvon S. Definition, classification and frequency of status
epilepticus. In: Shorvon S, ed. Status Epilepticus: Its Clinical
Features and Treatment in Children and Adults. Cambridge, UK:
Cambridge University Press; 1994:21–33.
145. Williamson PD, Spencer DD, Spencer SS, et al. Complex partial
status epilepticus: a depth-electrode study. Ann Neurol. 1985;18:
647–654.
146. Rohr-le-Floch J, Gauthier G, Beaumanoir A. États confusionelles
d’origine épileptique: interêt de l’EEG fait en urgence. Rev Neurol
(Paris). 1988;144:425–436.
147. Markand ON, Wheeler GL, Pollack SL. Complex partial status
epilepticus (psychomotor status). Neurology. 1978;28:189–196.
148. Mayeux R, Lueders H. Complex partial status epilepticus: case
report and proposal for diagnostic criteria. Neurology. 1978;28:
957–961.
149. Salanova V, Morris HH, Van Ness P, et al. Frontal lobe seizures:
electroclinical syndromes. Epilepsia. 1995;36:16–24.
150. Williamson PD, French JA, Thadani VM, et al. Characteristics of
medial temporal lobe epilepsy: II. Interictal and ictal scalp elec-
troencephalography, neuropsychological testing, neuroimaging,
surgical results, and pathology. Ann Neurol. 1993;34:781–787.
151. Foldvary N, Lee N, Thwaites G, et al. Clinical and electrographic
manifestations of lesional neocortical temporal lobe epilepsy.
Neurology. 1997;49:757–763.
152. Nowack WJ, Shaikh IA. Progression of electroclinical changes in
complex partial status epilepticus: filling in the blanks. Clin
Electroencephalogr. 1999;30:5–8.
153. Wieser HG. Temporal lobe or psychomotor status epilepticus: a
case report. Electroencephalogr Clin Neurophysiol. 1980;48:558–572.
154. Williamson PD, Boon PA, Thadani VM, et al. Parietal lobe
epilepsy: diagnostic considerations and results of surgery. Ann
Neurol. 1992;31:193–201.
155. Gloor P, Olivier A, Quesney LF, et al. The role of the limbic sys-
tem in experiential phenomena of temporal lobe epilepsy. Ann
Neurol. 1982;12:129–144.
156. Kudo T, Sato K, Yagi K, et al. Can absence status epilepticus be of
frontal lobe origin? Acta Neurol Scand. 1995;92:472–477.
157. Shneker BF, Fountain NB. Assessment of acute morbidity and
mortality in nonconvulsive status epilepticus. Neurology. 2003;
61:1066–1073.
158. Shorvon S, Baulac M, Cross H, et al. The drug treatment of sta-
tus epilepticus in Europe: consensus document from a workshop
at the first London Colloquium on Status Epilepticus. Epilepsia.
2008;49:1277–1284.
159. Meierkord H, Holtkamp M. Non-convulsive status epilepticus in
adults: clinical forms and treatment. Lancet Neurol. 2007;6:329–339.
160. Towne AR, Pellock JM, Ko D, et al. Determinants of mortality in
status epilepticus. Epilepsia. 1994;35:27–34.
161. McBride MC, Dooling EC, Oppenheimer EY. Complex partial sta-
tus epilepticus in young children. Ann Neurol. 1981;9:526–530.
162. Berg AT, Shinnar S. Do seizures beget seizures? An assessment of the
clinical evidence in humans. J Clin Neurophysiol. 1997;14:102–110.
163. Schwab RS. A case of status epilepticus in petit mal.
Electroencephalogr Clin Neurophysiol. 1953;5:441–442.
164. Niedermeyer E, Khalifeh R. Petit-mal status (“spike-wave stu-
por”). An electro-clinical appraisal. Epilepsia. 1965;6:250–262.
165. Wheless JW. Acute management of seizures in the syndromes of
idiopathic generalized epilepsies. Epilepsia. 2003;44(Suppl 2):
22–26.
166. Agathonikou A, Panayiotopoulos CP, Giannakodimos S, et al.
Typical absence status in adults: diagnostic and syndromic con-
siderations. Epilepsia. 1998;39:1265–1276.
642 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
167. Shorvon S, Walker M. Status epilepticus in idiopathic generalized
epilepsy. Epilepsia. 2005;46:73–79.
168. Thomas P, Snead OC. Absence status epilepticus. In: Engel J,
Pedley TA, eds. Epilepsy: A Comprehensive Textbook. 2nd ed.
Philadelphia, PA: Lippincott-Raven Publishers; 2008:693–703.
169. Genton P, Ferlazzo E, Thomas P. Absence status epilepsy: delin-
eation of a distinct idiopathic generalized epilepsy syndrome.
Epilepsia. 2008;49:642–649.
170. Kimura S, Kobayashi T. Two patients with juvenile myoclonic
epilepsy and nonconvulsive status epilepticus. Epilepsia. 1996;37:
275–279.
171. Osorio I, Reed RC, Peltzer JN. Refractory idiopathic status
epilepticus: a probable paradoxical effect of phenytoin and car-
bamazepine. Epilepsia. 2000;41:887–894.
172. Andermann F, Robb JP. Absence status: a reappraisal following
review of thirty-eight patients. Epilepsia. 1972;13:177–187.
173. Niedermeyer E, Ribeiro M. Considerations of nonconvulsive sta-
tus epilepticus. Clin EEG. 2000;31:192–195.
174. Baykan B, Gokyigit A, Gurses C, et al. Recurrent absence status
epilepticus: clinical and EEG characteristics. Seizure. 2002;11:
310–319.
175. Berkovic SF, Bladin PF. Absence status in adults. Clin Exp Neurol.
1983;19:198–207.
176. Treiman DM. Electroclinical features of status epilepticus. J Clin
Neurophysiol. 1995;12:343–362.
177. Hovinga CA, Chicella MF, Rose DF, et al. Use of intravenous val-
proate in three pediatric patients with nonconvulsive or convul-
sive status epilepticus. Ann Pharmacother. 1999;33:579–584.
178. Berkovic SF, Andermann F, Guberman A, et al. Valproate prevents
the recurrence of absence status. Neurology. 1989;39:1294–1297.
179. Fernandez-Torre JL. De novo absence status of late onset following
withdrawal of lorazepam: a case report. Seizure. 2001;10:433–437.
180. D’Agostino MD, Andermann F, Dubeau F, et al. Exceptionally
long absence status: multifactorial etiology, drug interactions and
complications. Epileptic Disord. 1999;1:229–232.
181. Nolan M, Bergazar M, Chu B, et al. Clinical and neurophysiologic
spectrum associated with atypical absence seizures in children
with intractable epilepsy. J Child Neurol. 2005;20:404–410.
182. Roger J, Lob H, Tassinari CA. Status epilepticus. In: Vinken P, Bruyn
G, eds. Handbook of Clinical Neurology, Vol 15, The Epilepsies.
Amsterdam: North Holland Publishing Co.; 1974:145–188.
183. Cortez MA, McKerlie C, Snead OC. A model of atypical absence
seizures. EEG, pharmacology, and developmental characteriza-
tion. Neurology. 2001;56:341–349.
184. Holmes GL, McKeever M, Adamson M. Absence seizures in chil-
dren: clinical and electroencephalographic features. Ann Neurol.
1987;21:268–273.
185. Matsumoto A, Kumagai T, Miura K, et al. Epilepsy in Angelman
syndrome associated with chromosome 15q deletion. Epilepsia.
1992;33:1083–1090.
186. Gastaut H, Roger J, Ouahchi S, et al. An electroclinical study of
generalized epileptic seizures of tonic expression. Epilepsia. 1963;4:
15–44.
187. Perez Velazquez JL, Huo JZ, Dominguez LG, et al. Typical versus
atypical absence seizures: network mechanisms of the spread of
paroxysms. Epilepsia. 2007;48:1585–1593.
188. Bancaud J, Talairach J, Morel P, et al. “Generalized” epileptic
seizures elicited by electrical stimulation of the frontal lobe in
man. Electroencephalogr Clin Neurophysiol. 1974;37:275–282.
189. Abend NS, Dlugos DJ. Nonconvulsive status epilepticus in a
pediatric intensive care unit. Pediatr Neurol. 2007;37:165–170.
190. Treiman DM, Walton NY, Kendrick C. A progressive sequence of
electroencephalographic changes during generalized convulsive
status epilepticus. Epilepsy Res. 1990;5:49–60.
191. Young GB, Jordan KG, Doig GS. An assessment of nonconvulsive
seizures in the intensive care unit using continuous EEG moni-
toring: an investigation of variables associated with mortality.
Neurology. 1996;47:83–89.
192. Barry E, Hauser W. Status epilepticus: the interaction of epilepsy
and acute brain disease. Neurology. 1993;43:1473–1478.
193. Krishnamurthy KB, Drislane FW. Depth of electroencephalo-
gram suppression and outcome in pentobarbital coma for refrac-
tory status epilepticus. Epilepsia. 1999;40:759–762.
194. Claassen J, Hirsch LJ, Emerson RG, et al. Continuous EEG mon-
itoring and midazolam infusion for refractory nonconvulsive sta-
tus epilepticus. Neurology. 2001;57:1036–1042.
195. So EL, Ruggles KH, Ahmann PA, et al. Clinical significance and
outcome of subclinical status epilepticus in adults. J Epilepsy.
1995;8:11–15.
196. Walker M. Status epilepticus: an evidence based guide. Br Med J.
2005;331:673–677.
197. Holtkamp M, Othman J, Buchheim K, et al. Diagnosis of psy-
chogenic nonepileptic status epilepticus in the emergency set-
ting. Neurology. 2006;66:1727–1729.
198. Wilner AN, Bream PR. Status epilepticus and pseudostatus
epilepticus. Seizure. 1993;2:257–260.
199. Howell SJ, Owen L, Chadwick DW. Pseudostatus epilepticus. Q J
Med. 1989;71:507–519.
200. Towne AR, Waterhouse EJ, Boggs JG, et al. Prevalence of noncon-
vulsive status epilepticus in comatose patients. Neurology.
2000;54:340–345.
201. Vespa PM, Nuwer MR, Nenov V, et al. Increased incidence and
impact of nonconvulsive and convulsive seizures after traumatic
brain injury as detected by continuous electroencephalographic
monitoring. J Neurosurg. 1999;91:750–760.
202. Privitera M, Hoffman M, Moore JL, et al. EEG detection of non-
tonic-clonic status epilepticus in patients with altered conscious-
ness. Epilepsy Res. 1994;18:155–166.
203. Claassen J, Mayer SA, Kowalski RG, et al. Detection of electro-
graphic seizures with continuous EEG monitoring in critically ill
patients. Neurology. 2004;62:1743–1748.
204. Jirsch J, Hirsch LJ. Nonconvulsive seizures: developing a rational
approach to the diagnosis and management in the critically ill
population. Clin Neurophysiol. 2007;118:1660–1670.
205. Vespa PM, O’Phelan K, Shah M, et al. Acute seizures after intrac-
erebral hemorrhage: a factor in progressive midline shift and
outcome. Neurology. 2003;60:1441–1446.
206. Claassen J, Jette N, Chum F, et al. Electrographic seizures and
periodic discharges after intracerebral hemorrhage. Neurology.
2007;69:1356–1365.
207. Pandian JD, Cascino GD, So EL, et al. Digital video-electroen-
cephalographic monitoring in the neurological–neurosurgical
intensive care unit: clinical features and outcome. Arch Neurol.
2004;61:1090–1094.
208. Jette N, Claassen J, Emerson RG, et al. Frequency and predictors
of nonconvulsive seizures during continuous electroencephalo-
graphic monitoring in critically ill children. Arch Neurol. 2006;
63:1750–1755.
209. Jirsch J, Hirsch LJ. Nonconvulsive status epilepticus in critically
ill and comatose patients in the intensive care unit. In: Kaplan
PW, Drislane FW, eds. Nonconvulsive Status Epilepticus. New
York, NY: Demos Medical Publishing; 2009:175–186.

210. Nei M, Lee JM, Shanker VL, et al. The EEG and prognosis in sta-
tus epilepticus. Epilepsia. 1999;40:157–163.
211. Hirsch LJ, Brenner RP, Drislane FW, et al. The ACNS subcommit-
tee on research terminology for continuous EEG monitoring: pro-
posed standardized terminology for rhythmic and periodic EEG
patterns encountered in critically ill patients. J Clin Neurophysiol.
2005;22:128–135.
212. Scheuer ML. Continuous EEG monitoring in the intensive care
unit. Epilepsia. 2002;43(Suppl 3):114–127.
213. Brenner RP. Is it status? Epilepsia. 2002;43(Suppl 3):103–113.
214. Pohlmann-Eden B, Hoch DB, Cochius JI, et al. Periodic lateralized
epileptiform discharges—a critical review. J Clin Neurophysiol.
1996;13:519–530.
215. Reiher J, Rivest J, Grand’Maison F, et al. Periodic lateralized epilep-
tiform discharges with transitional rhythmic discharges: associa-
tion with seizures. Electroencephalogr Clin Neurophysiol. 1991;78:
12–17.
216. Brenner RP, Schaul N. Periodic EEG patterns: classification, clin-
ical correlation, and pathophysiology. J Clin Neurophysiol. 1990;7:
249–267.
217. Treiman DM. Status epilepticus. Baillieres Clin Neurol. 1996;5:
821–839.
218. Krumholz A. Epidemiology and evidence for morbidity of noncon-
vulsive status epilepticus. J Clin Neurophysiol. 1999;16:314–322.
219. Husain AM, Mebust KA, Radtke RA. Generalized periodic
epileptiform discharges: etiologies, relationship to status epilep-
ticus, and prognosis. J Clin Neurophysiol. 1999;16:51–58.
220. Hirsch LJ, Claassen J, Mayer SA, et al. Stimulus-induced rhyth-
mic, periodic, or ictal discharges (SIRPIDs): a common EEG
phenomenon in the critically ill. Epilepsia. 2004;45:109–123.
221. Hirsch LJ, Pang T, Claassen J, et al. Focal motor seizures induced
by altering stimuli in critically ill patients. Epilepsia. 2008;49:
968–973.
222. Fisch BJ, Klass DW. The diagnostic specificity of triphasic wave
patterns. Electroencephalogr Clin Neurophysiol. 1988;70:1–8.
223. Sheridan PH, Sato S. Triphasic waves of metabolic encephalopathy
versus spike-wave stupor. J Neurol Neurosurg Psychiatry. 1986;49:
108–109.
224. Boulanger JM, Deacon C, Lecuyer D, et al. Triphasic waves versus
nonconvulsive status epilepticus: EEG distinction. Can J Neurol
Sci. 2006;33:175–180.
225. Jordan KG. Continuous EEG and evoked potential monitoring in
the neuroscience intensive care unit. J Clin Neurophysiol. 1993;
10:445–475.
226. Scheuer ML, Wilson SB. Data analysis for continuous EEG mon-
itoring in the ICU: seeing the forest and the trees. J Clin
Neurophysiol. 2004;21:353–378.
227. Agarwal R, Gotman J. Long-term EEG compression for inten-
sive-care settings. IEEE Eng Med Biol Mag. 2001;20:23–29.
228. Gotman J. Automatic seizure detection: improvements and eval-
uation. Electroencephalogr Clin Neurophysiol. 1990;76:317–324.
229. Gotman J. Automatic detection of seizures and spikes. J Clin
Neurophysiol. 1999;16:130–140.
Chapter 29 ■ Nonconvulsive Status Epilepticus 643
230. Wilson SB. Algorithm architectures for patient dependent seizure
detection. Clin Neuophysiol. 2006;117:1204–1216.
231. Wilson SB. A neural network method for automatic and incre-
mental learning applied to patient-dependent seizure detection.
Clin Neurophysiol. 2005;116:1785–1795.
232. Qu H, Gotman J. Improvement in seizure detection performance
by automatic adaptation to the EEG of each patient.
Electroencephalogr Clin Neurophysiol. 1993;86:79–87.
233. Abend NS, Dlugos D, Herman S. Neonatal seizure detection
using multichannel display of envelope trend. Epilepsia. 2008;
49:349–352.
234. Hirsch LJ. Continuous EEG monitoring in the intensive care
unit: an overview. J Clin Neurophysiol. 2004;21:332–340.
235. Bleck TP. Intensive care unit management of patients with status
epilepticus. Epilepsia. 2007;48(Suppl 8);59–60.
236. Walker M, Cross H, Smith S, et al. Nonconvulsive status epilepti-
cus: epilepsy research foundation workshop reports. Epileptic
Disord. 2005;7:253–296.
237. Litt B, Wityk RJ, Hertz SH, et al. Nonconvulsive status epilepticus
in the critically ill elderly. Epilepsia. 1998;39:1194–1202.
238. Krishnamurthy KB, Drislane FW. Relapse and survival after bar-
biturate anesthetic treatment of refractory status epilepticus.
Epilepsia. 1996;37:863–867.
239. Walker MC. Diagnosis and treatment of nonconvulsive status
epilepticus. CNS Drugs. 2001;15:931–939.
240. Claassen J, Hirsch LJ, Emerson RG, et al. Treatment of refractory
status epilepticus with pentobarbital, propofol, or midazolam: a
systematic review. Epilepsia. 2002;43:146–153.
241. Drislane FW, Lopez MR, Blum AS, et al. Detection and treatment
of refractory status epilepticus in the intensive care unit. J Clin
Neurophysiol. 2008;25:181–186.
242. Drislane FW. Evidence against permanent neurologic damage
from nonconvulsive status epilepticus. J Clin Neurophysiol. 1999;16:
323–331.
243. Lee SI. Nonconvulsive status epilepticus. Ictal confusion in later
life. Arch Neurol. 1985;42:778–781.
244. Scholtes FB, Renier WO, Meinardi H. Non-convulsive status
epilepticus: causes, treatment, and outcome in 65 patients. J Neurol
Neurosurg Psychiatry. 1996;61:93–95.
245. Adachi N, Kanemoto K, Muramatsu R, et al. Intellectual
prognosis of status epilepticus in adult patients: analysis with
Wechsler Adult Intelligence Scale-Revised. Epilepsia. 2005;46:
1502–1509.
246. Nevander G, Ingvar M, Auer R, et al. Status epilepticus in well-
oxygenated rats causes neuronal necrosis. Ann Neurol. 1985;18:
281–290.
247. Vespa PM, Miller C, McArthur D, et al. Nonconvulsive electro-
graphic seizures after traumatic brain injury result in a delayed,
prolonged increase in intracranial pressure and metabolic crisis.
Crit Care Med. 2007;35:2830–2836.
248. Ruegg SJ, Dichter MA. Diagnosis and treatment of nonconvul-
sive status epilepticus in an intensive care unit setting. Curr Treat
Options Neurol. 2003;5:93–110.

Anticipating Seizures Based on EEG
FERNANDO H. LOPES DA SILVA AND STILIYAN KALITZIN
INTRODUCTION
In this chapter, we deal with the possibility of anticipating
seizures based on the analysis of electroencephalogram (EEG)
recordings. Nonetheless it is interesting to note that there exist
many reports stating that epileptic patients may feel an
approaching seizure (1), but most of these are based on rela-
tively small patient populations. It is relevant to assess this issue
critically since it may be important in order to determine the
value of any automatic method based on EEG recordings.
Schulze-Bonhage et al. (2) carried out a comprehensive analysis
of subjective seizure anticipation reported by 500 consecutively
recruited outpatients with focal and generalized epilepsy. The
patients were asked to report any premonitory symptoms, with
respect to, at least 30 minutes prior to a seizure. No auras were
included. These authors found that 6.2% of patients were able
to report anticipating seizures. The median estimated time
interval between occurrence of premonitory symptoms and
seizure onset was 90 minutes. More recently, Scaramelli et al.
(3) reported a study where the existence of premonitory or pro-
dromal signs (PS) was investigated in a randomly selected pop-
ulation of 100 adult epileptic patients both with focal and
generalized epilepsies. Using a semistructured protocol and
personal interviews they found evidence for PS (behavioral,
cognitive and mood changes, excluding auras) in 39% of
patients, mostly in those having complex partial and general-
ized tonic-clonic seizures. In this study PS were reported to
have an insidious onset and a duration ranging from 30 min-
utes to several hours prior to seizure onset. We should note that
these results are certainly relevant although it is not yet possible
to estimate the corresponding sensitivity and specificity with
respect to the pre-ictal period. The fact that no EEG recordings
were generally carried out during the period where the patients
reported PS raises also the question as to whether EEG signal
changes might be associated with the presence of PS. In assess-
ing the clinical significance of EEG automatic methods of
analysis, it is important to keep these results on subjective expe-
riences of epileptic patients in mind.
In their thoughtful account of the “state of seizure predic-
tion,” Zaveri et al. (4) considered that three phases can be distin-
guished in this research field in the course of the last decades. In
the first phase in the seventies, a variety of attempts were tried,
from measuring epileptiform spike rates to estimating time-
varying EEG spectral features. In the second phase, particularly
in the late eighties and beginning nineties, the question of antic-
ipating epileptic seizures was approached within the frame of
mathematical methods derived from the theory of nonlinear
CHAPTER
30
dynamical systems. In the third phase, since about 2000, more
sophisticated studies are appearing particularly with respect to
theoretical models of seizure dynamics, to statistical methods of
evaluating clinical application studies, and to proposals of new
strategies. In any case the development of reliable methods to
anticipate seizures is the aim of a wide range of researchers hav-
ing in view the desideratum of being able to control the occur-
rence of seizures, ideally avoiding their onset.
Considering the relevance of the mathematical approaches,
however, we start by presenting a sketch of the analytical tools
used in this context. Before entering into the main question of
anticipation of epileptic seizures, we consider how these mathe-
matical analyses have been applied to characterize the dynamics
of the neuronal networks during seizures. This is useful to under-
stand better what happens in the dynamics of the corresponding
EEG signals in the course of the transition from the interictal to
the ictal or seizure state. We finish this chapter with the discus-
sion of alternative strategies specially in relation to theoretical
modeling of seizure dynamics. In this context, it is interesting to
note that the question of predicting seizures is related to similar
problems in other fields of science. Recently, Scheffer et al. (5)
reviewed this issue in a general way and concluded that there may
exist generic early-warning signals that may indicate for a wide
class of systems with multiple attractors if a critical threshold is
approaching, such as in the case of an approaching epileptic
seizure. The authors assume in this chapter, however, that transi-
tions occur due to crossing that threshold. In reality internal or
external fluctuations, or “noise,” can trigger a transition even
away from the critical threshold in a multi-attractor system.
Another aspect that should be noted is that in order to define
effectively a precursor, or early-warning signal, it is necessary to
have a specific model of critical transitions.
CHARACTERIZATION OF EEG SIGNALS
DURING EPILEPTIC SEIZURES AS
REALIZATIONS OF LOW-DIMENSION
NONLINEAR COMPLEX SYSTEMS
In the last decade, the interest for the characterization of epilep-
tic seizures based on EEG recordings using analytical tools
derived from the mathematical theory of nonlinear complex
systems has grown steadily. A large number of approaches have
been tried out with this aim in view.
A very abbreviated explanation of the corresponding basic
concepts is appropriate here. Traditionally EEG analysis employs
mathematical tools used in the analysis of linear systems as
645
646 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
discussed in detail in Chapter 54. More recently nonlinear meth-
ods of analysis of time series have received special attention, par-
ticularly since Takens (6) proposed the mathematical basis for
the reconstruction of dynamical systems using time series and
Grassberger and Procaccia (7,8) introduced methods to estimate
invariants of such systems from a single time series, such as the
correlation dimension (D2), Lyapunov exponents, and entropy (for
mathematical definitions see, for example, Ref. 9). The most
important aspects of nonlinear time series analysis are described
in Grassberger et al. (10), and a tutorial review of nonlinear
dynamical EEG analysis is that of Pritchard and Duke (9). The
basic principle is to reconstruct the state space of the system
using delay-vectors derived from the EEG time series. A funda-
mental concept is that the dynamics of a system must be studied
in a phase space; a point in this space characterizes the state of
the system at any moment of time. The phase space is a recon-
struction procedure that starts from raw data and builds vectors
by iterations of a time delay (delay-vectors). In this way invari-
ants of the signal dynamics, namely the correlation integral and
the correlation dimension D2, or the Lyapunov exponents can be
estimated to characterize the signal dynamics.
Given an EEG signal one can put in evidence nonlinear
deterministic dynamics, by testing whether the null hypothesis
that the signal is a realization of a Gaussian random process can
be rejected. This can be tested in various ways; one of the most
common tests consists in making a comparison of the invari-
ants of the signal, such as the correlation dimension D2 or the
Lyapunov exponents, with those of surrogate signals (random-
ized signals) as proposed by Theiler et al. (11), Pijn et al. (12),
Prichard and Theiler (13), and Schreiber and Schmitz (14). This
methodology has been applied by several authors to the analy-
sis of EEG signals recorded during different behavioral states,
including sleep (15) and alpha rhythms (16), by computing the
correlation dimension (D2) of EEG epochs and the correspon-
ding surrogates.
With respect to the study of EEG signals recorded from the
brain of epileptic patients, the basic question is whether differ-
ences in the dynamics of neuronal networks, in terms of invari-
ants such as the correlation dimension (D2) and/or Lyapunov
exponents, could be found in these signals during the period
preceding an epileptic seizure, with respect to reference periods
far from seizure occurrence.
This concept stems from previous studies where it was
shown that EEG signals recorded during epileptic seizures have
different properties in terms of nonlinear dynamics compared
to normal ongoing EEG activity. Hallmarks of such studies were
those of Babloyanz and Destexhe (17) who estimated nonlinear
dynamical parameters of absence seizures, namely the correla-
tion dimension and the largest Lyapunov exponent, and
reported a considerably lower value for the seizure periods in
comparison with normal EEG, what these authors interpreted
as indicating that the EEG during absence seizures has low-
dimensional chaotic nature. These authors claimed the “exis-
tence of a chaotic attractor” during the absence seizure.
Although this claim has been rejected by controlled studies per-
formed later, this study has set the tone for many later investi-
gations. Along the same line Iasemidis et al. (18–20) described
a sudden decrease in the value of the largest Lyapunov expo-
nent, an indication of deterministic chaos, at seizure onset in
subdural electrocorticogram (ECoG) recordings in seizures
with a temporal lobe epilepsy (TLE) onset, initially in those
channels close to the epileptogenic focal area, followed by other
channels. In an animal model of epilepsy (limbic kindling in
the rat) with a focal onset in the hippocampus, Pijn et al. (12)
demonstrated a very clear decrease of the value of the correla-
tion dimension D2, occurring at seizure onset. The drop in the
value of D2 was most pronounced in the “focal area.” The
process of seizure spreading could be followed by showing the
progressive decrease of D2 for signals recorded from various
brain areas at increasing distances from the “focal area.”
In this experimental animal study (12) and in a subsequent
study in patients with temporal lobe epilepsy (21,22) we found
that most EEG signals recorded during interictal periods could
not be distinguished from random time series, while most
epochs recorded during seizures were clearly distinguishable
from random noise. These results demonstrate that EEG signals
during epileptic seizures have different dynamics with respect
to the interictal EEGs and are characterized by low-correlation
dimension. This implies that the dynamical state of the neu-
ronal networks underlying an epileptic seizure can be described
by a relatively small number of degrees of freedom, in contrast
with the interictal EEG. Whether this corresponds to a state
characterized by a chaotic attractor cannot be concluded with-
out further evidence.
The estimation of correlation dimension D2 in the cases
reported by Pijn et al. (21) yields excellent results in terms of
delineating the epileptogenic zone in cases of TLE, as well as
tracing ictal spread throughout the brain. These studies, how-
ever, did not reveal any value of D2 able to predict reliably the
onslaught of a complex partial seizure. In fact these findings
indicate that a behavior reminiscent of a low-dimensional
chaotic attractor becomes manifest only well into the seizure.
In any case we should stress that these kind of measurements
can be influenced by a number of factors that have to be taken
into account in interpreting the results, namely the following:
the activity of the reference electrode may be of considerable
influence on estimates of D2; bipolar montage reconstruction is
the method of choice, whereas in case of monopolar recordings,
one should always consider the possibility that extraneous
“noisy” contamination through the reference electrode may
influence the estimation of D2. In addition the presence of
epileptiform spikes in the EEG signals, recorded interictally or
pre-ictally, may yield a very low D2, close to 0. EEG signals dis-
playing seizure activity always yielded D2 values different from
surrogate signals. The differences, however, varied from being
subtle to very large. Indeed there is usually a marked degree of
nonstationarity in most EEG signals during the course of a
seizure. Therefore, it is inappropriate to allocate a single value
for D2 for a whole seizure and thus only an analysis that takes
into account the evolution of D2 in the course of time might
reveal a correct picture. Another limiting factor in the nonlin-
ear EEG analysis is related to the nature of the acquisition tech-
nique. EEG signals are amplified and band-pass filtered during
recording, and this may influence the values of calculated D2 or

the Laypunov exponents. As those calculations are based on
defining neighborhoods in the reconstructed phase space of the
system, excluding low-frequency variations of the signals may
create spurious “closeness” of states and vice versa, taking a
broader band signal may separate states in the phase space of
the system.
Although many investigators have considered the possibility
that dynamical processes in the brain considered in general may
be of a low-dimensional chaotic nature (17,23,24), such has
never reliably been clearly proven above any doubt for “nor-
mal” ongoing EEG activity recorded from the scalp.
ANTICIPATING EPILEPTIC SEIZURES USING
SPONTANEOUS EEG SIGNALS
First we should have an operational definition of what is meant
by anticipating or predicting epileptic seizures. Predictability may
be defined operationally as “the demonstration that there exists
a statistically significant association between the outcomes of a
set of measurements preceding a seizure, and the time to the first
seizure following these measurements.” The main point that has
to be added is what kind of measurement is appropriate for the
required association. Here we pass in review a number of meas-
ures that have been proposed and explored.
An early and simple idea was to use as a measure the rate of
occurrence of epileptiform spikes, that is, to check whether
there exist any consistent association between variations in the
rate of occurrence of epileptiform spikes and the time of seizure
onset. While there were some early reports indicating a possible
relationship, a well controlled study by Gotman and Marciani
(25) in 44 patients with focal epilepsy clarified this issue. They
monitored continuously EEG and behavior of these patients,
quantified spiking rate by an automatic detection method, and
demonstrated that spiking rates do not change before seizures
but may increase markedly after them. Thus neither high nor
low spiking rates are associated with the occurrence of seizures,
and spike rating cannot be applied in order to predict seizure
onset. Other EEG statistical properties have been used with the
aim of detecting hidden information in the pre-ictal period that
might be used to predict the impending occurrence of a seizure.
Thus Viglione and Walsh (26) using pattern recognition ana-
lytic procedures of spectral data reported that EEG changes
characteristic of preseizure states could be detected a few sec-
onds before seizure onset. Using linear autoregressive models
fitted to ongoing EEG signals in 12 epileptic patients suffering
from “absences,” Rogowski et al. (27) showed that the location
of the poles in the z- and s-planes, as a function of time, pre-
sented a specific pattern linked with the occurrence of the
seizure. These authors noted further that the trajectory of the
“most mobile pole” during the preseizure period could aid in
the prediction of the seizure by several seconds. Later, Salant et
al. (28) used multivariate spectral estimation based on para-
metric modeling (determination of pole trajectories and coher-
ence functions) in order to detect EEG changes preceding the
outbreak of a seizure. Prediction of oncoming primary general-
ized seizures was based on the detection of increased pre-ictal
Chapter 30 ■ Anticipating Seizures Based on EEG 647
synchronization. Prediction times of 1 to 6 seconds were found
in several seizures from five patients. These methods, however,
were not applied in general in view of the short prediction win-
dow and the uncertainty about the precise moment of starting
of these seizures. These techniques were not tested thoroughly
in a general clinical setting.
Efforts to solve the problem of anticipation in a more gen-
eral context got a considerable impulse with the application of
mathematical concepts developed for the analysis of nonlinear
complex dynamical systems. Having passed above in review evi-
dence that during epileptic seizures EEG signals display low-
dimension correlation dimension values, the question is
whether this is also the case before seizure occurrence, that is, in
the pre-ictal period. In this respect memorable contributions
were those of Iasemidis et al. (18–20), Lehnertz and Elger (29),
and Martinerie et al. (30).
As indicated above, a seizure onset, in general, represents an
abrupt transition from a high-dimensional to a low-dimen-
sional state. This can be described as a phase transition in the
dynamics of the neuronal networks characteristic of the non-
linear nature of the underlying system. The question is thus
whether phase transitions occur already in the pre-ictal period?
Iasemidis et al. (18–20) analyzed long stretches of EEG sig-
nals, starting many hours before a seizure, recorded from sub-
dural and implanted electrodes in the hippocampus of TLE
patients. They reconstructed for each EEG epoch the phase
space and estimated the corresponding maximum Lyapunov
exponent (Lmax). The values of Lmax of EEG epochs recorded
before the seizure showed fluctuations over time. Noteworthy,
the Lmax values of EEG signals recorded from different cortical
sites started to converge (phase of entrainment) several minutes
before a seizure, and seconds prior to a seizure these values were
locked together and showed an abrupt transition to a more
ordered state. These authors noted that a critical number of
sites should display entrainment over a certain time period for
a seizure to occur. In this way these authors consider that this
analysis methodology allows the definition of a pre-ictal transi-
tion state and thus it may be used to detect an impending
seizure. Based on these data Iasemidis et al. (31) proposed an
adaptive seizure prediction algorithm to analyze continuous
EEG signals for the prediction of TLE seizures.
Lehnertz and Elger (29), applied the concept of EEG signal
“complexity,” defined as the variability of a set of consecutive
D2 estimates, in order to determine whether seizure onset could
be anticipated. They found that “complexity” was reduced in
EEG epochs recorded close to seizure onset and proposed that
“neuronal complexity loss” could be used to anticipate seizure
occurrence.
Martinerie et al. (30) analyzed 19 seizures from a homoge-
neous group of 11 patients with MTLE associated with
hippocampal sclerosis, using indwelling electrodes, by recon-
structing intracranial EEG signals as trajectories in phase space.
They introduced nonlinear indicators to characterize the signal
dynamics taking into account the EEG signals both as functions
of time and space. In this way they were able to pinpoint the
brain sites where seizure apparently started and showed that in
most cases seizure onset could be anticipated 2 to 6 minutes
648 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Figure 30.1 Nonlinear analysis of an intracranial EEG signal showing
changes in the “similarity index” computed between a reference period
and successive time windows preceding the seizure (D). The statistical
significance is indicated by color code (blue indicates significance lev-
els up to 5 standard deviations). The anticipation time with respect to
seizure onset for EEG channel PC1 (Precuneus gyrus) is 13.5 minutes
in this case. Channels with and without pre-ictal changes are depicted
in red and blue, respectively, on a diagram showing the locations of
intracranial electrodes (E). The first channel where seizure onset is
detected is indicated by a yellow star. AT, anterior temporal lobe; PT,
posterior temporal lobe; PC, precuneus gyrus; IC, infracalcarine; SC,
supracalcarine. (From Navarro V, Martinerie J, Le Van Quyen M, et al.
Seizure anticipation in human neocortical partial epilepsy. Brain.
2002;125(Pt 3):640–655.)(See color insert)
(Fig. 30.1). Confronted with the difficulties of applying this
methodology in practice particularly because of the nonsta-
tionarity of EEG signals that may cause many spurious detec-
tions, this group developed alternative methods choosing,
instead of a continuous sliding EEG window, a time window far
from the seizure as reference and compared the dynamical
properties of successive epochs of EEG signals to those of the
reference window. In this way a new measure was introduced:
the “dynamical similarity index,” which quantified changes in
dynamics relative to the reference window. These authors
reported a decrease of the dynamical similarity index several
minutes before seizures in both intracranial (32–37) and even
scalp EEG recordings (38). Chávez et al. (39) analyzed changes
in phase synchrony and in the degree of association between
EEG signals recorded from different sites (using nonlinear cor-
relation coefficient) and found a significant decrease of syn-
chrony in the focal area several minutes before seizures ( 30
minutes), mainly in the frequency band of 10 to 25 Hz.
More recently the same group (40) introduced a new
method for analyzing multivariate phase synchronization in
EEG signals, which basically tracks multivariate phase synchro-
nization in space and time–frequency domains; these authors
called this method frequency flow analysis (FFA). This method
is particularly suited to analyze nonstationary time–frequency
dynamics, since it allows the estimation of continuous phase
differences as function of time. Using this approach, groups
of EEG signals can be identified as belonging to common
frequency flows, and the problem of studying synchronization
becomes the problem of tracking frequency flows. This method
was applied to simulated signals and real EEG or MEG signals
recorded before and during epileptic seizures (35); it appeared
to be relevant in tracking transient collective dynamics fluctu-
ating in time, frequency, and space.
The reader may be overwhelmed by the large amount of
methods that we have passed in review in this short chapter.
This considerable series of methods is the result of the very
inventive work of several research groups pursuing the still eva-
sive objective of developing “the” optimal EEG analysis method
with respect to detecting relevant features of EEG signals in the
pre-ictal phase. An excellent critical description of the various
methods used in this field is given by Mormann et al. (41).
Recently Wendling et al. (42) made a critical comparison of
the performance of three families of methods regarding their
capacity to assess the degree of coupling between EEG signals,
namely linear and nonlinear regression, phase synchronization,
and generalized synchronization. This was evaluated using a
model where the relationships (coupling parameters) between
signals were precisely known. The main result was that there
was no “ideal” method, that is, none of the methods performed
better than all the other in all conditions. They pointed out,
however, that linear and nonlinear regression methods showed
sensitivity to the coupling parameter in all tested models with
average or good performances. Accordingly they concluded that
the strategy of choice with respect to the analysis of coupling of
EEG signals in epileptic brains is to first apply the regression
methods in order to characterize functional connectivity before
using more sophisticated methods that require more stringent
assumptions. We concur with this advice and stress that a prac-
tical strategy is to use more than one method applied to the
same EEG data, including always linear and nonlinear regres-
sion (21,43).
ASSESSING THE PERFORMANCE
OF ALGORITHMS IN ANTICIPATING
EPILEPTIC SEIZURES
The early studies were followed by a series of investigations
where researchers tried a variety of modifications of the
approaches described above and proposed additional algo-
rithms like the “accumulated energy” algorithm, based on wide
frequency band spectral energy (44), and presented critical
studies with the aim of evaluating the results obtained by means
of different methods in a rigorous way. In the context of a crit-
ical evaluation of the performance of the “dynamical similarity
index” Navarro et al. (37) analyzed the dynamics of long EEG
epochs recorded from patients with indwelling electrodes
undergoing monitoring for presurgical evaluation of refractory
partial epilepsy at the Montreal Neurological Institute. The
similarity index detected pre-ictal changes considered relevant
in two third of the seizures, these detections preceded a seizure
on average by about 12 minutes, but the index could display
fluctuations that were not directly followed by seizures. A large
number of these fluctuations were related to changes of vigilance

or behavior. These authors carefully note that the dynamic
changes detected pre-ictally may “represent physiologic changes
acting as facilitating factors or pathologic changes reflecting a
network dysfunction.”
An important assessment study in this context is that of
Aschenbrenner-Scheibe et al. (45). These authors investigated
the sensitivity and specificity of Lehnertz and Elger’s method
(29,46,47) based on estimating a “dimension complexity drop”
of sufficient amplitude and duration as indicative of the pre-
ictal state. In invasive 24-hour long EEG recordings from 21
patients with medically refractory partial epilepsies, who
underwent invasive presurgical monitoring, they found that
only one out of 88 seizures was preceded by a significant pre-
ictal dimension drop. They were not able to detect significant
differences of mean length and amplitude of dimension drops
between interictal and pre-ictal states. Furthermore they ana-
lyzed the results of the same algorithm applied to EEG signals
comprising 50 minutes of pre-ictal and 24 hours of interictal
periods. Using this data set they found that this algorithm
resulted in a sensitivity of 38% in hippocampal seizures and
33% in neocortical seizures, considering that a false-positive
rate of 0.1/hr was allowed. Regarding the clinical applicability
of this algorithm Aschenbrenner-Scheibe et al. (45) concluded
with the comment that “used as a pure warning system, a pre-
diction method of this quality would probably be ignored after
a short time.”
This study brought clear evidence that it is not sufficient to
study EEG signals during relatively short periods, namely 1
hour preceding a seizure in order to estimate the sensitivity and
specificity of automatic analysis methods. These estimates have
to be made on long-term EEG recordings lasting a large num-
ber of hours or preferably several days and nights including the
natural variations in wakefulness and sleep state (48,49). More
refined methods of assessing the statistical validity of algo-
rithms applied in this kind of data have been proposed (50,51).
Also with respect to other algorithms, subsequent tests
under controlled conditions in practice showed that results
reported earlier were difficult to reproduce (52). This was the
case for the results obtained using the “accumulated energy”
algorithm of Litt et al. (44) that were not reproduced by
Harrison et al. (53). Maiwald et al. (54) introduced a criterion
called the “seizure prediction characteristic” that incorporates
the assessment of sensitivity and false prediction rate to evalu-
ate the performance of seizure prediction algorithms (55).
Applying this criterion, three algorithms were evaluated on a
large EEG data set comprising seizures of 21 patients. The
“dynamical similarity index” (37) yielded 1 to 3.6 false predic-
tions per day and had a sensitivity between 21% and 42%; the
“accumulated energy” (44) yielded a sensitivity between 18%
and 31% for the extended, prospective version; and values
between 13% and 30% were found for the “effective correlation
dimension” (46,56). Performances at this level preclude clinical
applications in a general setting. Similarly the results obtained
using the Lyapunov exponent (18) could not be reproduced in
the study of Lai et al. (57) on account of finite-time statistical
fluctuations and noise. Furthermore Lai et al. (58) examined
more in detail the predictive power of Lyapunov exponents
Chapter 30 ■ Anticipating Seizures Based on EEG 649
using “control tests” with artificial signals and also applying this
method to real ECoG signals and concluded that the use of
Lyapunov exponents for seizure prediction is “practically
impossible as the brain dynamical system generating the ECoG
signals is more complicated than low-dimensional chaotic sys-
tems, and is noisy.”
This accumulation of control studies reporting a rather weak
performance of the initially proposed algorithms for seizure
prediction damped the expanding optimistic mood of this
research field into what Mormann et al. called (41) “the rise of
skepticism.”
It is recommendable that new algorithms in the field of
seizure prediction, possibly after a first stage to present a “proof-
of-principle,” must be tested in a large EEG data sample.
Although a “learning set” may initially be appropriate for algo-
rithm development, the results of interest must be obtained in a
“test set” recorded from nonselected patients undergoing EEG-
video monitoring in the course of preparation for brain surgery
which are recorded for many hours or days continuously.
Notwithstanding the difficulties in developing reliable algo-
rithms all studies support the idea that deterministic nonlinear
processes are involved in pre-ictal neural reorganization in
transition to seizure.
THE SEARCH FOR NOVEL BIOMARKERS
OF EPILEPTIC NETWORKS
In addition to the endeavor of applying analytical methods to
ongoing EEG signals in general, in the last decade special atten-
tion was given also to particular EEG features that appear to be
specifically associated with epilepsy, whether directly related to
seizure occurrence. This is the case of the fast oscillations that
were described in the hippocampus and entorhinal cortex of
rats with spontaneous seizures (59) that were called fast ripples
(FRs) because they have a higher frequency range (250 to 500
Hz) than the normal ripple oscillations (100 to 200 Hz)
described previously in normal brain (60). Similar FRs were
also encountered in the epileptic zones of patients with tempo-
ral lobe epilepsy (61). It should be noted that quite frequently
focal seizures begin with high-frequency oscillations (61,62).
The interpretation is that these FRs result from the activity of
clusters of interconnected neurons capable of sustaining syn-
chrony of discharges, surrounded by a strong inhibitory ring;
overcoming interneuron feedback inhibition may elicit the for-
mation of similar clusters in target areas, ultimately leading to
seizures (63).
Thus FRs could be considered biomarkers of epileptogenic
networks, although we should add that the fact that many of these
events are restricted to tiny volumes of brain tissue makes it more
difficult to detect FRs with standard electrophysiologic techniques
in the human brain. Jirsch et al. (64) showed that FRs can be
recorded from human epileptic brain using depth macroelec-
trodes and concluded that these “macro-FRs” reflect the partial
synchronization of very local oscillations as recorded by means of
microwires. Jacobs et al. (65) evaluated whether the occurrence of
FRs (250 to 500 Hz) showed any association with the pre-ictal
650 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
period in seven consecutive MTLE patients. Total rates of FRs
were significantly higher in the seizure-onset zone than outside,
but they did not change in a systematic way in the pre-ictal period.
Thus we may conclude that the occurrence and rate of FRs does
not appear to be associated with pre-ictal EEG periods.
THEORETICAL AND EXPERIMENTAL
MODELS OF THE TRANSITION BETWEEN
INTERICTAL AND SEIZURE STATES
A basic problem subjacent to the difficulties in developing reli-
able methods of anticipating seizures is that, at present, rela-
tively little is known about the dynamics of the neuronal
processes leading to seizures. Nonetheless some theoretical con-
cepts have been formulated that are essential in order to gain
insight into these processes, as well as to put in perspective
experimental findings in animals and the human.
A discussion about general principles of brain dynamics is rel-
evant in this context, since it is difficult to develop and interpret
EEG data obtained in epileptic patients without having a general
theoretical framework of the dynamics of brain activity that may
account both for normal and epileptic neuronal activities. In this
context, we assume that in the epileptic brain, some neuronal
networks possess an abnormal set of control parameters that can
produce different kinds of parallel dynamical states. In other
words, these networks may have “bi”- or “multi”-stable proper-
ties. This means that, in addition to a normal steady state, they
also have an abnormal one characterized by widespread synchro-
nous activity and that the transition between these two states may
occur more or less abruptly. This accounts for the two main char-
acteristics of epilepsy: (i) that an epileptic brain can function
apparently normally between seizures, that is, during the interic-
tal state; and (ii) that the seizures occur in a paroxysmal way,
thereby impairing brain functioning to a lesser or greater extent.
In this sense, epileptic disorders may be considered special cases
of the large class of dynamical diseases, meaning those patho-
physiologic states that can be characterized by intermittent
occurrence of abnormal dynamics, a theoretical concept pro-
posed by Glass and Mackey (66), and Bélair et al. (67), which we
(68–70) and others (71) have used in the context of epilepsy.
The theory of nonlinear dynamics offers the possibility to
understand, in formal terms, how the occurrence of the manifes-
tations of dynamical diseases takes place. In the case of epilepsy,
the basic question is how changes in the dynamics of a neuronal
network may occur such that paroxysmal widespread synchro-
nous oscillations abruptly emerge. It is often difficult to study
these changes in the human brain due to the inherent limita-
tions of obtaining detailed neuronal data from recordings in the
human brain. Therefore it is important to investigate these phe-
nomena in experimental animals that provide reliable models of
some types of epilepsy and in appropriate computer models.
In general, we can distinguish (69,70) two basic models that
can account for transitions to an epileptic seizure. Above we have
already presented some of the properties of complex nonlinear
dynamical systems, but a further element should be put forward.
It is a fact that these systems possess attractors, that is, sets of
values to which a dynamical system settles some time after being
perturbed or in mathematical terminology an asymptotically sta-
ble manifold in the phase space of the system. Topologically, an
attractor can be a point, a curve-like a limit cycle, a manifold, or
even a complicated set with a fractal structure known as a strange
attractor; the latter is typical of a chaotic dynamical system. With
respect to epilepsy in general we proposed (69,70) the following
general scheme:
Model 1: Bifurcation Model
“Bi”- or “multi”-stable systems are systems where jumps
between two or more coexisting attractors can take place,
caused by stochastic fluctuations, such as noise, of any input. In
this case the transition between interictal state and a seizure
may occur at random due to some fluctuations in input condi-
tions or in some system’s parameters, or in other words seizures
can be generated autonomously due to internal instability caus-
ing intermittent behavior (72) (Fig. 30.2).
Model 2: Deformation Model
Systems with deformable parameters such that the dynamics
may evolve from one attractor that represents the normal or the
interictal state to another attractor or to a series of attractors.
The parametric deformation may be caused by the fact that
some parameters may be, at a biologic level, unstable or may be
sensitive to some endogenous modulating influences, as for
example, chemical neurotransmitters, hormones, and composi-
tion of extracellular medium (pH), or may be induced by an
external stimulus, for example, sensory stimulus in reflex
epilepsies. An interesting observation in this respect was pre-
sented by Baumgartner et al. (73) who performed pre-ictal
SPECT scans fortuitously while video-EEG monitoring was
being carried out in two TLE patients. These authors found in
pre-ictal SPECT a significant increase in rCBF in the epileptic
temporal lobe, which was not associated with significant
Figure 30.2 Phase portraits of the thalamo-cortical system in two
cases: normal and epileptic brain. Sample trajectories (thin lines) from
normal (left) and epileptic model (right) projected onto a two-dimen-
sional slice of state space. The axes are values of two state variables:
cortical excitatory and inhibitory activity. The thick line is a separatrix
separating two attractors. (From Lopes da Silva FH, Blanes W, Kalitzin
SN, et al. Dynamical diseases of brain systems: different routes to
epileptic seizures. IEEE Trans Biomed Eng. 2003;50(5):540–548 and
Lytton WW. Computer modelling of epilepsy. Nat Rev Neurosci.
2008;9(8):626–637.)

changes of the ongoing EEG. They interpreted this finding as
indicating that rCBF changes observed on peri-ictal SPECT
may reflect a change in neuronal activity precipitating the tran-
sition from the interictal to the ictal state. Also a study by
Federico et al. (74) attempted to determine whether changes in
fMRI (BOLD signal) could be detected in the preseizure period.
The results in three patients indeed showed changes in pre-ictal
BOLD signal before seizure onset but these were rather com-
plex and not easy to interpret.
In addition we can also consider systems displaying a combi-
nation of these two processes, that is, scenarios where the system
“deforms” from single attractor to multistable dynamics (bifur-
cation). One feature of both the deformation and bifurcation
models is that they are not, strictly speaking, autonomous. They
need external input or endogenous noise in order to allow for
transitions. This is not, however, a major conceptual problem
(75). The living brain is hardly ever a closed dynamic system,
since it is always subjected to external influences. The so-called
spontaneous seizures, for example, simply indicate our lack of
knowledge about the precipitating factor or event.
Recently the possibility of a third class of models of
autonomous epileptic seizure generation has been investigated.
In these models the system possesses a feature called dynamic
intermittency (76–78). In its essence such systems are character-
ized by complex attractors and partially unstable dynamics.
Areas in the phase space (around the so-called saddle points)
can deflect the system into various types of behavior. In those
systems the transitions to epileptiform activity can occur pure
autonomously, without the help from any endogenous or exoge-
nous factor. Although intermittency has been suggested as a pos-
sible mechanism for various occurrences of seizure generation
(79), a realistic neuronal model of intermittency expressed in
terms of changes in neuronal dynamics is yet to be built.
An example of Model 1 is the occurrence of spike-and-wave
discharges (SWDs) in the thalamocortical system, typical of
absence seizures. This model is in agreement with experimental
observations carried out in order to understand how SWDs are
generated. Typical SWDs start and end abruptly. We found in a
rat genetic model of absences, the so called WAG/Rij rat (43)
that SWDs are initiated in the facial somatosensory cortex and
from there propagate to other cortical areas and to the thalamus.
This was confirmed in another experimental rat model of
absences, the Genetic Absence Epilepsy Rats of Strasbourg
(GAERS), where Polack et al. (80) found that the display of
SWDs in the local ECoG coincides with rhythmic membrane
depolarizations superimposed on a tonic hyperpolarization of
layer IV neurons, which occurs abruptly. Furthermore it was
demonstrated in a detailed microelectrophysiolgic study (81),
also in GAERS, that layer VI corticothalamic neurons act as “dri-
vers” in the generation of SWDs in the somatosensory thalamo-
cortical system. These observations reveal also that SWDs occur
abruptly in this animal model. A general conclusion that may be
drawn from these experimental studies is that the SWDs, char-
acteristic of absence seizures, tend to occur without a clear pro-
ictal electrophysiologic ECoGraphic pattern. In order to
understand in more general terms how this abrupt change of
neuronal dynamics takes place a computational model has been
Chapter 30 ■ Anticipating Seizures Based on EEG 651
constructed and proved to have not only face value but also pre-
dictive value. Without entering here in details (see for detailed
model Refs. 82–85) we may state that the model studies revealed
that (i) SWDs result from dynamical bifurcations that occur in
a bistable neuronal network; (ii) the durations of interictal
epochs have exponential distributions, indicating that transi-
tions between these two stable states occur randomly over time
with constant probabilities (see Ref. 86 for statistical distribu-
tions of seizures in epileptic patients); (iii) the probabilistic
nature of the onset of paroxysmal activity implies that it is not
possible to predict its occurrence; (iv) the bistable nature of the
dynamical system allows that an ictal state may be aborted by a
single counter-stimulus, even by a single stimulation pulse.
Kalitzin et al. (75) have developed this concept further including
continuously operating external modules designed to constantly
monitor and react to epileptic transitions.
An example of Model 2 is the occurrence of partial seizures
in MTLE. This model is in accordance with experimental stud-
ies. One of these in particular is relevant. This is the investigation
carried out by Cohen et al. (87) in hippocampal slices, in vitro,
that shed light into the basic question, namely whether in this
kind of epileptic seizures a pro-ictal state can be identified, in
contrast to what was observed with respect to absence seizures.
This study showed that before a full-blown seizure may be
detectable in local field potentials, there is a gradual change in
electrophysiological properties that becomes manifest as a
build-up of neuronal firing frequency variance. This implies that
a pro-ictal state with special properties can be identified at least
in this experimental condition. Thus these observations show
that the evolution of the excitability state of the neuronal popu-
lation in these hippocampal slices differs sharply from what was
seen in the case of SWDs in absence epilepsy. Assuming that
these in vitro observations are representative of the in vivo situ-
ation in patients with MTLE, and taking these findings together,
we may draw the conclusion that it should be possible, in prin-
ciple, to identify a pro-ictal state in this and similar cases.
Within the framework of these theoretical and experimental
models we can state that the transition to a seizure can occur
according to two scenarios: (i) as a sudden leap; in this case the
transition may be unpredictable as explained in detail by the
model studies shortly described above; (ii) as a continuous
sequence of changes in the dynamics of neuronal networks. In
this type of ictal transition, it is conceivable that the seizure may
be anticipated in its early, pre-ictal (preclinical) phase. Whether
these dynamical changes at the neuronal level may be detectable
at the ongoing EEG level is, of course, the crux of all the discus-
sions concerning the algorithms developed to anticipate
seizures based on the analysis of ongoing EEG signals, which we
discussed above. An important advance in this respect is the
model of Wendling et al. (88,89) that shows how the deforma-
tion of certain parameters can cause the neuronal network
dynamics to evolve though a succession of phases and ulti-
mately to display a full-blown seizure (Fig. 30.3).
An important clue to understand this basic question was
obtained when we investigated the process leading to an absence
seizure induced by intermittent photic stimulation (IPS) (90,91).
This investigation revealed that the phase clustering of harmonically
652 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Figure 30.3 Computational model (left-hand side) based on a lumped approach of a hippocampal network (based on
the representation of neuronal subpopulations of pyramidal neurons and two populations of interneurons) and a real
neuronal network (right-hand side, based on the explicit representation of cells and interconnections) approach.
Simulated activity (local field potentials) can be quantitatively compared to real activity using information processing
techniques. This quantitative analysis allows for identifying the parameters settings for which models best reproduce
real data. Performing this identification over a sliding time window can trace the evolution of model parameters, pro-
viding information into the mechanisms responsible for the transition between normal and seizure activity. (From
Wendling F. Computational models of epileptic activity: a bridge between observation and pathophysiological interpre-
tation. Expert Rev Neurother. 2008;8(6):889–896.)(See color insert)

related frequency components of a subject’s MEG/EEG responses
evoked by IPS increases significantly seconds before a seizure
occurs. This led to the introduction of a quantity—the phase clus-
tering index (PCI, see Appendix for definition)—by means of
which this enhancement can be quantified. We may argue that this
quantity reflects the degree of excitability of the underlying dynam-
ical system, and it can indicate the presence of nonlinear dynamics.
Applying this form of analysis we found that the patients who
develop seizures during IPS present an enhancement of the PCI at
the gamma frequency band, compared with that at the driving fre-
quency; this is why the index is called relative PCI or rPCI. An
important lesson to be drawn from this finding is that the change in
dynamics of EEG or MEG signals preceding the transition to the
seizure may be more readily detected by using an active probe, that
is, a stimulus to provoke a response of the system than by a passive
analysis of the ongoing signals. These observations led to test
whether the use of active probes could be also used in other forms
of epilepsy according to Model 2 scenario, such as in MTLE, as
described in the section “New Approaches in Anticipating Seizures.”
As a footnote we may add that it may be more appropriate
to use the term pro-ictal than to use the more common term
pre-ictal; indeed pre-ictal denotes just the activity immediately
preceding in time the seizure, which is, of course, always pres-
ent, whether or not it may be causally related to the transition
to a seizure; pro-ictal implies that the activity reflects a state of
excitability of the underlying neuronal networks that can lead
to a seizure. If the system is detected to be in a pro-ictal state
then an impending seizure may be anticipated. Here we must
add that such a state itself does not determine for 100% the
occurrence of the seizure, only the predisposition of the system
for a transition to an epileptic state.
NEW APPROACHES IN ANTICIPATING
SEIZURES: “ACTIVE” VERSUS
“PASSIVE” ALGORITHMS
The findings of photosensitive epilepsy described above
(90,91) prompted us to investigate whether rPCI of EEG sig-
nals recorded in patients with MTLE could also have predictive
value with respect to the anticipation of seizures. This implies
the application of a stimulus, since the estimation of rPCI is

based on the EEG activity modulated by an external “carrier”
signal. In the case of photosensitive epilepsy we used, intermit-
tent light. In the case of patients with MTLE we had to resort
to deep brain electrical stimulation using the available elec-
trodes that were implanted for long-term EEG monitoring and
recording seizures. Thus we developed an active electrical
direct-stimulation paradigm. With this objective we stimulated
patients with indwelling electrodes in the hippocampal forma-
tion using short, low-intensity (up to 1 mA) bursts of electri-
cal pulses, 5 seconds long, at the frequency of 20 Hz, repeated
at intervals of 20 seconds over long period of hours, and even
days at which time the intracranial EEG was sampled in paral-
lel. We call this stimulation paradigm a carrier frequency mod-
ulation probe (92). The relative phase clustering index (rPCI)
computed for all signals and all stimulated epochs of six
patients was found to be larger for electrode sites near to the
seizure-onset site. Even more interesting was the finding that it
was possible to forecast the probability of a seizure occurring
within a certain time based on the values of rPCI estimated at
a given moment in time (Fig. 30.4). For example in the patient
population analyzed in this study, a value of rPCI 0.6 pre-
dicts that a seizure will occur in less than 2 hours, with an
accuracy 80%. This study has still to be considered as a
“proof-of-principle” since no further controlled studies were
published so far. Nonetheless the application of “active probes”
opens a novel strategy in the field of seizure anticipation.
Figure 30.4 Results of the analysis of intracranial EEG signals from
temporal lobe epileptic patients. The relative Phase Clustering Index
(rPCI) was computed for several EEG channels of six patients at vari-
ous times before seizure onset. Upper image: median values and 95%
confidence intervals for the six patients are shown along the horizontal
axis; along the vertical axis the time interval to the next seizure is plot-
ted. Note that the values of rPCI increase with a decrease in the dura-
tion of this interval. Lower image: plot of the rPCI value (vertical axis)
against the prediction horizon, that is, the time interval to seizure
onset. In color the error rate of the prediction is indicated, from a low
error (dark blue) to a large error (red, brown). Note that for values of
rPCI 0.6 there is a probability of a seizure occurring in less than 2
hours with an accuracy of 80% . (From Kalitzin SN, Velis DN,
Suffczynski P, et al. Electrical brain-stimulation paradigm for estimating
the seizure onset site and the time to ictal transition in temporal lobe
epilepsy. Clin Neurophysiol. 2005;116(3):718–728.)(See color insert)
Chapter 30 ■ Anticipating Seizures Based on EEG 653
Although this method requires brain electrical stimulation, it
is not more complex than electrical stimulation paradigms
used for clinical purposes during intracranial EEG investiga-
tions to delineate epileptogenic areas and seizure-onset zones
by determining thresholds for eliciting afterdischarges (ADs)
or by triggering seizures. In fact our method was also success-
ful in identifying the seizure-onset zone as the location pro-
ducing increased rPCI values. This let us to conclude that the
rPCI quantity reflects the ability of the neuronal system to gen-
erate epileptic discharges. Remarkably, the measurement of
rPCI does not involve provoking those discharges, unlike other
stimulation-based techniques. To understand the mechanisms
that allow for such effectiveness we have resorted again to real-
istic computer models.
The emphasis on “active paradigms” for testing changes of
excitability of neuronal networks that according to either the
“deformation model” or the “combined model” will occur in
the pro-ictal period as the onset of a seizure approaches has also
been recently corroborated by simulations, using a computa-
tional model of seizure generation. Suffczynski et al. (84) inves-
tigated (i) how changes in model parameters that lead to a
transition from the normal ongoing EEG to an ictal pattern are
reflected in the properties of the simulated EEG output signals
and (ii) how the evolution of neuronal excitability towards
seizures can be reconstructed from EEG data using an “active”
paradigm, rather than the traditional “passive” algorithms that
use only ongoing EEG signals. The simulations indicate that an
“active” paradigm combined with appropriate analytical tools,
such as rPCI, yields more readily relevant information about
the change in excitability that characterizes the pre-ictal, or bet-
ter the pro-ictal, period than algorithms applied to the ongoing
EEG. These findings give strong support to the development
and application of “active” paradigms with the aim of charac-
terizing the change of excitability during the pro-ictal period
and thus anticipating the occurrence of a transition to an
epileptic seizure.
A quite different study gives support to the concept that
“active paradigms” may yield relevant information regarding
the characterization of the pre-ictal state. This is the investiga-
tion of Badawy et al. (93) who used TMS in patients with dif-
ferent forms of epilepsy (generalized, focal or focal with
secondary generalization) to determine the cortical motor
threshold and paired pulse recovery curves (at short, 2 to 15
msec, and long, 50 to 400 msec interstimulus intervals) at 24
hours before a seizure (also postseizure). They found
decreased motor threshold, increased intracortical facilitation,
and decreased intracortical inhibition, indicative of an increase
in cortical excitability and/or a decrease in cortical inhibitory
processes, at short and long interstimulus intervals, in the 24
hours before a seizure, and the opposite changes in the 24
hours after a seizure. These effects were bilateral in tonic-
clonic seizures in idiopathic generalized epilepsy and also in
secondarily generalized seizures in patients with focal epilepsy.
Similar changes were seen in the hemisphere ipsilateral to the
seizure focus in focal seizures that did not secondarily general-
ize. It appears that by means of TMS it is possible to detect
changes in cortical excitability preceding seizure by hours.
654 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Nonetheless these results are based on measurements of aver-
ages of the TMS responses obtained in experimental sessions
that lasted for 60 to 90 minutes; these measurements were not
followed along time as done by Kalitzin et al. (94), so it is not
known whether these changes covary with the progressive
approximation of a seizure. Therefore we can assume that
those findings corroborate our own findings and our hypoth-
esis of existence of a pro-ictal state that precedes in a nondeter-
ministic way the transition to epileptic seizure.
CONCLUDING REMARKS
The road toward the development of strategies aiming at a reli-
able detection of a pro-ictal state has been full of exciting the-
oretical challenges and cunning experimental approaches.
Despite the fact that the scientific community learned a lot
about the dynamics of neuronal networks in the context of
epileptic seizures in these last decades, we are still far from hav-
ing a generally accepted and robust strategy. A positive devel-
opment is that the community has set up a number of
Workshops on Seizure Prediction and a digital database that
can be used to develop and test suitable algorithms in a real
collaborative effort. The need of new strategies is becoming
even more pressing since the access to deep brain stimulation
techniques for a variety of brain diseases is now also open to
epileptic patients, particularly to those who are pharmacologi-
cally refractory and yet have either been rejected for a surgical
intervention or have had failed epilepsy surgery. Thus it would
be desirable to have reliable methods capable of seizure antic-
ipation so that a dedicated “closed-loop” stimulation proce-
dure might be developed to avoid the occurrence of seizures on
the basis of continuous and reliable analysis of the running
EEG record. The success of such procedures depends on the
possibility to either foresee an impending seizure or to detect
its occurrence at the earliest possible stage and only then to
apply a counter-stimulus in order to arrest it. A few “closed-
loop” approaches are starting to be explored, namely using
properties of the measured EEG signal to determine the exact
timing of the counter-stimulation (95) or generating an appro-
priate seizure suppression stimulation waveform (96). We may
refer to the latter as “state-reactive control.” This matter is dis-
cussed theoretically in more detail by Kalitzin et al. (75) who
give arguments to justify different scenarios for seizure control,
required according to different dynamical models of seizure
generation. Multistable and possibly some intermittent sys-
tems may be controlled by an acute intervention that aborts
the transition to the seizure and resets the system to its normal
state. Parametric deformation of the system’s attractor can be
compensated by continuously stimulating the system and
reducing the ictal attractor to a stable steady state. All of the
aforementioned approaches require indwelling systems that
deliver electrical current to selected targets in the brain.
Although a considerable body of evidence exists on the basis of
which intermittent high-frequency deep brain stimulation in
extrapyramidal disorders is deemed safe, the long-term effects
of chronic simulation administered to the human hippocam-
pus are still largely unknown. In a review of the experience in
nine chronically implanted patients with mesial temporal lobe
epilepsy who underwent almost continuous high-frequency
hippocampal electrical stimulation (97) no memory deficits
could be identified on neuropsychological testing with a fol-
low-up of up to 7 years.
This is a hard, though challenging field that requires the
convergence of a variety of scientific research approaches to
achieve results that may have practical usefulness in clinical
epilepsy.
APPENDIX
We introduce here the phase clustering index (PCI) (91) and
the associated relative PCI (rPCI) as a measure of the ability of
the system to phase-lock its response to an external stimulus.
We assume that the system is exposed to the same stimulus
at discrete time points ts H5t1, p , tN s6 , where N s is the number
of stimuli. If we denote the system’s response (this can be any
measured multidimensional physiologic signal such as EEG,
MEG, etc.) as Sc(t), c = 1, p , N c , N c indicating the number of
channels, then the stimulus-triggered response (STR) function
is defined as
Rsc(t) S c(t ts) (30.1)
In the particular case of periodic stimulations where all con-
secutive stimuli are separated by a constant period T ts 1
ts, s 1, p , N s, it is natural to select the response interval t H
[0,T) in Eq. (30.1). Our further definitions are not restricted to
this choice but the previous applications of PCI and the exam-
ples given below assume this simple nonredundant representa-
tion. The general theory is also perfectly applicable to general
nonperiodic stimuli. Below we also omit the channel index c for
notation simplicity.
Next we perform a discrete Fourier transform (30.1) along
the t-axis:
Z fs fft(Rs(t) (30.2)
where f denotes the discrete frequency associated with the com-
plex coefficients Z. In our applications we assume f H{1,2, p
N f}f0, where f0 is the fundamental frequency, typically the fre-
quency of the stimulation in case of periodic stimuli and N f is
the highest harmonic frequency considered. We will also omit
the f0 notation, unless otherwise stated, and we shall refer to the
frequency as an integer number.
Averaging either Eq. (30.1) or (30.2) over the stimuli yields
the classic averaged evoked response (or evoked potential if the
signal represents local field potentials). We introduce here the
definition of the PCI:
Zfs s PCIf Z fs s PCIf Af (30.3)
Here and later s means averaging over all periods (stimuli)
and denotes taking the magnitude of a complex number. In
essence the factorization in Eq. (30.3) represents the spectral-
evoked response as a product of an averaged amplitude spec-
trum and the PCI. In this way one can obtain one more degree
of information related to the evoked responses. Explicitly the
factors are

Z fs s 7. Grassberger P, Procaccia I. Measuring the strangeness of strange
PCIf ` `
Z fs s (30.4)
Af Z fs s
From Eqs. (30.3) and (30.4) it is clear that in Fourier repre-
sentation the STR is a product of two factors. One of the factors
represents the averaged spectral power of the response and the
second, the PCI, accounts for the phase of the response relative
to the stimulus.
We can easily see that the PCI is exactly equal to 1 if the
response complex amplitudes are of the same phase, irrespec-
tive of the values of the magnitudes. If on the contrary, the
phases of the individual responses are randomly scattered and
the magnitudes are equal, the PCI quantity is exactly 0.
We have shown previously (90,91) that a particular combi-
nation of the PCI spectrum, that we called rPCI, emerged as a
good identifier of an impending transition to seizures both in
patients with photosensitive epilepsy submitted to periodic
visual stimulation and in TLE patients suffering from sponta-
neous seizures (Fig. 30.4). We define the rPCI as
rPCI maxf (PCIf ) PCI1 (30.5)
In this definition the spectral maximum of PCI at frequency
f is compared to the phase consistency of the first Fourier com-
ponent or “the base driving.”
Our final remark in this section concerns the statistical signif-
icance test of the measured PCI values. Using the same magni-
tudes Af as defined in Eqs. (30.3) and (30.4) but generating a
large number of surrogate phases for each frequency component
of the response, we can compute a set of PCI values in the
absence of phase-locking to the stimulus. Then the 95th per-
centile PCI value defines the threshold for P 0.05 significant
PCI. This “surrogate phases test” will exclude falsely high PCI val-
ues that are generated by a single complex amplitude in which a
very high magnitude dominates the averaging in Eq. (30.4).
REFERENCES
1. Hughes J, Devinsky O, Feldmann E, et al. Premonitory symptoms
in epilepsy. Seizure. 1993;2:201–203.
2. Schulze-Bonhage A, Kurth C, Carius A, et al. Seizure anticipation
by patients with focal and generalized epilepsy: a multicentre
assessment of premonitory symptoms. Epilepsy Res. 2006;70(1):
83–88.
3. Scaramelli A, Braga P, Avellanal A, et al. Prodromal symptoms in
epileptic patients: clinical characterization of the pre-ictal phase.
Seizure. 2009;18(4):246–250.
4. Zaveri HP, Frei MG, Osorio I. State of seizure prediction: a report
on informal discussions with participation of the 3rd
International Workshop on seizure prediction. In: Schelter B,
Timmer J, Schulze-Bonhage A, eds. Seizure Prediction in Epilepsy.
Weinheim: John Wiley – VCH Verlag GmbH & Co. KGaA;
2008:325–330.
5. Scheffer M, Bascompte J, Brock WA, et al. Early-warning signals
for critical transitions. Nature. 2009;461(7260):53–59.
6. Takens F. Detecting strange attractors in turbulence. In: Rand DA,
Young LS, eds. Lecture Notes in Mathematics. Vol. 898. Berlin:
Springer-Verlag; 1981:366–381.
Chapter 30 ■ Anticipating Seizures Based on EEG 655
attractors. Physica D. 1983;9:189–208.
8. Grassberger P, Procaccia I. Characterization of strange attractors.
Phys Rev Lett. 1983;50(5):236–349.
9. Pritchard WS, Duke DW. Measuring chaos in the brain. A tutorial
review of nonlinear dynamical EEG analysis. Int J Neurosci.
1992;67:31–80.
10. Grassberger P, Schreiber T, Schaffrath C. Nonlinear time sequence
analysis. Int J Bifurc Chaos. 1991;1:521–547.
11. Theiler J, Eubank S, Longtin A, et al. Testing for nonlinearity in
time series: the method of surrogate data. Physica D. 1992;58:
77–94.
12. Pijn JP, van Neerven DN, Noest A, et al. Chaos or noise in EEG sig-
nals: dependence on state and brain site. Electroenceph Clin
Neurophysiol. 1991;79:371–381.
13. Prichard D, Theiler J. Generating surrogate data for time series
with several simultaneously measured variables. Phys Rev Lett.
1994;73(7):951–954.
14. Schreiber T, Schmitz A. Improved surrogate data for nonlinearity
tests. Phys Rev Lett 1996;77(4):635–638.
15. Achermann P, Hartmann R, Gunzinger A, et al. Correlation
dimension of the human sleep electroencephalogram: cyclic
changes in the course of the night. Eur J Neurosci. 1994;6(3):
497–500.
16. Stam CJ, Pijn JP, Suffczynski P, et al. Dynamics of the human alpha
rhythm: evidence for non-linearity? Clin Neurophysiol.
1999;110(10):1801–1813.
17. Babloyantz A, Destexhe A. Low-dimensional chaos in an instance
of epilepsy. Proc Natl Acad Sci USA. 1986;83(10):3513–3517.
18. Iasemidis LD, Sackellares JC, Zaveri HP, et al. Phase space topogra-
phy and the Lyapunov exponent of electrocorticograms in partial
seizures. Brain Topogr. 1990;2(3):187–201.
19. Iasemidis LD, Olson LD, Savit RS, et al. Time dependencies in the
occurrences of epileptic seizures. Epilepsy Res. 1994;17(1):81–94.
20. Iasemidis LD, Principe JC, Czaplewski JM, et al. Spatiotemporal
transitions to epileptic seizures: a nonlinear dynamical analysis
of scalp and intracranial EEG recordings. In: Lopes da Silva F,
Principe JC Almeida LB, eds. Spatiotemporal Models in
Biological and Artificial Systems. Amsterdam: IOS Press;
1996:81–88.
21. Pijn JP, Velis DN, van der Heyden MJ, et al. Nonlinear dynamics of
epileptic seizures on basis of intracranial EEG recordings. Brain
Topogr. 1997;9(4):249–270.
22. van der Heyden MJ, Velis DN, Hoekstra BP, et al. Non-linear analy-
sis of intracranial human EEG in temporal lobe epilepsy. Clin
Neurophysiol. 1999;110(10):1726–1740.
23. Skarda CA, Freeman WJ. How brains make chaos in order to make
sense of the world. Behav Brain Sci. 1987;10:161–195.
24. Freeman WJ. Neural networks and chaos. J Theor Biol.
1994;171(1):13–18.
25. Gotman J, Marciani MG. Electroencephalographic spiking activ-
ity, drug levels, and seizure occurrence in epileptic patients. Ann
Neurol. 1985;17(6):597–603.
26. Viglione SS, Walsh GO. Proceedings: epileptic seizure prediction.
Electroencephalogr Clin Neurophysiol. 1975;39(4):435–436.
27. Rogowski Z, Gath I, Bental E. On the prediction of epileptic
seizures. Biol Cybern. 1981;42:9–15.
28. Salant Y, Gath I, Henriksen O. Prediction of epileptic seizures from
two-channel EEG. Med Biol Eng Comput. 1998;36(5):549–556.
29. Lehnertz K, Elger CE. Spatio-temporal dynamics of the primary
epileptogenic area in temporal lobe epilepsy characterized by neu-
656 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
ronal complexity loss. Electroencephalogr Clin Neurophysiol.
1995;95(2):108–117.
30. Martinerie J, Adam C, Le Van Quyen M, et al. Epileptic seizures can
be anticipated by non-linear analysis. Nat Med. 1998;4(10):
1173–1176.
31. Iasemidis LD, Shiau DS, Chaovalitwongse W, et al. Adaptive
epileptic seizure prediction system. IEEE Trans Biomed Eng.
2003;50(5):616–627.
32. Le Van Quyen M, Martinerie J, Baulac M, et al. Anticipating
epileptic seizures in real time by a non-linear analysis of similarity
between EEG recordings. Neuroreport. 1999;10(10):2149–2155.
33. Le Van Quyen M, Adam C, Martinerie J, et al. Spatio-temporal
characterizations of non-linear changes in intracranial activities
prior to human temporal lobe seizures. Eur J Neurosci.
2000;12(6):2124–2134.
34. Le Van Quyen M. Anticipating epileptic seizures: from mathemat-
ics to clinical applications. C R Biol. 2005;328(2):187–198.
35. Le Van Quyen M, Amor F, Rudrauf D. Exploring the dynamics of
collective synchronizations in large ensembles of brain signals.
J Physiol Paris. 2006;100(4):194–200.
36. Navarro V, Martinerie J, Le Van Quyen M, et al. Seizure anticipa-
tion in human neocortical partial epilepsy. Brain. 2002;125
(Pt 3):640–655.
37. Navarro V, Martinerie J, Le Van Quyen M, et al. Seizure anticipa-
tion: do mathematical measures correlate with video-EEG evalua-
tion? Epilepsia. 2005;46(3):385–396.
38. Le Van Quyen M, Martinerie J, Navarro V, et al. Anticipation of
epileptic seizures from standard EEG recordings. Lancet.
2001;357(9251):183–188.
39. Chávez M, Le Van Quyen M, Navarro V, et al. Spatio-temporal
dynamics prior to neocortical seizures: amplitude versus phase
couplings. IEEE Trans Biomed Eng. 2003;50(5):571–583.
40. Rudrauf D, Douiri A, Kovach C, et al. Frequency flows and the
time-frequency dynamics of multivariate phase synchronization
in brain signals. Neuroimage. 2006;31(1):209–227.
41. Mormann F, Andrzejak RG, Elger CE, et al. Seizure prediction: the
long and winding road. Brain. 2007;130(Pt 2):314–333.
42. Wendling F, Ansari-Asl K, Bartolomei F, et al. From EEG signals to
brain connectivity: a model-based evaluation of interdependence
measures. J Neurosci Methods. 2009;183(1):9–18.
43. Meeren HK, Pijn JP, Van Luijtelaar EL, et al. Cortical focus drives
widespread corticothalamic networks during spontaneous absence
seizures in rats. J Neurosci. 2002;22(4):1480–1495.
44. Litt B, Esteller R, Echauz J, et al. Epileptic seizures may begin hours
in advance of clinical onset: a report of five patients. Neuron.
2001;30(1):51–64.
45. Aschenbrenner-Scheibe R, Maiwald T, Winterhalder M, et al. How
well can epileptic seizures be predicted? An evaluation of a nonlin-
ear method. Brain. 2003;126(Pt 12):2616–2626.
46. Elger CE, Lehnertz K. Seizure prediction by non-linear time series
analysis of brain electrical activity. Eur J Neurosci. 1998;10(2):
786–789.
47. Lehnertz K, Andrzejak RG, Arnhold J, et al. Nonlinear EEG analy-
sis in epilepsy: its possible use for interictal focus localization,
seizure anticipation, and prevention. J Clin Neurophysiol.
2001;18(3):209–222.
48. Litt B, Lehnertz K. Seizure prediction and the preseizure period.
Curr Opin Neurol. 2002;15(2):173–177.
49. Litt B, Echauz J. Prediction of epileptic seizures. Lancet Neurol.
2002;1(1):22–30.
50. Schelter B, Winterhalder M, Maiwald T, et al. Testing statistical sig-
nificance of multivariate time series analysis techniques for epilep-
tic seizure prediction. Chaos. 2006;16(1):013108.
51. Schelter B, Feldwisch-Drentrup H, Timmer J, et al. A common
strategy and database to compare the performance of seizure pre-
diction algorithms. Epilepsy Behav. 2010;17(2):154–156.
52. Harrison MA, Osorio I, Frei MG, et al. Correlation dimension and
integral do not predict epileptic seizures. Chaos. 2005;15(3):33106.
53. Harrison MA, Frei MG, Osorio I. Accumulated energy revisited.
Clin Neurophysiol. 2005;116(3):527–531.
54. Maiwald T, Winterhalder M, Aschenbrenner-Scheibe R, et al.
Comparison of three nonlinear seizure prediction methods by
means of the seizure prediction characteristic. Physica D.
2004;194(3–4):357–368.
55. Winterhalder M, Maiwald T, Voss HU, et al. The seizure prediction
characteristic: a general framework to assess and compare seizure
prediction methods. Epilepsy Behav. 2003;4(3):318–325.
56. Lehnertz K, Elger CE. Neuronal complexity loss in temporal lobe
epilepsy: effects of carbamazepine on the dynamics of the epilepto-
genic focus. Electroencephalogr Clin Neurophysiol. 1997;103(3):
376–380.
57. Lai YC, Harrison MA, Frei MG, et al. Inability of Lyapunov expo-
nents to predict epileptic seizures. Phys Rev Lett. 2003;91:068102.
58. Lai YC, Harrison MA, Frei MG, et al. Controlled test for predictive
power of Lyapunov exponents: their inability to predict epileptic
seizures. Chaos. 2004;14:630–642.
59. Bragin A, Engel J Jr, Wilson CL, et al. Electrophysiologic analysis of
a chronic seizure model after unilateral hippocampal Epilepsia KA
injection. Epilepsia. 1999;40(9):1210–1221.
60. Buzsáki G, Horváth Z, Urioste R, et al. High-frequency network
oscillation in the hippocampus. Science. 1992;256(5059):
1025–1027.
61. Bragin A, Engel J Jr, Wilson CL, et al. High-frequency oscillations
in human brain. Hippocampus. 1999;9(2):137–142.
62. Traub RD, Draguhn A, Whittington MA, et al. Axonal gap junctions
between principal neurons: a novel source of network oscillations,
and perhaps epileptogenesis. Rev Neurosci. 2002;13(1):1–30.
63. Bragin A, Wilson CL, Staba RJ, et al. Interictal high-frequency
oscillations (80–500 Hz) in the human epileptic brain: entorhinal
cortex. Ann Neurol. 2002;52(4):407–415.
64. Jirsch JD, Urrestarazu E, LeVan P, et al. High-frequency oscillations
during human focal seizures. Brain. 2006;129(Pt 6):1593–1608.
65. Jacobs J, Zijlmans M, Zelmann R, et al. Value of electrical stimula-
tion and high frequency oscillations (80–500 Hz) in identifying
epileptogenic areas during intracranial EEG recordings. Epilepsia.
2010;51(4):573–582.
66. Glass L, Mackey MC. (1988). The Rhythms of Life. New Jersey:
Princeton University Press; 1988:248.
67. Bélair J, Glass L, van der Heiden U, et al. Dynamical disease: iden-
tification, temporal aspects and treatment strategies of human ill-
ness. Chaos. 1995;5(1):1–7.
68. Lopes da Silva FH, Pijn JPM, Gorter JA, et al. Rhythms of the
brain: between randomness and determinism. In: Lehnertz K,
Arhold J, Grassberger P, et al., eds. Chaos in Brain. Singapore:
World Scientific; 2000:63–76.
69. Lopes da Silva FH, Blanes W, Kalitzin SN, et al. Epilepsies as
dynamical diseases of brain systems: basic models of the transition
between normal and epileptic activity. Epilepsia. 2003;44(s12):
72–83.
70. Lopes da Silva FH, Blanes W, Kalitzin SN, et al. Dynamical diseases
of brain systems: different routes to epileptic seizures. IEEE Trans
Biomed Eng. 2003;50(5):540–548.
71. Andrzejak RG, Widman G, Lehnertz K, et al. The epileptic process
as nonlinear deterministic dynamics in a stochastic environment:

an evaluation on mesial temporal lobe epilepsy. Epilepsy Res.
2001;44:129–40.
72. Lytton WW. Computer modelling of epilepsy. Nat Rev Neurosci.
2008;9(8):626–637.
73. Baumgartner C, Serles W, Leutmezer F, et al. Preictal SPECT in
temporal lobe epilepsy: regional cerebral blood flow is increased
prior to electroencephalography-seizure onset. J Nucl Med.
1998;39(6):978–982.
74. Federico P, Abbott DF, Briellmann RS, et al. Functional MRI of the
pre-ictal state. Brain. 2005;128:1811–1817.
75. Kalitzin SN, Velis DN, Lopes da Silva FH. Stimulation-based antic-
ipation and control of state transitions in the epileptic brain.
Epilepsy Behav. 2010;17(3):310–323.
76. Ohayon EL, Kwan HC, McIntyre Burnham W, et al. Adaptable
intermittency and autonomous transitions in epilepsy and cogni-
tion, advances in cognitive neurodynamics. ICCN Proceedings.
Springer; 2007;485–490.
77. Ohayon EL, Kalitzin S, Suffczynski P, et al. Charting epilepsy by
searching for intelligence in network space with the help of evolv-
ing autonomous agents. J Physiol. 98(4–6):507–529.
78. Zalay OC, Bardakjian BL. Mapped clock oscillators as ring devices
and their application to neuronal electrical rhythms. IEEE Trans
Neural Syst Rehabil Eng. 2008;16(3):233–244.
79. Velazquez JL, Khosravani H, Lozano A, et al. Type III intermittency
in human partial epilepsy. Eur J Neurosci. 1999;11(7):2571–2576.
80. Polack PO, Guillemain I, Hu E, et al. Deep layer somatosensory
cortical neurons initiate spike-and-wave discharges in a genetic
model of absence seizures. J Neurosci. 2007;27(24):6590–6599.
81. Pinault, D, O’Brien TJ. Cellular and network mechanisms of
genetically-determined absence seizures. Thalamus Relat Syst.
2007;03(03):217–226.
82. Suffczynski P, Kalitzin S, Lopes Da Silva FH. Dynamics of non-
convulsive epileptic phenomena modeled by a bistable neuronal
network. Neuroscience. 2004;126(2):467–484.
83. Suffczynski P, Lopes da Silva FH, Parra J, et al. Dynamics of epilep-
tic phenomena determined from statistics of ictal transitions. IEEE
Trans Biomed Eng. 2006;53(3):524–532.
84. Suffczynski P, Kalitzin S, Lopes da Silva F, et al. Active paradigms
of seizure anticipation: computer model evidence for necessity of
stimulation. Phys Rev E Stat Nonlin Soft Matter Phys. 2008;78
(5 Pt 1):051917.
85. Bouwman BM, Suffczynski P, Lopes da Silva FH, et al. GABAergic
mechanisms in absence epilepsy: a computational model of
Chapter 30 ■ Anticipating Seizures Based on EEG 657
absence epilepsy simulating spike and wave discharges after viga-
batrin in WAG/Rij rats. Eur J Neurosci. 2007;25(9):2783–2790.
86. Milton JG, Gotman J, Remillard GM, et al. Timing of seizure
recurrence in adult epileptic patients: a statistical analysis.
Epilepsia. 1987;28:471–478.
87. Cohen I, Huberfeld G, Miles R. Emergence of disinhibition-
induced synchrony in the CA3 region of the guinea pig hippocam-
pus in vitro. J Physiol. 2006;570(Pt 3):583–594.
88. Wendling F, Bartolomei F, Bellanger JJ, et al. Epileptic fast activity
can be explained by a model of impaired GABAergic dendritic
inhibition. Eur J Neurosci. 2002;15(9):1499–1508.
89. Wendling F. Computational models of epileptic activity: a bridge
between observation and pathophysiological interpretation.,
Expert Rev Neurother. 2008;8(6):889–896.
90. Parra J, Kalitzin SN, Iriarte J, et al. Gamma-band phase clustering
and photosensitivity: Is there an underlying mechanism common
to photosensitive epilepsy and visual perception? Brain.
2003;126(Pt 5):1164–1172.
91. Kalitzin SN, Parra JG, Velis DN, et al. Enhancement of phase cluster-
ing in the EEG/MEG gamma frequency band anticipates transitions
to paroxysmal epileptiform activity in epileptic patients with known
visual sensitivity. IEEE Trans BioMed Eng. 2002;49(11):1279–1285.
92. Kalitzin SN, Velis DN, Suffczynski P, et al. Electrical brain-
stimulation paradigm for estimating the seizure onset site and the
time to ictal transition in temporal lobe epilepsy. Clin Neurophysiol.
2005;116(3):718–728.
93. Badawy R, Macdonell R, Jackson G, et al. The peri-ictal state: cor-
tical excitability changes within 24 h of a seizure. Brain. 2009;132:
1013–1021.
94. Kalitzin SN, Parra J, Velis DN, et al. Quantification of unidirec-
tional nonlinear associations between multidimensional signals.
IEEE Trans Biomed Eng. 2007;54(3):454–461.
95. Osorio I, Frei MG. Real-time detection, quantification, warning,
and control of epileptic seizures: the foundations for a scientific
epileptology. Epilepsy Behav. 2009;16(3):391–396.
96. Ullah G, Schiff SJ. Tracking and control of neuronal Hodgkin-
Huxley dynamics. Phys Rev E Stat Nonlin Soft Matter Phys.
2009;79(4 Pt 1):040901.
97. Velasco AL, Velasco F, Velasco M, et al. Electrical stimulation of the
hippocampal epileptic foci for seizure control: a double-blind,
long-term follow-up study. Epilepsia. 2007;48(10):1895–1903.

Nonepileptic Attacks
MEGAN SELVITELLI, TRUDY PANG, DONALD L. SCHOMER, AND ERNST NIEDERMEYER
S
udden, brief loss of consciousness, sudden behavior
changes, or strange subjective sensations are not neces-
sarily epileptic. This should be a truism, but there is a
present trend to ascribe too many transient conditions to an
epileptic mechanism. Here the differential diagnosis of epilep-
tic seizures is discussed briefly. Topics covered will include:
1. Circulatory insufficiency giving rise to acute cerebral
ischemia
2. Psychogenic alterations
3. Movement disorders imitating epileptic seizures (often asso-
ciated with loss of consciousness)
4. Paroxysmal disturbances in infants and children imitating
epileptic seizures
5. Toxic–metabolic disturbances
6. Disturbances related to sleep disorders (such as
narcolepsy–cataplexy complex and parasomnias)
7. Cerebrovascular disorders
8. Migraines
The first four topics will be discussed in this chapter, while
toxic–metabolic disturbances, sleep disorders, cerebrovascular
disorders, and migraines are covered in Chapters 21, 41, 19, and
24, respectively.
SYNCOPAL ATTACKS
Syncope has been defined as a sudden brief loss of conscious-
ness due to a discrete episode of generalized cerebral ischemia;
it is hence a manifestation of acute insufficiency of the cerebral
circulation (1). Syncope is frequently encountered and encom-
passes up to 3% of emergency room evaluations and 6% of hos-
pital admissions per year in the United States (2). As syncope
may result from benign causes such as vasovagal syncope to
more serious etiologies such as cardiac arrhythmias and auto-
nomic dysfunction, a thorough understanding of the symp-
toms, pathophysiology, evaluation, and differentiation from
seizures is warranted.
Classic descriptions of syncope note prodromal symptoms,
followed by loss of consciousness and rapid resumption of
awareness shortly thereafter. Prodromal symptoms may include
blurry vision, decreased hearing, warmth, nausea, diaphoresis,
pallor, lightheadedness, and generalized weakness.
If cerebral perfusion is prolonged, as may occur with artifi-
cial maintenance of an upright posture, abnormal involuntary
movements may be observed that may be misinterpreted as
epileptic. Lempert et al. induced syncope in healthy volunteers
CHAPTER
31
and noted myoclonus in 90% of subjects, including multifocal
and generalized myoclonus (3). Other involuntary movements
were also witnessed, including head turning, lip-smacking,
chewing, and fumbling movements (79%); vocalizations
(40%); or righting movements (76%) (3). Similar involuntary
movements were also described in 14/15 patients undergoing
tilt table testing for unexplained syncope (4). In these patients,
rigid flexed postures were observed with associated eye rolling
and head extension, with an additional 30% of patients experi-
encing bilateral myoclonic rhythmic jerks and 27% experienc-
ing urinary incontinence. Thus, clinical description of events
may be insufficient for accurate diagnosis of either syncope or
seizure. For a summary of these features, see Table 31-1.
Given the difficulties in clinically differentiating convulsive
syncope from epileptic convulsions, Sheldon et al. utilized
symptom-based questionnaires to identify features more com-
mon in patients with diagnosed syncope or epilepsy. Epilepsy
was the more frequent etiology of loss of consciousness if the
events occurred often, were triggered by stressors, or were asso-
ciated with head deviation or unresponsiveness (5). Conversely,
syncope was more frequently associated with diaphoresis before
or after the event or associated with events triggered by pro-
longed sitting or standing (5).
Despite clinical features that may help differentiate between
seizures and syncope, these two entities may rarely co-occur. In
this rare entity, rhythmic ictal discharges give rise to bradycar-
dia, leading to decreased cardiac output and syncope. Schuele et
al. identified 10 patients with ictal asystole from a cohort of
6825 patients undergoing video-electroencephalography (EEG)
monitoring for medically refractory partial seizures (6). These
10 patients represented 0.27% of patients monitored and
demonstrated an initial ictal onset of 29.9 seconds prior to the
development of asystole (6). In these patients, loss of postural
tone was more commonly seen late in the seizure. Schuele et al.,
as well as other investigators, have not been able to convincingly
identify a significant laterality to ictal asystole, although
involvement of the temporal lobes, particularly the insular cor-
tex, and the orbitofrontal cortex has been suggested (7). The
exact frequency of ictal asystole is not known and is presumed
to be underreported, as the identification requires concomitant
EEG and EKG recording. Identification of ictal asystole is
important, however, as it may contribute to sudden unex-
plained syncope in epilepsy. Treatment more commonly
includes placement of a permanent pacemaker and medication
with antiepileptic drugs, with care taken to avoid antiepileptic
medications that may contribute to arrhythmias.
659
660 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
Tabl e 31 . 1
List of Historical and Event Features that Differentiate Syncope and Seizure
Syncope
Preictal features
Diaphoresis More common
Ictal features
Loss of consciousness Common
Myoclonus Possible
Frequency of ictus Rare
Precipitation by stress Uncommon
Head turning Uncommon
Occurrence after prolonged
sitting or standing More common
Postictal features
Diaphoresis More common
Etiologies
Any decrease in cerebral perfusion may result in syncope. Various
etiologies may include neurally mediated syncope, orthostasis,
cardiac arrhythmias, or cardiac outflow obstructions.
Vasovagal faints, carotid sinus hypersensitivity, and situa-
tional faints all are examples of neurally mediated syncope,
resulting from inappropriate bradycardia and hypotension.
Pressure on hypersensitive carotid sinus baroreceptors increases
vagal tone, causing hypotension or bradycardia, or both, thereby
leading to cerebral hypoperfusion. Situational faints may occur
with physical pain, receipt of frightening news, defecation, mic-
turition, or cough; these typically occur in healthy individuals
with normal blood pressure and frequently consistent triggers.
Orthostasis may also cause syncope, due to either autonomic
dysfunction or decreased blood volume due to dehydration or
blood loss. Autonomic dysfunction may be seen with neurode-
generative disorders, such as Shy–Drager syndrome, Parkinson
disease, or postural orthostatic syndrome. Alternatively, sys-
temic diseases such as diabetes, alcoholism, or amyloidosis may
cause an autonomic neuropathy that results in orthostasis. In
addition, medications that cause peripheral vasodilatation (e.g.,
nitroglycerin, beta blockers) may cause a relative decrease in
blood volume.
Finally, cardiac disease may also contribute to syncope,
through either arrhythmias or decreased cardiac output.
Bradyarrhythmias, such as atrial fibrillation with slowed ven-
tricular response, atrioventricular block, and long QTc syn-
drome, and tachyarrhythmias, such as supraventricular
tachycardia, ventricular tachycardia, or torsades de pointes, may
decrease cardiac output and thereby decrease cerebral perfusion
episodically. Conversely, a more persistent decrease in cardiac
Seizure
Uncommon
Common, depends on
seizure type
Possible
Common
More common
More common
Less common
Uncommon
output may be seen with obstructive cardiac disease, such as
hypertrophic obstructive cardiomyopathy or aortic stenosis,
and may also be seen with low ejection fraction from prior
myocardial infarctions.
EEG in Syncopal Attacks
If clinical features alone do not help differentiate between syn-
cope and seizures, an EEG and simultaneous EKG during the
event may help clarify the diagnosis. Numerous EEG studies of
syncope have revealed the following sequence: initial alpha
depression, followed by low-voltage fast activity, generalized
high-amplitude theta, and then delta slowing occurring con-
comitantly with the loss of consciousness. In general, a period of
10 seconds elapses before high-voltage delta waves appear. If the
loss of consciousness is prolonged ( 15 seconds), the EEG may
demonstrate electrocerebral flattening. The patient may experi-
ence myoclonic jerks, with no corresponding epileptiform dis-
charges (4,8). After the event, the EEG demonstrates recovery to
the normal background rhythms. Conversely, epileptic loss of
consciousness may reveal epileptiform discharges consistent
with the seizure type, either focal or generalized.
BREATH-HOLDING ATTACKS OF EARLY
CHILDHOOD AS A SPECIAL FORM OF
SYNCOPE
Breath-holding attacks are related to syncope and occur rather
frequently in older infants and small children (9). Emotional
stress, frustration, minor trauma, and crying trigger these
attacks. Cyanotic spells typically begin at 6 to 18 months, result-
ing in breath-holding during expiration, with subsequent loss

of consciousness and flaccidity. These may be accompanied by
increased tonic and clonic movements. In contrast, pallid spells
occur in a similar age group, resulting in decreased tone, loss of
pulse, and pallor, with recovery in less than 1 minute. Pallid
breath-holding is felt to be secondary to bradycardia or asys-
tole, resulting in decreased cerebral perfusion. Of note, breath-
holding spells may be associated with iron deficiency and
respond to supplementation (10). The children almost always
have normal EEG findings. Khurana et al. demonstrated that an
increased RR interval from baseline of 0.5 seconds during 10
seconds of ocular compression with simultaneous EEG and
EKG recording helps differentiate breath-holding spells and
syncope from epileptic events in children (11).
PSYCHOGENIC ATTACKS
Concept of Nonepileptic Seizures
The symptomatology of the hysterical attack has been well
known since the work of Briquet (12) and Charcot (13).
Briquet’s approach to this problem has proved to be a more
solid foundation for modern theories than the work of Charcot,
which is beset with iatrogenic artifacts.
Other concepts were essentially based on psychodynamic
mechanisms such as dissociation (14) and conversion (15–18).
The concept of Kretschmer with emphasis on “hyponoic” and
“hypobulic” mechanisms warrants particular attention (19,20).
Hysterical seizures are associated with loss of impulse, usu-
ally in stressful situations. These attacks used to be attributed
solely (and by definition) to females; this clearly has been
proved to be untrue, although female adolescents and young
adults are most often affected.
The term pseudoseizures denotes a conglomeration of non-
cerebral or nonepileptic attacks. It was introduced by Liske and
Forster, who were aware of the fact that many, but not all, of the
patients had psychogenic (hysterical) attacks (21). Unfortunately,
this term is not conducive to a differentiation of the nonepileptic
attacks with certain distinctive features.
The term pseudoseizures has been more or less equated with
psychogenic attacks. This may have been the result of a search
for a euphemistic term since older terms such as hysterical or
psychogenic attacks have been thought to have a derogatory con-
notation. Interestingly, the term pseudoseizures has been fading
since 1990, while the term psychogenic seizures once again is
being used more frequently.
By definition, nonepileptic seizures (NES) connote paroxys-
mal events that behaviorally or symptomatically mimic epileptic
seizures but are not associated with concomitant epileptiform
activity on EEG. Presumably, these events are secondary to a
psychologic mechanism, with conversion disorder often being
cited. The incidence of NES is variable and estimated between
0.91 and 3.03/100,000 in the general population (22,23). Correct
diagnosis is often delayed in these patients, leading to increased
risk for iatrogenic complications of anticonvulsant use, delayed
referral to psychiatric treatment, and potential exposure to
unnecessary emergency therapies, including intubation and
CNS-suppressant medications. Furthermore, earlier diagnosis of
Chapter 31 ■ Nonepileptic Attacks 661
NES is associated with improved outcomes and better NES event
control. Nonetheless, the differentiation between these entities is
often difficult based on historical features alone and is best diag-
nosed with video-EEG monitoring. Complicating the discrimi-
nation between NES and epileptic seizures is their common
co-occurrence.
Differentiation of Nonepileptic
Seizures from Epileptic Seizures
Numerous researchers have attempted to find a diagnostic test
that may easily differentiate between NES and epileptic seizures.
Excellent reviews by Cragar and colleagues, as well as by Reuber,
provide a summary of differentiating features based on demo-
graphic variables, medical history, and seizure semiology (24,25).
Of note, patients with NES are more likely to be women, and
have a prior history of psychiatric treatment and a prior history
of abuse or sexual abuse; however, men with NES frequently have
a history of work-related problems or posttraumatic stress disor-
der. NES patients are typically older at the onset of the events and
had been treated with fewer anticonvulsants than patients with
epilepsy. With regard to seizure semiology, NES patients had
greater variability in behaviors during the events, greater fre-
quency of pelvic thrusting and eye closure during the event.
Reuber also cited commonly held beliefs that NES events often
manifest with gradual onset, excessive limb movements, often
asynchronous and asymmetric, with possible opisthotonus, side-
to-side head shaking, crying, resistance to eye opening, preserva-
tion of pupillary light responses, and purposeful movements
(25). Interestingly, there were no significant differences between
patients with NES or epileptic seizures with regards to age, years
of education, seizure frequency, or family history of epilepsy or
psychiatric disorders. Similarly, there were no significant differ-
ences between these groups regarding presence of urinary incon-
tinence or tongue biting during the ictus. Epileptic seizures were
typically of shorter duration, more commonly associated with
injuries during the event, more frequently occurred during sleep,
and were associated with postictal confusion.
More recent studies have similarly looked for historical and
event behaviors that may predict whether the events represent
NES or epileptic seizures. Benbadis established that patients
evaluated in an epilepsy clinic for refractory seizures were more
likely to exhibit NES if they had a history of fibromyalgia or
chronic pain or had experienced a typical event in the waiting
or examining room (26). Opherk and Hirsch found that during
staring spells, patients with NES had no change in heart rate,
while patients with complex partial seizures had a marked
tachycardia; a similar pattern was seen comparing generalized
tonic–clonic seizures with nonepileptic convulsive seizures
(27). Oliva and colleagues noted a low sensitivity (8%) but high
specificity (100%) for epileptic seizures when both oral lacera-
tions and urinary incontinence were observed during inpatient
video-EEG monitoring compared to their occurrence in NES
events (28). Postictally, whispering responses, partial motor
responses to commands, and shallow, irregular, rapid breathing
are more commonly seen with NES events (29,30). For a sum-
mary of these features, see Table 31-2.
662 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

Tabl e 31 . 2

List of Historical and Paroxysmal Event Characteristics That Discriminate Between Nonepileptic
Attacks and Epileptic Seizures

Nonepileptic Attacks Epileptic Seizures

Historical features
Age of onset Older Younger
Gender predominance Female None
History of abuse Common Less common
History of pain syndromes Common Uncommon
or event in waiting
or examining room
History of prior psychiatric treatment Common Less common
History of anticonvulsant therapy Typically fewer medication Typically more medication
trials and less polytherapy trials and greater polytherapy
Event features
Duration of event Longer (typically 2 minutes) Shorter (typically 1–2 minutes)
Sleep occurrence Uncommon Common
Ictal behaviors (pelvic thrusting, Common Uncommon
side-to-side head movements,
gradual onset, asymmetric/
asynchronous limb movements,
variable amplitude, crying)
Ictal heart rate Stable Tachycardia
Ictal urinary incontinence Uncommon Common
and tongue bite
Ictal eye closure Common Uncommon
Postictal confusion Uncommon Common
Postictal whisper or partial Common Not seen
motor responses
Postictal breathing patterns Often rapid, shallow, soft, Deep and prolonged inspiratory/
and irregular expiratory phases, regular and
loud, often with snoring

Electroencephalographic Observations sharply contoured normal background alpha rhythms, wicket

Given the variable clinical manifestations of NES, video-EEG
monitoring is considered the gold standard for diagnosis of
NES. However, video-EEG monitoring is a labor-intensive and
costly procedure and numerous investigators have evaluated
alternative options for diagnosis of NES.
Interictal EEGs are the more commonly used modality in the
initial evaluation of patients with suspected seizures. In video-
EEG-confirmed NES patients, up to 32% of patients had a prior
history of an abnormal “epileptiform” EEG in one study (31).
On visual inspection of 37% of these outside EEGs, none had
epileptiform discharges. Normal EEG features that were over-
stated as epileptiform abnormalities included fluctuations of
spikes, hypnagogic hypersynchrony, and hyperventilation-
induced slowing. In a separate study, Reuber et al. performed
both a retrospective analysis of interictal EEGs and a blinded,
controlled investigation of interictal routine EEG abnormalities
in video-EEG-confirmed NES compared to age- and sex-
matched healthy controls (32). In the retrospective analysis,
53.8% of NES patients demonstrated nonspecific EEG abnor-
malities while 12.3% demonstrated epileptiform discharges. In
the prospective study, 9/50 NES patients demonstrated inter-
mittent slow rhythms and 1 patient had sharp and slow-wave
activity. Comparatively, 5/50 controls also demonstrated inter-
mittent slow activity, with 1 patient demonstrating isolated left

temporal sharp waves. Thus, in the prospective study, there did
not appear to be a difference in the frequency of interictal non-
specific slowing or epileptiform discharges between the NES
patients and control group. These studies would suggest that
the incidence of epileptiform discharges and nonspecific slow-
ing in NES patients is similar to that found in the normal pop-
ulation and implies that a normal EEG is more commonly seen
in patients with NES compared to epileptic seizures.
As interictal EEGs may also be normal in many patients with
epileptic seizures, the gold standard for diagnosis is video-EEG
monitoring. As inpatient monitoring involves high labor costs,
procedures to decrease monitoring time have been devised.
Benbadis et al. found that 1 to 2 hours of outpatient video-EEG
monitoring confirmed the diagnosis in suspected NES in 69%
of patients, when induction techniques including prolonged
hyperventilation, photic stimulation, and verbal description of
the events were used (33). Wassmer and colleagues were able to
induce NES with saline infusion and verbal suggestion in 62%
of patients with clinically suspected NES and 24 hours of non-
diagnostic video-EEG testing (34). Ethical considerations have
arisen with the use of induction tests, although the use of
hyperventilation and photic stimulation, which are known to
induce seizures, appears to remove the deceit inherent in some
induction techniques (e.g., saline infusion, alcohol pad place-
ment, and use of tuning fork).
Nonetheless, video-EEG monitoring, without the use of
induction techniques, was demonstrated to confirm the diag-
nosis of either NES or epileptic seizures within 2 days of inpa-
tient monitoring in 88% of patients (35).
However, a small number of patients cannot be diagnosed by
video-EEG monitoring due to ictal semiology (partial symp-
toms or symptoms suggesting frontal lobe seizures that may
manifest without EEG alterations), absence of events during
monitoring, or obscuration of the EEG due to movement arti-
fact (36). Vinton et al. describe a technique in which the domi-
nant EEG frequency for 1- to 2-second epochs throughout the
episode was evaluated during NES and epileptic seizures (37).
Interestingly, there was little variability of the dominant EEG
frequency during NES, suggesting that movement artifact asso-
ciated with NES is constant throughout the event. Furthermore,
the authors noted that NES events were associated with an
“on–off” pattern of intermittent and sudden variation between
the dominant alpha rhythm and the movement artifact.
Conversely, ictal EEG dominant frequencies evolved through-
out the seizures in two dominant patterns. Thus, despite move-
ment artifact obscuring the underlying EEG abnormalities,
assessment of the dominant EEG frequency over time can dif-
ferentiate between NES and epileptic seizures.
Co-occurrence of NES and Epilepsy
Aggravating the difficulties discriminating between NES and
epileptic seizures is their frequent co-occurrence. Various esti-
mates are suggested, 5% to 47%, with lower estimates seen in
studies requiring video-EEG-confirmed NES and epilepsy and
higher estimates seen when epilepsy was presumed on the basis
of clinical history (38,39). Of note, Mari and associates noted
64% of patients with concomitant NES and epilepsy had
Chapter 31 ■ Nonepileptic Attacks 663
similar ictal semiology with both events, making clinical dis-
tinction challenging in these patients (38).
Thus, continued studies are needed to determine whether a
combination of historical variables, ictal and postictal behav-
iors, and interictal EEG abnormalities is helpful in the differen-
tiation of NES and epileptic seizures, while the gold standard of
diagnosis currently remains video-EEG demonstration of typi-
cal ictal behaviors with absence of epileptiform abnormalities.
MOVEMENT DISORDERS IMITATING
EPILEPTIC SEIZURES
Hyperkinetic movement disorders can often mimic focal motor
seizures and present a diagnostic dilemma in the clinical set-
ting. However, the history and phenomenology of the abnormal
movements play a critical role in helping to differentiate the two
conditions. Clearly identifiable triggers and movements that are
not stereotyped are more suggestive of a movement disorder,
rather than epilepsy. In addition, EEG recording while the
patient is symptomatic is helpful as it is normal in movement
disorders. A variety of conditions and their distinction from
partial seizures are discussed below, but see a recent publication
for a more complete description of several of the disorders (40).
Chorea is derived from the Greek word chorea, meaning “a
dance.” It is a hyperkinetic disorder characterized by brief, jerky,
nonrhythmic movements that involve various body parts in a
random fashion. Bilateral involvement is typically encountered.
A wide variety of conditions affecting the basal ganglia may
result in chorea, but it is most often associated with Huntington
disease, Sydenham chorea after streptococcal infections, or
strokes. Certain systemic disorders may also be associated with
chorea, such as hyperthyroidism, systemic lupus erythemato-
sus, primary antiphospholipid antibody syndrome, or paraneo-
plastic syndromes. Oral contraceptive use and pregnancy result
in hormonal changes that have been recognized as a cause of
chorea. While choreiform movements are nonrhythmic, jerky,
and brief, focal motor seizures are stereotyped, rhythmic move-
ments and can affect a single body part or spread to other body
parts in a progressive fashion. When the pattern of progression
follows the motor homunculus, this is known as the
“Jacksonian march.”
Athetosis consists of slow, continuous writhing movements
involving the limbs or face (unilaterally or bilaterally), trunk,
head, or tongue. It often results from lesions of the basal gan-
glia, due to stroke, tumor, demyelination, and infection, or it
may be drug-induced. Perinatal injury to the basal ganglia and
thalamus has been recognized as an etiology in the pediatric
population. In many cases, athetosis occurs in conjunction with
chorea and this is known as choreoathetosis. The slow and
writhing nature of the movements is quite distinct from focal
motor seizures that are generally fast and rhythmic.
Ballismus is another hyperkinetic movement disorder.
Ballismus or ballism is derived from the Greek word meaning
“to throw.” In contrast to focal seizures that are stereotyped,
rhythmic, and predictable, ballismus is characterized by
rapid, violent, high-amplitude flinging or flailing limb move-
ments that are unpredictable. The proximal muscles are
664 Part IV ■ Clinical EEG in Epilepsy and Related Disorders
predominantly affected and the movements are usually unilat-
eral, resulting in hemiballismus. It can be caused by a variety
of conditions affecting the striatum or subthalamic nucleus,
including systemic processes, such as hyperglycemia, or
structural lesions, such as strokes and tumor (41,42).
Medications have been known to cause ballismus as well, and
the commonly described agents include phenytoin, levodopa,
and oral contraceptives.
Dystonia is another condition that may be confused with
partial seizures, and the correct diagnosis is important in
directing treatment. It is characterized by paroxysmal, sustained
co-contractions of the agonist and antagonist muscles of a body
part resulting in twisting, repetitive movements, or abnormal
postures (43). Although the movements are usually slow, they
can also be quick and jerky, and may even have a rhythmic com-
ponent. The limbs are most often involved, but facial involve-
ment has also been seen. Anxiety or stress can trigger or worsen
the dystonia, but some patients have developed “sensory tricks,”
which are maneuvers or positions that improve or minimize it.
Dystonia is caused by abnormal basal ganglia function and is
felt to be associated with dysfunction of the inhibitory
interneuronal circuits in the brainstem and spinal cord, as well
as thalamic control of cortical motor planning and execution
(44). Patients can present with primary dystonia or secondary
dystonia, due to stroke, tumor, neurodegenerative processes,
medications, and various toxic substances.
Myoclonus is one of the most frequently encountered move-
ment disorders, characterized by sudden, brief involuntary
movements, due to either muscle contractions, resulting in pos-
itive myoclonus, or sudden muscle inhibition, resulting in neg-
ative myoclonus. Myoclonus is referred to as segmental when it
is restricted to certain muscle groups or body segments, but it
can be diffuse and involve the entire body. Since there are
numerous causes of myoclonus, it is commonly classified into
the following four major groups: (1) physiologic, such as hic-
cups; (2) essential, also known as hereditary or idiopathic; (3)
symptomatic, with an underlying abnormality (toxic, meta-
bolic, infectious, anoxic causes); (4) epileptic, with myoclonus
as part of an epileptic syndrome such as juvenile myoclonic
epilepsy. Physiologic myoclonus is considered to be the most
benign form. Essential myoclonus, also known as benign essen-
tial myoclonus, is an inherited, autosomal dominant disorder in
which myoclonic jerks affect the arms or trunk and there may
be associated dystonia (45,46). In symptomatic myoclonus,
there is an identifiable underlying cause, such as a systemic dis-
turbance, as seen in infections and toxic or metabolic derange-
ments, or it may be a manifestation of a variety of neurologic
conditions, including neurodegenerative disorders, strokes, or
injuries to the spinal cord (spinal segmental myoclonus). In sys-
temic disturbances or neurodegenerative disorders,
encephalopathy is the predominant feature and seizures may
occur. Finally, last category of myoclonus, epileptic myoclonus,
is a feature that may be seen in a wide variety of seizure types,
for example, juvenile myoclonic epilepsy. To distinguish
between epileptic and nonepileptic myoclonus, EEG is an
essential tool since nonepileptic myoclonus does not have any
EEG correlate, whereas those with an underlying epilepsy syn-
drome demonstrate clear epileptiform discharges. On occasion,
electromyography (EMG) may also be helpful in that the EMG
burst due to epileptic myoclonus is usually short, lasting less
than 50 msec, compared to a much longer duration of 200 to
300 msec as seen in nonepileptic myoclonus (47). Tics represent
an extremely heterogeneous group of disorders and consist of
brief, intermittent, involuntary movements (motor tics) or
sounds (phonic tics) that may be exacerbated by anxiety. They
may also occur in normal individuals. Classically, patients expe-
rience an urge that builds and is relieved by performance of the
tic. Although tics can be temporarily suppressed, the urge can
become so overwhelming with time that ultimately the tic can-
not be suppressed. The myriad of etiologies include structural
lesions, genetic abnormalities, psychiatric disorders, or strepto-
coccal infections (48,49). The disorder commonly begins in
childhood and peaks in adolescence. Complex motor tics such
as eye blinking may occasionally be confused with absence
seizures; however, patients do not have any alteration of aware-
ness. Other complex motor tics such as head shaking, jumping,
and obscene gestures are less likely to be confused with seizures,
in that they are usually not stereotyped and can be briefly sup-
pressed by suppressing the urge to tic. The EEG during the
attacks is normal.
Hemifacial spasm is characterized by irregular contractions
of the facial muscles and may be mistaken for focal motor
seizures. It is typically on one side of the face, although bilateral
involvement has also been reported (50). It results from spon-
taneous firing of the facial nerve, which may occur sporadically
or with vascular compression of the nerve after it exits from the
brainstem, and, very infrequently, with a demyelinating brain-
stem plaque due to multiple sclerosis. An association with a past
history of Bell’s palsy or injury to the facial nerve has also been
reported, in which synkinesis of the facial muscles due to aber-
rant regeneration of the facial nerve may sometimes mimic
focal motor seizures. Differentiating between hemifacial spasm
and focal motor seizures based on EEG alone may be difficult
since focal motor seizures involving a small area of the cerebral
cortex may not have a surface EEG correlate. Thus, a detailed
history and brain imaging will play an important role in the
diagnostic process.
Paroxysmal kinesigenic dyskinesia (PKD), a hyperkinetic
disorder, is characterized by recurrent attacks of any combina-
tion of movements, including dystonic posture, chorea, atheto-
sis, and ballism (51). The movements may be unilateral or
bilateral, and are usually brief, lasting seconds, and always less
than 5 minutes. A number of triggers have been reported,
including sudden movement, hyperventilation, and startle. The
primary form carries both an autosomal dominant and an
autosomal recessive pattern of inheritance and the locus has
been mapped to chromosome 16 (52,53). The secondary form
may result from other neurologic disorders such as multiple
sclerosis (54). There is an association with infantile generalized
convulsions in patients with familial PKD (55), while some
reported cases by Hirata et al. and Lombroso demonstrated
rhythmic discharges during attacks (56,57). Sadamatsu et al.
performed video-EEG monitoring in two patients with classic
PKD and did not find ictal discharges during attacks (58). The

clinical cases presented by Lombroso and Sadamatsu may be
distinct as the former case involved a patient with mental retar-
dation and the attacks were precipitated loud noises and men-
tal stress in addition to movement, whereas the latter cases by
Sadamatsu involved developmentally normal patients who
never developed generalized convulsions after PKC attacks;
thus, Sadamatsu concluded that PKC in their patients is distinct
from reflex epilepsy. Paroxysmal nonkinesigenic dyskinesia
(PNKD) is similar to PKD, but as the name implies, it is not
triggered by movement. Coffee, alcohol, hunger, fatigue, and
tobacco have been reported as precipitating factors (59). It typ-
ically lasts longer, on the order of minutes to hours rather than
seconds to minutes and consciousness is preserved. A number
of etiologies have been implicated, including psychogenic cere-
brovascular disease, multiple sclerosis, encephalitis, cerebral
trauma, peripheral trauma, migraine, and kernicterus (60).
Tardive dyskinesia and tardive dystonia are two movement
disorders commonly associated with dopamine antagonists,
such as antipsychotics, antiemetics, and even antidepressants.
While tardive dyskinesia refers to choreiform movements of the
mouth, tongue, head, neck, and limbs, tardive dystonia refers to
dystonic posturing or twisting of body parts as a result of sus-
tained muscle contractions (61). The first step in the treatment
of these conditions is removing the offending agent; however,
sometimes additional pharmacologic treatment is required to
control or improve the symptoms. The history, clinical appear-
ance of the movements, and normal EEG help to distinguish
the two disorders from epilepsy.
Hyperekplexia (startle disease) is a rare hereditary condi-
tion, caused by an abnormal gene for the subunit of glycine
receptor on chromosome 5q31.2 (62,63), and follows an auto-
somal dominant pattern of inheritance. Less commonly, auto-
somal recessive cases have also been reported (64). Excessive
and persistent startle to sudden auditory, visual, or sensory
stimuli is the hallmark feature of this condition (65,66). It typ-
ically begins in infancy causing stiff baby syndrome, as sudden
stimuli can trigger leg or body stiffening that are sometimes
severe enough to cause apnea (67). Affected babies demonstrate
the head retraction reflex, whereby gentle tapping of the nose or
forehead results in brisk retraction of the head that does not
habituate with repetition. In children, sudden stimuli or strong
emotions such as excitement or fear can trigger a typical attack,
resulting in falls with preserved awareness. At night, patients
may also experience spontaneous clonus lasting a few minutes
in lower extremities (65,66).
The clinical features of hyperekplexia may resemble startle
epilepsy, which also can present with falls and body stiffening.
Thus, it is important to differentiate hyperekplexia from the
epileptic disorder. To compound the confusion, treatment of
hyperekplexia consists of clonazepam or valproate that has
been shown to be effective for symptomatic relief (65,68).
However, as hyperekplexia is typically an inherited condition,
particularly in infants and children, a positive family history is
often present and examination of the parents aids the diagno-
sis. Startle epilepsy is characterized by brief seizures, typically
less than 30 seconds, triggered by unexpected auditory or tactile
stimuli, usually a sudden sound (69). The seizures consist of a
Chapter 31 ■ Nonepileptic Attacks 665
startle response followed by tonic contraction (usually asym-
metric); patients may fall and have clonic jerks as well. These
patients typically have a neurologic deficit such as hemiparesis
and the weak side is typically involved in the seizure. The EEG
is very helpful in distinguishing the two disorders as the EEG in
hyperekplexia is normal, while the EEG abnormality in startle
epilepsy usually involves the supplementary motor area; thus,
frontal, central, or frontocentral spikes can be seen (69).
Startle disease should be distinguished from another syn-
drome known as the “Jumping Frenchmen of Maine” in which
excessive startle reflex is a key feature. It was first described by
Beard, in a group of French-Canadian lumbermen in the
Moosehead Lake region of Maine (70). Patients with this disor-
der have the characteristic startle, but have additional features,
including echolalia and echopraxia, which are thought to be tic
related, and again the EEG does not demonstrate epileptiform
activity. The underlying molecular basis is unknown.
PAROXYSMAL DISTURBANCES IN
INFANTS AND CHILDREN IMITATING
EPILEPTIC SEIZURES
Hyperekplexia, also known as startle disease or stiff baby syn-
drome, is described in the section “Movement Disorders
Imitating Epileptic Seizures.”
Shuddering attacks in children are brief, fast tremors of the
head, shoulder, or trunk, lasting seconds, and often occur dur-
ing eating. The condition begins in infancy or childhood and
resolves spontaneously by the teenage years. Neurologically and
developmentally, the affected children are normal, and the EEG
is normal during these attacks (71). There is an apparent asso-
ciation with a family history of essential tremor (72).
Opisthotonos consists of intermittent generalized body stiff-
ening or extension posturing. This may be due to a gastroe-
sophageal reflux disease known as Sandifer syndrome, which
may be mistaken for tonic seizures. These spells tend to occur
within 30 minutes of a feeding and children may exhibit apnea,
staring, and minor jerking of the extremities during them. The
clear association with feedings is a good indicator of the diag-
nosis and the EEG during these spells is normal (73).
Repetitive movements known as stereotypies refer to various
stereotypic movements such as head banging, body rocking,
and hand flapping in children. Although these may be seen in
normal children, they are commonly encountered in children
with neurologic deficits, such as mental retardation and autism.
The behaviors are thought to be “self-stimulatory” for the pur-
pose of relaxation and can be present during the waking state or
during transition to sleep.
Spasmus nutans consists of a triad of clinical symptoms of
recurrent monocular nystagmus, head nodding, and head tilt,
which may be mistaken for seizures. The typical age of onset is
4 months and often resolves within a few years. The exact
underlying cause is still unknown and the EEG during these
spells is normal.
Daydreaming is a common phenomenon in children which
may be mistaken for absence or complex partial seizures. Unlike
666 Part IV ■ Clinical EEG in Epilepsy and Related Disorders

seizures, during these spells, children can respond to questions 23. Szaflarski JP, Ficker DM, Cahill WT, et al. Four-year incidence of
appropriately when their attention is redirected and there are psychogenic nonepileptic seizures in adults in Hamilton county,
no postictal symptoms such as confusion or lethargy. OH. Neurology. 2000;55:1561–1563.
24. Cragar DE, Berry DTR, Fakhoury TA, et al. A review of diagnostic
techniques in the differential diagnosis of epileptic and nonepilep-
REFERENCES tic seizures. Neuropsychol Rev. 2002;12:31–64.
25. Reuber M. Psychogenic nonepileptic seizures: answers and ques-
1. Naquet R, Vigouroux RA. Acute cerebral anoxia and syncopal
tions. Epilepsy Behav. 2008;12:622–635.
attacks. In: Remond A, ed. Handbook of Electroencephalography
26. Benbadis SR. A spell in the epilepsy clinic and a history of “chronic
and Clinical Neurophysiology. Vol 14A. Amsterdam: Elsevier;
pain” or “fibromyalgia” independently predict a diagnosis of psy-
1972:68–71.
chogenic seizures. Epilepsy Behav. 2005;6:264–265.
2. Sarasin FP, Louis-Simonet M, Carballo D, et al. Prospective evalu-
27. Opherk C, Hirsch LJ. Ictal heart rate differentiates epileptic from
ation of patients with syncope: a population-based study. Am
non-epileptic seizures. Neurology. 2002;58:636–638.
J Med. 2001;111:177–184.
28. Oliva M, Pattison C, Carino J, et al. The diagnostic value of oral
3. Lempert T, Bauer M, Schmidt D. Syncope: a videometric analysis
lacerations and incontinence during convulsive “seizures”.
of 56 episodes of transient cerebral hypoxia. Ann Neurol.
Epilepsia. 2008;49:962–967.
1994;36:233–237.
29. Chabolla DR, Shih JJ. Postictal behaviors associated with psy-
4. Ammirati F, Colivicchi F, Di Battista G, et al. Electro -
chogenic nonepileptic seizures. Epilepsy Behav. 2006;9:307–311.
encephalographic correlates of vasovagal syncope induced by
30. Azar NJ, Tayah TF, Wang L, et al. Postictal breathing pattern distin-
head-up tilt testing. Stroke. 1998;29:2347–2351.
guishes epileptic from nonepileptic convulsive seizures. Epilepsia.
5. Sheldon R, Rose S, Connolly S, et al. Diagnostic criteria for vaso-
2008;49:132–137.
vagal syncope based on a quantitative history. Eur Heart J. 2006;27:
31. Benbadis SR, Tatum WO. Overinterpretation of EEGs and misdi-
344–350.
agnosis of epilepsy. J Clin Neurophysiol. 2003;20(1):42–44.
6. Schuele SU, Bermeo AC, Alexopoulos AV, et al. Video-electrographic
32. Reuber M, Fernandez G, Bauer J, et al. Interictal EEG abnormali-
and clinical features in patients with ictal asystole. Neurology. 2007;
ties in patients with psychogenic nonepileptic seizures. Epilepsia.
69:434–441.
2002;43(9):1013–1020.
7. Carinci V, Barbato G, Baldrati A, et al. Asystole induced by partial
33. Benbadis SR, Seigrist K, Tatum WO, et al. Short-term outpatient
seizures: a rare cause of syncope. Pacing Clin Electrophysiol. 2007;30:
EEG with induction in the diagnosis of psychogenic seizures.
1416–1419.
Neurology. 2004;63:1728–1730.
8. Gastaut H, Fischer-Williams M. Electro-encephalographic study
34. Wassmer E, Wasmer SR, Donati F. Saline infusion: a diagnostic and
of syncope; its differentiation from epilepsy. Lancet. 1957;
therapeutic tool in nonepileptic attacks? Epilepsy Behav. 2003;4:
273(7004):1018–1025.
500–506.
9. Fejerman N. Nonepileptic disorders imitating generalized idio-
35. Lobello K, Morgenlander JC, Radtke RA, et al. Video/EEG moni-
pathic epilepsies. Epilepsia. 2005;46(suppl 9):80–83.
toring in the evaluation of paroxysmal behavioral events: dura-
10. Mackay M. Fits, faints and funny turns in children. Aust Fam
tion, effectiveness and limitations. Epilepsy Behav. 2006;8:261–266.
Physician. 2005;34:1003–1008.
36. Benbadis SR. The EEG in nonepileptic seizures. J Clin
11. Khurana DS, Valencia I, Kruthiventi S, et al. Usefulness of ocular
Neurophysiol. 2006;23:340–352.
compression during electroencephalography in distinguishing
37. Vinton A, Carino J, Vogrin S, et al. “Convulsive” nonepileptic
breath-holding spells and syncope from epileptic seizures. J Child
seizures have a characteristic pattern of rhythmic artifact distin-
Neurol. 2006;21:907–910.
guishing them from convulsive epileptic seizures. Epilepsia.
12. Briquet P. Traité de l’Hystérie. Paris; 1859.
2004;45:1344–1350.
13. Charcot JM. Leçons du Mardi à la Salpêtrière, Policlinique Paris:
38. Mari F, Di Bonaventura C, Vanacore N, et al. Video-EEG study of
Bureaux du Progrès Médical. Paris: Lecroisnier and Babe;
psychogenic nonepileptic seizures: differential characteristics in
1887–1889.
patients with and without epilepsy. Epilepsia. 2006;47(S5):64–67.
14. Janet P. L’état Mental des Hystériques. Paris; 1893/1894.
39. Martin R, Prasad BA, Powell T, et al. Frequency of epilepsy in
15. Breuer J. Studien über Hysterie. 2nd ed. Vienna: Deuticke; 1895.
patients with psychogenic seizures monitored by video-EEG.
16. Breuer J, Freud S. Studies on hysteria. In: Standard Edition of the
Neurology. 2003;61:1791–1792.
Complete Psychological Works of Sigmund Freud. Vol 7. London:
40. Schmitz B, Tettenborn B, Schomer DL. The Paroxysmal Disorders.
Hogarth; 1895, 1955.
2nd ed. Cambridge: Cambridge University Press; 2010.
17. Chodoff P, Lyons H. Hysteria, the hysterical personality and “hys-
41. Lee MS, Marsden CD. Movement disorders following lesions of
terical” conversion. Am J Psychiatry. 1958;114:734.
the thalamus or subthalamic region. Mov Disord. 1994;9:493–507.
18. Ziegler F, Imboden J, Meyer E. Contemporary conversion reac-
42. Lai PH, Tien RD, Chang MH, et al. Chorea–ballismus with nonke-
tions: a clinical study. Am J Psychiatry. 1960;116:901.
totic hyperglycemia in primary diabetes mellitus. AJNR Am
19. Kretschmer E. Hysterie, Reflex und Instinkt. 6th ed. Stuttgart:
J Neuroradiol. 1996;17:1057–1064.
Thieme; 1923, 1958.
43. Jankovic J, Tolosa E. Dystonic disorders. In: Parkinson’s Disease and
20. Kretschmer E. Medizinische Psychologie. 11th ed. Stuttgart:
Movement Disorders. 4th ed. Philadelphia, PA: Lippincott Williams
Thieme; 1956.
& Wilkins; 2002:14,721.
21. Liske E, Forster FM. Pseudoseizures: a problem in the diagnosis
44. Berardelli A, Rothwell JC, Hallett M, et al. The pathophysiology of
and management of epileptic patients. Neurology. 1964;14:41–49.
primary dystonia. Brain. 1998;121:1195–1212.
22. O’Sullivan SS, Spillane JE, McMahom EM, et al. Clinical characteris-
45. Mahloudji M, Pikielny RT. Hereditary essential myoclonus. Brain.
tics and outcome of patients diagnosed with psychogenic nonepilep-
1967;90:669–674.
tic seizures: a 5-year review. Epilepsy Behav. 2007;11:77–84.

46. Vidailhet M, Tassin J, Durif F, et al. A major locus for several phe-
notypes of myoclonus-dystonia on chromosome 7q. Neurology.
2001;56:1213–1216.
47. Hallett M. Myoclonus: relation to epilepsy. Epilepsia. 1985;26:
S67–S77.
48. Kumar R, Lang AE. Tourette syndrome. Secondary tic disorders.
Neurol Clin. 1997;15:309–331.
49. Swedo SE, Leonard HL, Garvey M, et al. Pediatric autoimmune
neuropsychiatric disorders associated with streptococcal infec-
tions: clinical description of the first 50 cases. Am J Psychiatry.
1998;155:264–271.
50. Tan EK, Jankovic J. Bilateral hemifacial spasm: a report of five
cases and a literature review. Mov Disord. 1999;14:345–349.
51. Jankovic J, Demirkiran M. Classification of paroxysmal dyskine-
sias and ataxias. Adv Neurol. 2002;89:387–400.
52. Bennett LB, Roach ES, Bowcock AM. A locus for paroxysmal kine-
sigenic dyskinesia maps to human chromosome 16. Neurology.
2000;54:125–130.
53. Goodenough DJ, Fariello RG, Annis BL, et al. Familial and
acquired paroxysmal dyskinesias. A proposed classification with
delineation of clinical features. Arch Neurol. 1978;35:827–831.
54. Tranchant C, Bhatia KP, Marsden CD. Movement disorders in
multiple sclerosis. Mov Disord. 1995;10:418–423.
55. Swoboda KJ, Soong B, McKenna C, et al. Paroxysmal kinesigenic
dyskinesia and infantile convulsions: clinical and linkage studies.
Neurology. 2000;55:224–230.
56. Hirata K, Katayama S, Saito T, et al. Paroxysmal kinesigenic
choreoathetosis with abnormal electroencephalogram during
attacks. Epilepsia. 1991;32:492–494.
57. Lombroso CT. Paroxysmal choreoathetosis: an epileptic or non-
epileptic disorder? Ital J Neurol Sci. 1995;16:271–277.
58. Sadamatsu M, Masui A, Sakai T, et al. Familial paroxysmal kinesi-
genic choreoathetosis: an electrophysiologic and genotypic analy-
sis. Epilepsia. 1999;40:942–949.
59. Bruno MK, Lee HY, Auburger GW, et al. Genotype–phenotype
correlation of paroxysmal nonkinesigenic dyskinesia. Neurology.
2007;68:1782–1789.
60. Demirkiran M, Jankovic J. Paroxysmal dyskinesias: clinical fea-
tures and classification. Ann Neurol. 1995;38:571–579.
Chapter 31 ■ Nonepileptic Attacks 667
61. Burke RE, Fahn S, Jankovic J, et al. Tardive dystonia: late-onset and
persistent dystonia caused by antipsychotic drugs. Neurology.
1982;32:1335–1346.
62. Ryan SG, Sherman SL, Terry JC, et al. Startle disease, or hyperek-
plexia: response to clonazepam and assignment of the gene (STHE)
to chromosome 5q by linkage analysis. Ann Neurol. 1992;31:
663–668.
63. Shiang R, Ryan SG, Zhu YZ, et al. Mutations in the alpha 1 subunit
of the inhibitory glycine receptor cause the dominant neurologic
disorder, hyperekplexia. Nat Genet. 1993;5:351–358.
64. Rees MI, Andrew M, Jawad S, et al. Evidence for recessive as well as
dominant forms of startle disease (hyperekplexia) caused by
mutations in the alpha 1 subunit of the inhibitory glycine recep-
tor. Hum Mol Genet. 1994;3:2175–2179.
65. Andermann F, Keene DL, Andermann E, et al. Startle disease or
hyperekplexia: further delineation of the syndrome. Brain.
1980;103:985–997.
66. Kurczynski TW. Hyperekplexia. Arch Neurol. 1983;40:246–248.
67. Cioni G, Biagioni E, Bottai P, et al. Hyperekplexia and stiff-baby
syndrome: an identical neurological disorder? Ital J Neurol Sci.
1993;14:145–152.
68. Dooley JM, Andermann F. Startle disease or hyperekplexia: adoles-
cent onset and response to valproate. Pediatr Neurol. 1989;5:
126–127.
69. Aguglia U, Tinuper P, Gastaut H. Startle-induced epileptic
seizures. Epilepsia. 1984;25:712–720.
70. Beard GM. Remarks upon ‘jumpers or jumping Frenchmen’. J Nerv
Ment Dis. 1878;5:526.
71. Holmes GL, Russman BS. Shuddering attacks. Evaluation using
electroencephalographic frequency modulation radiotelemetry
and videotape monitoring. Am J Dis Child. 1986;140:72–73.
72. Vanasse M, Bedard P, Andermann F. Shuddering attacks in chil-
dren: an early clinical manifestation of essential tremor. Neurology.
1976;26:1027–1030.
73. Kabakus N, Kurt A. Sandifer syndrome: a continuing problem of
misdiagnosis. Pediatr Int. 2006;48:622–625.
 Part V Complementary and Special Techniques
Nasopharyngeal, Anterotemporal,
and Sphenoidal Electrodes
ANDRES M. KANNER, TRAVIS STOUB, AND STEVE BILD
INTRODUCTION
Temporal lobe epilepsy (TLE) is the most frequent type of
epilepsy of focal origin affecting adults. It is a heterogeneous
type of epilepsy as the epileptogenic zone can be restricted to
mesial temporal structures and/or involve temporal lateral neo-
cortex. Accordingly, the yield of detection of epileptiform activ-
ity with scalp electrodes depends on the following variables: (i)
the source and extent of the epileptogenic area relative to the
position of the scalp electrodes, (ii) the electric field generated
by the epileptiform activity and the orientation of the electric
vectors, and (iii) the extent of propagation of the epileptiform
activity from mesial to temporal lateral regions. For example,
epileptiform discharges originating in temporal lateral neocor-
tex would be readily recorded with the standard 10-20 midtem-
poral scalp electrodes. On the other hand, such electrodes may
often fail to detect epileptiform activity of mesial temporal ori-
gin. For one, the amygdala nucleus is not a laminar structure
and, therefore, it generates very restricted electrical field poten-
tials. Likewise, the hippocampus, with its ‘‘double C shape,’’
appears as a closed loop electrically and generates epileptiform
discharges with a vertical dipole that could only be detected by
frontotemporal and temporal lateral electrodes if the epilepti-
form activity were to propagate to those regions.
Several studies have shown that the standard 10-20 interna-
tional system of scalp electrode placement may often fail to
identify epileptiform activity of mesial-temporal lobe origin
(1–4). Indeed, the overall findings of these studies suggest that
the lateral and anterior portion of the temporal lobe is not ade-
quately covered using the standard 10-20 system. In the 10-20
system, electrodes F7, F8, T7, T8 (also known as T3-T4), and
A1-A2 have been used to record anterior temporal electrical
activity. Yet, electrodes F7 and F8 are positioned over the infe-
rior frontal gyrus and the rostral pars triangularis (1), while
electrodes T7 and T8 are positioned over the middle and supe-
rior temporal gyri posterior to the rolandic fissure (1). Hence,
the angle subtended by these electrodes relative to the mesial-
temporal source of the epileptiform discharges may often be
inadequate (lest it propagates to temporal lateral regions).
This problem was recognized in the first half of the 20th cen-
tury, which led to the development of sphenoidal electrodes (SE)
and nasopharyngeal electrodes (NPE). Later on, anterotemporal
(ATE), minisphenoidal (MSE), and basal-temporal (scalp) elec-
trodes (BTE) started to be used in an attempt to avert the incon-
venience and discomfort associated with the insertion of SE.
Studies carried out in the 1980s and 1990s reported that SE
and ATE identified a significantly greater number of epileptiform
CHAPTER
32
discharges that were not recorded with the standard temporal
electrodes included in the 10-20 system. The noninvasive and
user-friendly placement of ATE led to their use in most epilepsy
centers. Yet, in the last two decades an ongoing debate has been
taking place on whether SE and NPE add any additional data to
those yielded by ATE. The purpose of this chapter is to review
the conditions in which SE, NPE, and ATE should be used in
patients suspected of suffering from TLE of mesial-temporal
origin.
TECHNICAL ASPECTS
Sphenoidal Electrodes
Jones was the first to report the use of SE in 1950 (5). At that
time, the electrodes consisted of a fine insulated needle electrode
inserted under local anesthesia. The point of entry was under the
zygomatic arch through the mandibular notch and targeted a
position under the base of the skull lateral to the foramen ovale
(FO). In 1957, Marshall introduced SE made of flexible wire ini-
tially made of stainless steal to allow for longer periods of recor-
ding (6), particularly during presurgical EEG monitoring studies,
which over the following two decades underwent further modifi-
cations. More recently, SE consisted of multistranded silver and
platinum wires, the latter to avert its lateral displacement (7).
Typically, the insertion of SE is carried out at the bedside,
under local anesthesia with lidocaine. The SE wire is mounted
on a #22-gauge lumbar puncture needle and the recording end
is inserted into the lumen of the needle. The needle is inserted
through the skin at a point 3 cm anterior to the external audi-
tory canal, beneath the zygomatic arch, and through the
mandibular notch oriented in a 10 posterior direction target-
ing the lateral border of the FO (Fig. 32.1). The tip of the nee-
dle is advanced to a depth of 4 to 5 cm beneath the skin’s surface
or until the patient reports mandibular pain (8).
Yet, the blind insertion of SE can have several problems that
may negate any advantage of their use. First, when SE are
inserted blindly, the recording tips may end up in a position at
a distance from the FO, either anterior or posterior to it.
Recordings from SE positioned posterior to the FO are ham-
pered by recording through the thick (2 cm) petrous pyramid,
while a position anterior or lateral to the foramen has the dou-
ble disadvantage of recording through bone and at an increased
distance from target mesial-temporal structures. To overcome
this potential problem, Kanner et al. suggested the insertion of
SE under fluoroscopic guidance (Fig. 32.2, left and right panels)
(9,10). This technique ensured that the recording tip is
669
670 Part V ■ Complementary and Special Techniques

Figure 32.1 Phantom image of the head showing a

22-gauge carrier needle entering the skin (concentric target
lines, S) in front of the condylar process of the mandible
passing under the zygomatic arch (Z) and through the
mandibular notch (MN), en route to V3 emerging from the
foramen ovale (FO, seen on the edge). PP denotes the
pterygoid plate. The SE that is mounted on the superior
surface of the needle is not shown. (Reprinted with permis-
sion from Kanner AM, Ramirez L, Jones JC. The utility of
placing electrodes under the foramen ovale with fluoro-
scope guidance. J. Clin. Neurophysiol. 1995;12:72–81.)

positioned inferior to FO (Fig. 32.3). Furthermore, the variable Minisphenoidal Electrodes

shape of the sphenoid wing among patients can result in an
obstacle to the insertion of the needle when done blindly, which
in turn results in additional discomfort to patients. The inser-
tion under fluoroscopic guidance averts this problem. On the
other hand, the higher costs associated with the insertion of SE
under fluoroscopic guidance is a disadvantage of this insertion
technique.
MSE were introduced by Laxer in 1984 as an alternative to SE (11)
in a study of 100 patients, 34 of whom had epileptiform activity of
temporal lobe origin. In 21 of these patients, the epileptiform
activity was only detected by MSE. MSE consist of stainless steel
0.5 cm needles inserted under the zygomatic arch, yet the inser-
tion of these electrodes is associated with less discomfort than that
of SE (12). In a study of 100 patients, Buchhalter et al. found that

Figure 32.2 Left panel: Submental-vertex radiograph shows the normal relationship of the FO (open arrow)
to the foramen spinosum (arrowhead). The electrode/ needle system (solid arrow) is positioned anterior and
slightly lateral to the FO. (Reprinted with permission from Kissani N, Alarcon G, Dad M, et al. Sensitivity of
recordings at sphenoidal electrode site for detecting seizure onset: evidence from scalp, superficial and deep
foramen ovale recordings. Clin Neurophysiol. 2001;112:232–240.) Right panel: Submental-vertex radiograph
shows the position of both SE wires (open arrows) anterior and lateral to their respective FO (solid arrows).
(Reprinted with permission from Fenton DS, Geremia GK, Dowd AM, et al. Precise placement of sphenoidal
electrodes via fluoroscopic guidance. Am J Neuroradiol. 1997;18:776–778.)
 Chapter 32 ■ Nasopharyngeal, Anterotemporal, and Sphenoidal Electrodes 671

Figure 32.3 Coronal CT through FO, showing the rela-

tion between the SE (E) and the inferomesial temporal
lobe. M, mandible; A, amygdala; P, parahypocampal
gyrus; OT, occipitotemporal gyrus; I, inferior temporal
gyrus; V, third ventricle; NP, nasopharynx; L, lentiform
nucleus.

MSE and SE yielded comparative data in nine patients; however, electrical activity of anterotemporal origin (17). These ATE,
SE identified interictal epileptiform activity not detected by MSE labeled T1 and T2, are positioned “one centimeter above a point
in six patients (12). On the other hand, other investigators, how- one third of the distance along a line from the external auditory
ever, failed to find any advantage of MSE to ATE (3,13). meatus to the outer canthus of the ipsilateral eye.” A disadvan-
tage of T1 and T2 is that their locations are not proportional to
Nasopharyngeal Electrodes 10-20 or 10-10 system positions and thus display dispropor-
In 1948, Roubicek and Hill (14) introduced NPE, and their tionate voltages when utilized in bipolar derivations (Fig. 32.5).
increased yield of localization of epileptiform activity of mesial Accordingly, referential derivations may be preferable when
temporal origin was confirmed by several authors (3,14). using these electrodes (Figs. 32.6 and 32.7).
These electrodes consist of a solid silver rod with a small silver Homan et al. reported a study of 80 patients in whom a total
ball at the tip (Fig. 32.4) of 0.67 and 0.055 inches in diameter of 58 epileptiform foci were found. ATE significantly improved
for use in adults and children, respectively (15,16). The shaft is
flexible and malleable and is insulated up to the ball tip with
six or seven coats of insulex varnish. The distal 1 inch is angled
about 20 , but such angle can be changed to fit any nasal pas-
sage. Unlike SE, NPE can be inserted by EEG technologists
through the nasal cavity to target the posterior pharyngeal
wall. When positioned optimally, the ball tip is expected to lie
in an area beneath the foramen lacerum.

Scalp Anterotemporal and Basal-Temporal

Electrodes
In 1960, Silverman suggested a position for scalp electrodes
outside of the 10-20 international system aimed at recording

Figure 32.5 Position of ATE and BTE within the 10-10 electrode
Figure 32.4 Picture of nasopharyngeal electrodes. placement system.
672 Part V ■ Complementary and Special Techniques

Figure 32.6 Interictal recordings

of a patient with a partial seizure
disorder of right mesial temporal
origin. Interictal spikes are detected
only at SP2 in the bipolar montage,
while in the referential montages
sharp transients can be identified at
T2 and T4 electrodes in the first
two discharges.

Figure 32.7 Ictal recordings of a

patient with a partial seizure disorder
of right mesial temporal origin. The
ictal onset was identified 6 seconds
at SP2 before it was recorded at ATE.
 Chapter 32 ■ Nasopharyngeal, Anterotemporal, and Sphenoidal Electrodes
detection of the epileptiform activity of mesial-temporal origin
compared to the standard temporal electrodes (2). Other
authors confirmed these observations (3).
In addition to T1 and T2 electrodes, scalp BTE that are part
of the 10-10 system have been used with increasing frequency
to record epileptiform activity of mesial-temporal origin. Each
BTE is located in the coronal plane defined by their superior
counterparts (i.e., F9 and F10 are in the plane defined by F7, F3,
Fz, F4, and F8 [Fig. 32.4]). Interelectrode distances within the
coronal plane are determined by the superior electrodes as well
(i.e., F9–F7 is 2.5 cm when F7–F3 is 5 cm). Distances between
basal electrodes should be as equidistant as possible (T9 and
T10 will usually have to be placed either anterior or posterior to
their ideal position to avoid the ear auricles).
CHOICE OF ELECTRODES
Several studies have compared the localizing yield of ATE, NPE,
SE, and MSE over the 10-20 scalp temporal electrodes in the
recording of interictal and ictal epileptiform activity of mesial-
temporal origin. These studies have confirmed that any of these
electrodes are superior to the standard 10-20 temporal elec-
trodes in the recording of epileptiform activity of mesial-tem-
poral origin.
Studies that have compared the effectiveness of spike detec-
tion between NPE and ATE showed that ATE are as good as or
better at detecting epileptiform activity than NPE (4,17,18). In
fact, one of these studies showed NPE detecting significantly
fewer spikes (4). In addition, most of these studies report
patient discomfort with NPE and more muscle and movement
artifact associated with NPE than ATE. Given these data, ATE
have replaced the use of NPE in most epilepsy centers.
When only two ATE are being placed, T1 and T2 may be
preferable to 10-10 system electrodes because of their ease of
placement. The closest counterparts to T1 and T2 in the 10-10
system are electrodes FT9 and FT10. Locating FT9 and FT10
requires the calculation of at least four other coordinates in the
10-10 system.
ATE and scalp BTE are included today in EEG and video-
EEG studies in patients with presumed epilepsy of mesial tem-
poral origin in all major epilepsy centers worldwide. A big
debate remains on whether SE add any localizing yield to the
interictal and ictal data obtained with ATE and BTE. Table 32.1
summarizes the data from studies that compared the localizing
yield of interictal and ictal foci with SE and ATE (9,10,19–28).
As seen from the table, some studies show some advantage of
SE over ATE, while others do not.
How can we explain such discrepant findings? Most epilepsy
centers insert SE under blind conditions, in which cases the
recording tips commonly may end up in a position at a distance
from the FO, either anterior or posterior to it. As stated above, the
petrous pyramid is interposed between the SE recording tip and
mesial-temporal structures when it ends up in a posterior posi-
tion to the FO, while a position anterior or lateral to the foramen
has the disadvantage of recording through bone and at an
increased distance from the presumed epileptogenic area. A large
electric field of interictal and ictal activity minimizes the impact
673
of longer electrode–generator distances and the effect of bone
attenuation. In such circumstances, recordings obtained with SE
positioned at a distance from the FO would look the same as
ATE/BTE recordings. In addition, SE placed anterior to the FO
and ATE/BTE yield comparable subtended angles to mesial-basal
generators and, hence, epileptiform activity would not differ at
either electrode, no matter the extent of its electric field.
Thus, the question that needs to be asked is not “whether SE
are superior to ATE or BTE in recording epileptiform activity of
mesial-temporal origin” but “in what circumstances should SE
be used?” Data from three studies carried out by one of the
authors (AMK) may provide answers to these questions
(9,10,28). The first study included 17 patients with TLE of
mesial-temporal origin in which SE inserted blindly (SEB) and
ATE failed to identify a focal ictal onset (9). Recordings with SE
inserted under fluoroscopy (SEF) detected a unilateral
anterotemporal interictal focus in four patients in whom
SEB/ATE had failed to record any interictal spikes and detected
bitemporal independent interictal foci in one patient where
SEB/ATE had only identified unilateral spikes. In nine other
patients, the spike count obtained with SEF recordings
increased by more than 100%, relative to the spike count
obtained with SEB/ATE recordings. In addition, SEF recorded
seizures with an anterotemporal focal onset pattern in 10
patients where SEB/ATE had failed to do so. These data raised
the hypothesis that the detection of interictal and ictal activity
with only SEF was related to a restricted electric field of the
activity recorded with SEF (Figs. 32.6 and 32.7).
To test this hypothesis, a second study was carried out
using the interictal data of the previous study. The study con-
sisted of a comparison of spike voltages at SEF, SEB, and ATE
in sets of five randomly selected spikes per interictal focus,
recorded in the course of the two monitoring studies (10).
The voltage differences were expressed as ratios, VATE/SEF and
VATE/SEB and for each spike set, a mean ratio was calculated.
The spike voltages were almost identical at SEB and ATE
(mean VATE/SEB = 0.94), while they were significantly higher at
SEF than at ATE (mean VATE/SEF = 0.66; P 0.001). In the SEF
recordings, the narrowest electric field contours were found
among interictal foci that had only been recorded with SEF,
which, in turn, were significantly smaller than those of inter-
ictal foci in which SEF yielded an increment of more than
100% in spike counts (relative to those obtained with
SEB/ATE). These, in turn, had a significantly smaller contour
than those of the interictal foci where SEF failed to yield any
advantage over SEB/ATE (P 0.001). In addition, VATE/SEF of
interictal foci in which SEF recorded seizures with a focal
onset pattern were significantly lower than those of foci where
SEF failed to do so (P = 0.016). Finally, VATE/SEF did not differ
from VATE/SEB among interictal foci where SEF failed to yield
any advantage over SEB/ATE in either interictal (P = 0.240) or
ictal (P = 0.311) recordings. The data of the second study con-
firmed our hypothesis and proved that SEF yield additional
localizing data when recording epileptiform activity with a
restricted field, provided that its recording tip is positioned
below the FO. When distant from the FO, SE can be expected
to yield comparable data to those obtained with ATE.
674 Part V ■ Complementary and Special Techniques
Tabl e 3 2 . 1
Comparison of Interictal and Ictal Recordings among ATE, SE, and NPE
Authors
Sperling and Engel (19)
Binnie et al. (20)
Sperling et al. (21)
Binnie et al. (22)
Sperling and Engel (23)
Fernández Torre et al. (24)
Ives et al. (25)
Wilkus et al. (26)
Pacia et al. (27)
Kanner et al. (9)
Kanner and Jones (10)
Mintzer et al. (28)
Electrodes Compared
Interictal discharges at sphenoidal,
earlobe, and nasopharyngeal electrodes
Interictal discharges at multipolar
sphenoidal and surface electrodes
(near current T1 and T2 sites)
Interictal discharges at sphenoidal and
invasive electrodes
Interictal discharges at sphenoidal
electrodes stereotactically placed into
the foreman ovale and surface electrodes
Ictal onset at sphenoidal and scalp
electrodes
Ictal onset at sphenoidal, infrazygomatic,
anterior temporal, and surface temporal
electrodes
Ictal onset at sphenoidal and T1/T2
electrodes
Ictal onset at sphenoidal, T1/T2,
and anterior temporal electrodes
Ictal onset at foreman ovale–placed
sphenoidal electrodes and anterior
temporal electrodes
Interictal and ictal activity with
sphenoidal electodes placed with
fluoroscopic guidance just below the
foramen ovale and standard
blind–placed sphenoidal electrodes
Interictal and ictal activity at sphenoidal
electodes placed with fluoroscopic
guidance just below the foramen ovale,
standard blind–placed sphenoidal
electrodes, and anterior temporal
electrodes
Ictal onset at sphenoidal electodes placed
with fluoroscopic guidance just below
the foramen ovale and anterior temporal
electrodes
Results
99% of discharges were detected with sphenoidal,
57% with nasopharyngeal, and 54% with
earlobe electrodes
Sphenoidal electrodes offer no improvement in
sensitivity compared with surface electrodes
Sphenoidal electrodes were synchronous with
invasive electrodes in the neocortical areas but
not in the amygdala or hippocampus
4% of discharges were seen at sphenoidal electro-
des only, while 3% were seen at surface only
19% of the seizure onset was seen only at
sphenoidal electrodes. The sphenoidal electro-
des led the scalp by more than 5 seconds in
70% of the seizure onsets
Ictal onset was seen earliest at the
sphenoidal and then at infrazygomatic and
anterior temporal leads
40% of the seizures were seen first at sphenoidal
electrodes and 60% of the seizures were seen at
both recording sites
No difference was seen between electrodes
Improved identification and timing at the
sphenoidal electrodes 7% of the time
Fluoroscopically placed sphenoidal electrodes
added additional information such as increased
spike count and better seizure onset
localization
Significantly higher voltage spikes and significan-
tly narrower electric field contour were seen
during ictal and interictal activity recorded
from fluoroscopic-guided sphenoidal
electrodes adding additional data
Fluoroscopically placed sphenoidal electrodes
increased rater reliability, increased the number
of seizures correctly localized, and detected
seizures 5 seconds before they were seen at
the anterior temporal electrodes in 43% of
the seizures
 Chapter 32 ■ Nasopharyngeal, Anterotemporal, and Sphenoidal Electrodes
The question of when should SE be inserted was addressed
in a third study by determining the frequency with which SEF
detected localizing data not obtained with ATE/BTE recordings
in a group of 40 unselected patients whose presurgical evalua-
tion had demonstrated a unilateral (n = 32) seizure focus or
bilateral independent (n = 8) seizure foci in anterotemporal
regions (28). The study consisted of a comparison of the ictal
onset pattern of 156 seizures obtained with SEF and ATE/BTE
by four electroencephalographers who reviewed the 312 ictal
recordings independently, and blind to the patients’ identity, to
any presurgical data, and to whether the recordings’ montage
included ATE or SEF. Inter-rater agreement among the four
raters was significantly greater with SEF than ATE recordings
(P 0.0001). The number of seizures correctly localized by at
least three raters was significantly greater with SEF (n = 144)
than ATE (n = 99; P 0.0001). All the seizures (n = 36, or 23%)
originating from 14 ictal foci (29%) in 11 patients (27.5%) were
only localized with SEF. This was a significantly more common
occurrence in patients with bilateral independent ictal foci (P =
0.04). The ictal onset was detected at SEF five or more seconds
earlier than at ATE in 67 seizures (43%) originating from 33
(69%) foci in 30 patients (75%). In addition, SEF had a signifi-
cant advantage over ATE in the recording of seizures of patients
with bilateral independent foci and with a normal MRI, but not
in seizures of patients with a structural lesion. These data sug-
gested that SEF facilitate the localization of anterotemporal ictal
onset and in about one-fourth of patients SEF identify ictal data
not localized or mislocalized with ATE recordings. The follow-
ing conclusions can be derived from these data:
(1) If SE are to be inserted, they should be placed under fluo-
roscopic guidance.
(2) SE need not be inserted in all patients with suspected TLE
of mesial temporal origin.
(3) The indications to insert SEF are as follows.
In case of hippocampal atrophy, the decision to insert SEF can be
delayed until the first seizure is recorded and the interictal
recordings of the first 24 hours have been reviewed. SEF will not
be necessary in the case of patients with unilateral atrophy if
ATE recordings demonstrate:
• A majority of spikes ipsilateral to the atrophic hippocampus.
• The first seizure has a focal or regional onset pattern (pro-
vided that all of the patient’s seizures are identical by history).
However, SEF should be inserted:
• If ATE interictal recordings reveal bilateral independent spike
foci with a relative frequency of 50/50 to 60/40.
• If ATE recordings reveal an ictal onset pattern that is lateral-
ized or not localized.
• In case the patient’s seizures are not identical by history
(excluding complex partial seizures that also evolve to sec-
ondarily generalized tonic–clonic seizures).
• In case of bilateral hippocampal atrophy.
In case of a unilateral anterotemporal structural lesion, the
same criteria as outlined above can be used.
In case of a normal brain MRI, SEF should be inserted.
675
THE ADVANTAGE OF JOINT USE
OF SE AND ATE/ BTE
As stated above, epileptic seizures of TLE may originate in
mesial temporal structures or temporal lateral neocortex. With
the widespread availability of digital EEG systems, a spatial-
temporal analysis of epileptiform discharges with SE and ATE
may help establish the source in mesial structures by demon-
strating the initial negativity at the SE. On the other hand, a
synchronous discharge at SE and ATE/BTE does not exclude a
mesial source.
CONCLUDING REMARKS
The recordings of epileptiform activity of presumed mesial
temporal origin should always include ATE/BTE. The available
data do not appear to support the use of NPE, they do not
appear to yield any advantage over ATE/BTE, and they are asso-
ciated with some discomfort. SE should be considered in special
circumstances, and their insertion should be done under fluo-
roscopic guidance.
REFERENCES
1. Homan RW, Herman J, Purdy P. Cerebral localization of the 10-20
system electrode placement. Electroencephalogr Clin Neurophysiol.
1987;66:373–382.
2. Homan RW, Jones MC, Tawat S. Anterior temporal electrodes in
complex partial seizures. Electroencephalogr Clin Neurophysiol.
1988;70:105–109.
3. Goodin DS, Aminoff MJ, Laxer KD. Detection of epileptiform
activity by different noninvasive EEG methods in complex partial
epilepsy. Ann Neurol. 1990;27:330–334.
4. Sadler RM, Goodwin J. Multiple electrodes for detection of spikes
in partial complex seizures. Can J Neurol Sci. 1989;16:326–329.
5. Jones DP. Recording of the basal electroencephalogram with sphe-
noidal electrodes. Electroencephalogr Clin Neurophysiol.
1951;3:100.
6. Marshall C. Sphenoidal electrodes: an effort toward their popular-
ization. Electroencephalogr Clin Neurophysiol. 1957;9:379–382.
7. Rovit RL, Gloor P, Rasmussen T. Sphenoidal electrodes in the elec-
trographic study of patients with temporal lobe epilepsy: an eval-
uation. J Neurosurg. 1961;18:151–158.
8. Ives JR, Gloor P. Update: chronic sphenoidal electrodes.
Electroencephalogr Clin Neurophysiol. 1978;44:789–790.
9. Kanner AM, Ramirez L, Jones JC. The utility of placing electrodes
under the foramen ovale with fluoroscope guidance. J Clin
Neurophysiol. 1995;12:72–81.
10. Kanner AM, Jones JC. When do sphenoidal electrodes yield addi-
tional data to that obtained with antero-temporal electrodes?
Electroencephalogr Clin Neurophysiol. 1997;102:12–19.
11. Laxer KD. Mini-sphenoidal electrodes in the investigation of
seizures. Electroencephalogr Clin Neurophysiol. 1984;58:127–129.
12. Buchhalter J, Schomer DL, Schachter S. A comparative study on
the sensitivity of the mini and standard sphenoidal electrodes.
J Epilepsy. 1991;4:29–32.
13. Chu NS. Surface sphenoidal electrode for recording anterior tem-
poral spikes. Clin Electroencephalogr. 1992;23(4):190–195.
14. Roubicek J, Hill D. Electroencephalography with pharyngeal elec-
trodes. Brain. 1948;71:77–87.
676 Part V ■ Complementary and Special Techniques
15. deJesus PV Jr., Marsland WS. The role of nasopharyngeal elec-
trodes in clinical electroencephalography. Neurology. 1970;20:
869–878.
16. Kashing DM, Celesia GG. Nasopharyngeal electrodes in the diag-
nosis of partial seizures with complex symptoms. Arch Neurol.
1976;35:519–520.
17. Silverman D. The anterior temporal electrode and the ten-twenty
system. Electroencephalogr Clin Neurophysiol. 1960;12:735–737.
18. Nowack WJ, Janati A, Metzer WS, et al. The anterior temporal elec-
trode in the EEG of the adult. Clin Electroencephalogr. 1988;4:
199–204.
19. Sperling MR, Engel J Jr. EEG from the temporal lobes: a compari-
son of ear, anterior temporal and nasopharygeal electrodes.
Neurology. 1985;17:510–513.
20. Binnie CD, Dekker E, Smit A, et al. Practical considerations in the
positioning of EEG electrodes. Electroencephalogr Clin Neuro-
physiol. 1982;53:453–458.
21. Sperling MR, Mendiuys JR, Engel J Jr. Mesial temporal spikes: a
simultaneous comparison of sphenoidal, nasopharyngeal and ear
electrodes. Epilepsy. 1986;27:81–86.
22. Binnie CD, Marston D, Polkey CE, et al. Distribution of temporal
spikes in relation to the sphenoidal electrode. Electroencephalogr
Clin Neurophysiol. 1989;73:403–409.
23. Sperling MR, Engel J. Sphenoidal electrodes. J Clin Neurophysiol.
1986;3:67–73.
24. Fernández Torre JL, Alarcón G, Binnie CD, et al. Comparison of
sphenoidal, foramen ovale, and anterior temporal placements for
detecting interictal epileptiform discharges in presurgical assess-
ment for temporal lobe surgery. Clin Neurophysiol. 1999;110:
890–895.
25. Ives JM, Drislane FW, Schachter SC, et al. Comparison of coronal
sphenoidal versus standard anteroposterior temporal montage in
the EEG recording of temporal lobe seizures. Electroencephalogr
Clin Neurophysiol. 1996;98:417–421.
26. Wilkus RJ, Vossler DG, Thompson PM. Comparison of EEG
derived from sphenoidal, infrazygomatic, anterior temporal, and
midtemporal electrodes during complex partial seizures.
J Epilepsy. 1993;6:152–161.
27. Pacia SV, Jung WJ, Devinsky O. Localization of mesial temporal
lobe seizures with sphenoidal electrodes. J Clin Neurophysiol.
1998;15:256–261.
28. Mintzer S, Nicholl JS, Stern JM, et al. Relative utility of sphenoidal
and temporal surface electrodes for the localization of ictal onset
in temporal lobe epilepsy. Clin Neurophysiol. 2002;113:911–916.
29. McClean PD. A new nasopharyngeal lead. Electroencephalogr Clin
Neurophysiol. 1949;1:110–112.
30. Bach-Y-Rita G, Lion J, Reynolds J, et al. An improved nasopharyn-
geal lead. Electroencephalogr Clin Neurophysiol. 1969;26:220–221.
31. Kissani N, Alarcon G, Dad M, et al. Sensitivity of recordings at
sphenoidal electrode site for detecting seizure onset: evidence
from scalp, superficial and deep foramen ovale recordings. Clin
Neurophysiol. 2001;112:232–240.
32. Kanner AM, Parra J, Gil-Nagel A, et al. The localization yield of
sphenoidal and anterior temporal electrodes in ictal recordings: a
comparison study. Epilepsia. 2002;43:1–8.
33. Fenton DS, Geremia GK, Dowd AM, et al. Precise placement of
sphenoidal electrodes via fluoroscopic guidance. Am J Neuroradiol.
1997;18:776–778.

Intracranial Monitoring: Depth,
Subdural, and Foramen Ovale Electrodes
MARGITTA SEECK, DONALD L. SCHOMER, AND ERNST NIEDERMEYER
INTRODUCTION
In most Western societies, epilepsy affects around 0.7% of the
population, that is, 2.1 million in the United States, 3.5 million
in Europe, and 47 million worldwide are affected. In about 20%
to 30%, drug treatment cannot control seizure occurrence and
the possibility of surgical treatment is then considered. Although
the optimal profiles of surgical candidates continue to be
defined, the precise localization of a focal onset remains a cor-
nerstone of successful surgical treatment. In many patients,
localization requires EEG recording directly from the cortex.
This is done either by peroperative corticography, that is, directly
from the brain in the operating room (see Chapter 34), or by
chronically implanted electrodes for a variable duration lasting
from a few days to several weeks. The major goals of intracranial
EEG (IEEG) recordings are to determine the seizure onset and
distinguish the focus from vital cortex (Table 33.1). There are no
true evidence-based studies on the yield of chronic IEEG moni-
toring, that is, studies assigning subjects randomly to “intracranial
Tabl e 3 3 . 1
Requirements and Goals of Intracranial Monitoring
• Noninvasive exams are inconclusive concerning the site
and extent of a surgical intervention
• The clinical status of the patient is compatible with
prolonged intracranial recordings, for example, no major
behavioral disturbances, absence of psychosis, and
bleeding disorder
• The patient accepts the procedure and the risk of
inconclusive results after intracranial monitoring
• The precise focus identification requires the following
information:
• Lateralization of onset (e.g., left or right temporal lobe
epilepsy)
• Identification of lobe of onset (e.g., temporal vs. frontal
lobe epilepsy)
• Precise region of onset (e.g., left anterior vs. posterior
temporal onset)
• Localization of vital or eloquent cortex in adjacent cortex
is mandatory (e.g., determination of language cortex in a
patient with left posterior temporal lobe seizure onset)
CHAPTER
33
monitoring” or “no intracranial monitoring.” However, the
yield of precisely mapping the cortex in terms of focus localiza-
tion and vital cortex is evident even without formal studies.
Most patients who are candidates for IEEG fall into an inter-
mediate category of operability: a focal onset is suspected or at
least not completely ruled out on the basis of the results of non-
invasive EEG exams, but immediate resective surgery is consid-
ered too risky in terms of neurologic complications. Statistically
speaking, the need for IEEG is an unfavorable predictor for
postoperative success (1).
The definition of candidacy for IEEG has changed over the
past decades, due in large part to more sophisticated brain
imaging. Today, in most centers, a patient with a glioma in the
anterior mesial temporal lobe and consistent EEG findings
would not be considered an IEEG candidate, even if he/she had
contralateral interictal spikes. In contrast, a patient with multi-
ple lesions and discordant EEG findings, or with no MRI abnor-
mality at all, would be considered an appropriate IEEG
candidate. However, to the best of our knowledge, there are no
published consensus guidelines regarding the utility of IEEG.
The need for invasive recordings has been discussed extensively
for different clinical scenarios for patients with temporal lobe
epilepsy (TLE). It has been proposed that the concordance of
unilateral hippocampal atrophy and interictal EEG is sufficient
to proceed to surgery, that is, even without (scalp) ictal record-
ings, which may be sometimes misleadingly showing bilateral
or even contralateral epileptic discharges (2). However, some
centers would envision not only ictal recordings but even IEEG
in order to determine unambiguously EEG onset (3). In any
case, careful evaluation of each individual patient is mandatory,
including functional imaging and neuropsychological data.
Ictal scalp recordings help to better identify the patient’s seizure
semiology, verify the presumed ictal onset by analysis of the
clinical picture, and estimate the burden of epilepsy as well as
following up on postoperative persistent seizures. Apart from
persistent epileptic seizures, postoperative nonepileptic (psy-
chogenic) seizures have been described in 2% to 8%. These can
be difficult to diagnose if the preoperative semiology is not
known (4).
Despite improvements to our understanding about who ben-
efits most from this otherwise costly and labor-intensive proce-
dure, each center has its own strategies, which depend on the
experience of the epilepsy surgery team, the availability and qual-
ity of the noninvasive exams, and overall national health care sys-
tems that facilitate (or not) the access to the highest level epilepsy
surgery centers and IEEG. Requirements for the equipment and
staffing of such centers have been published (5). The diversity in
677
678 Part V ■ Complementary and Special Techniques
presurgical epilepsy evaluation across nations or continents
became evident in the recent survey. The Pediatric Epilepsy
Surgery Subcommittee of the International League Against
Epilepsy conducted a survey in Europe, the United States, and
Australia (6). It appeared that in the United States children are
more often evaluated with intracranial electrodes, even in the
presence of an MRI lesion. There is also more access to functional
imaging in the United States compared to their European and
Australian counterparts. However, more patients are treated with
a vagal nerve stimulator, a palliative procedure, in the United
States compared to Europe and Australia. Although other charac-
teristics of the patient populations were not different (underlying
pathology, affected lobe, etc.), the reasons for these differences are
not entirely clear but are probably of multiple origins, which
include different national “schools” with different algorithms of
epilepsy care or different regulations from health care providers.
However, while sophisticated epilepsy centers exist today in
all Western countries, it should not be forgotten that there are
still countries where epilepsy surgery is not done or is only
rarely available, which prevents patients from obtaining a
potentially very effective epilepsy treatment (7).
In the past 10 to 20 years, the development of new noninva-
sive imaging techniques was readily embraced in the presurgical
epilepsy evaluation (8–13), starting with the entry of high-def-
inition MRI and special imaging protocols for epilepsy patients
(14). Indeed, the knowledge obtained from these techniques has
given us patient populations and helped to specify patient pop-
ulations where IEEG is more likely to be diagnostic. For exam-
ple, 10 years ago, invasive monitoring was considered in
patients with hippocampal sclerosis, but bilateral temporal
spikes or seizures. However, the presence of an MRI abnormal-
ity is a good prognostic indicator for seizure-free outcome post-
operatively (15), and patients with bilateral spikes or even
seizures may very well benefit from surgery on the sclerotic hip-
pocampus (16,17), even without IEEG. However, truly bitem-
poral lobe epilepsy, as determined by IEEG and MRI, is
associated with a lower success rate and only 50% became
seizure-free according to a recent study (18). On the other
hand, severe unilateral hippocampal atrophy, the so-called
burned-out hippocampus, may present with aberrant, even
contralateral, scalp ictal discharges. Those patients nevertheless
benefit from surgery (19). In some of these patients, IEEG was
used to verify ictal onset in the affected hippocampus.
More recently, it has become evident that the presence of a
localized or lateralized temporal hypometabolism equals the
presence of an MRI lesion. Both MRI lesional and nonlesional
TLE have the same postoperative prognosis, that is, a chance of
seizure control of approximately 80%, if the temporal hypome-
tabolism in the PET is obvious and unilateral (20). Thus, if this
finding is confirmed in more patients, nonlesional TLE patients
may not need IEEG anymore if they underwent good-quality
PET-imaging and its results are concordant with the remainder
of the evaluation. Over the next few years, other investigative
measures such as studies on other pathologies, for example,
dysplasia or tuberous sclerosis, identifying reliable noninvasive
markers of the epileptogenic zone may further reduce the over-
all number of patients who need IEEG (21–23).
BASIC CONCEPTS
IEEG should be considered if it is felt that surgery can cure or
at least significantly ameliorate the epilepsy disorder, but perti-
nent localizing information is still missing from the noninva-
sive evaluation. This implies that every patient should first
undergo a noninvasive evaluation in order to understand the
size and nature of the problem. A clear clinical hypothesis is
mandatory before implantation, because implanted electrodes
record only from a limited cortical volume. Although the exact
volume is not known, there are estimates indicating that a cor-
tical area of 10 to 20 cm 2 is necessary to generate a scalp spike
(24). Detailed analysis of scalp discharges and the simultane-
ously recorded subdural EEG showed that only the early com-
ponents of spikes and seizures arise from a limited volume,
which is, however, still larger than 10 cm 2. With further propa-
gation, the active brain areas attain 30 cm 2 or more (25). In this
context, imaging methods that suggest point sources of the
underlying spikes represent rather a simplification than the true
estimation of the size of epileptogenic cortex (26).
Moreover, the surgical act and the fact that the electrodes
remain in place for several days or weeks are associated with a
significant risk of morbidity but a relatively small risk for mor-
tality (see below), requiring careful elaboration of the goal of
the investigation and the sites of implantation.
Several concepts need to be recognized and defined as much
as possible before any epilepsy surgical procedure is proposed
(27). The epileptogenic area, that is, the area whose resection is
necessary and sufficient to abolish further seizures, needs to be
clearly identified. This zone is not to be confounded with
• the lesional zone, that is, the area that appears morphologi-
cally altered but may be smaller or even remote to the epilep-
togenic zone
• the irritative zone, that is, the area of interictal spikes, which
can be more widespread than the epileptogenic zone and
include even contralateral structures
• the symptomatic zone, that is, the area whose activation pro-
duces the clinical symptoms, such as nausea or dystonic
posturing
• the functional deficit zone, that is, area whose activation/
deactivation leads to neurologic or neuropsychological
deficits, for example, verbal memory impairment or Todd
paresis
Overall, IEEG data look different compared to scalp EEG data
and appear to be more “epileptogenic,” due to lack of the filter
qualities of the skull, although the same rhythms known from
scalp EEG can also be retrieved in the IEEG. The determination
of ictal EEG pattern is usually at the center of the investigation;
however, careful analysis of interictal epileptogenic discharges is
probably more important than previously appreciated.
The overall yield of IEEG depends on the criteria that deter-
mine at which degree of inconsistency intracranial evaluation is
offered. There are only very few data on this subject. A recent
study, with access to modern imaging tools, described the
results in 77 patients with mesial temporal (39%), lateral tem-
poral (12%), and extratemporal (43%) epilepsy (12). In the
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 679

remaining five patients (6%), seizures could not be captured Today, depth and subdural electrodes are the most estab-
and they remained unclassified. Localizing results were lished electrodes types. Most centers have a preference for one
obtained in 74%, which corresponds to numbers from other or the other electrode type, which depends on personal experi-
studies (28,29) and to the authors’ experience. Thus, despite the ence or technical equipment in the neurosurgical clinic. Around
invasive nature of the procedure and the physical and psycho- the same time of the arrival of subdural electrodes, foramen
logical investment by the patient, it is not always certain that ovale electrodes (FOE) (36) and epidural peg electrodes were
IEEG will be followed by a resective procedure. In around 25%, described by the Zürich group of Wieser et al. and the
surgery cannot be proposed, which needs to be addressed and Cleveland Clinic group of Lüders et al. (37), respectively. The
reiterated during preparatory talks with the patients. use of the latter two types remains relatively limited and is often
used as an intermediate step in determining if, and which,
HISTORY larger set of intracranial electrode setups will be more adequate
or as complementary electrode type, if depth or subdurals alone
IEEG focus evaluation has a relatively brief history, starting in do not appear sufficient.
Germany with Berger’s discovery of the EEG published in 1929
(30). Otfrid Foerster used preoperative EEG during epilepsy
surgery and became a pioneer in using EEG for the determina- DETERMINATION OF PRECISE
tion of the epileptogenic zone. Foerster worked with Wernicke ELECTRODE PLACEMENT
and published a brain atlas, and as an aside he was the treating
Independent of the chosen electrode type, knowledge of the
neurologist for Lenin when he suffered a stroke in 1922. The ris-
exact position of the electrodes with respect to the individual
ing Nationalist government basically destroyed Germany’s sci-
brain anatomy is of utmost importance for successful surgery in
entific culture for the next decades and many researchers left
terms of complete focus removal while sparing eloquent cortex.
the country. Dr Foerster was the mentor of Wilder Penfield who
Previously, only radiographs of the skull were possible. These
himself became widely known for the corticography of the
provided only a gross idea of the electrode position (38,39).
human neocortex. Thanks to his research on patients, the corti-
Now 3D reconstruction of CT and MRI with coregistration of
cal representations of the motor cortex (homunculus) became
X-rays provides superior electrode localization results (39).
standard knowledge. Dr Penfield was a neurosurgeon and
Coregistration of CT, MRI before and after electrode place-
founded the Montreal Neurological Institute. Herbert Jasper
ment, or digital photos taken intraoperatively with the patient’s
joined him as a neurologist and clinical neurophysiologist, and
MRI (40–45) provides excellent results in terms of electrode
together they trained numerous neurosurgeons and neurolo-
recording sites localization. Mean mismatch between real and
gists in the field of epilepsy surgery. They probably represent
virtual electrode placement was found between 2 and 4 mm
the first example of a successful relationship between a neurol-
(42,44). Additional external fiduciary markers improve this pre-
ogist and a neurosurgeon. This combination remains a prereq-
cision (44). Coregistration of different MR sequences, notably
uisite for any successful epilepsy surgery team (31).
Flair and T2, allow the visualization of smaller lesion inside the
Wires and a type of “strip” electrode, placed in the epidural
OR (46). Implantable fiduciary markers were also used, which
space, were the first to be used in epilepsy patients by Jasper and
led to an even smaller localizing error (~0.5 mm) (47). Other
Penfield. Later, subdural electrodes and other electrode types
difficulties inherent to all methods of coregistration between a
were developed and applied (31). Bancaud and Talairach, another
preimplantation and a postimplantation image are the transi-
successful team of neurologist and neurosurgeon from Hôpital
tory changes of the brain anatomy due to cerebral edema,
Sainte-Anne in Paris, introduced multicontact depth electrodes
hematoma, and brain shift. In most algorithms, the electrode
penetrating directly into the cerebral tissue (32,33), allowing a
artifacts are used to identify their exact localization, on the basis
systematic investigation of deep structures. This was a major step
of either a CT or an MRI. In a second step, the extracerebral tis-
forward in epilepsy surgery investigations, since at that time MRI
sue (skull, scalp) is stripped away, and brain reconstructed in a
was not available and the evaluation of surgical candidates relied
3D fashion, showing the surface with the individual gyri and
more on neurophysiologic methods. Ten years later, Wyler devel-
sulci. In Figure 33.1, coregistration algorithm is proposed.
oped and introduced another electrode type into the field: the
The exact localization of depth electrodes is simpler, given
subdural electrode, which was being placed onto the brain’s sur-
that there is less of brain shift. Otherwise, similar algorithms to
face without the risks of an intracerebral insertion (34). Initially,
reconstruct the position of the electrodes are employed.
strips with four electrodes were described, evolving later to more
complex and larger arrangements (see below).
Although the number of inserted electrodes steadily DEPTH ELECTRODES
increased, or at least became technically possible, simultaneous
recording of 100 electrodes was difficult for EEG systems, Technical Aspects
requiring the selection of recorded electrodes and even changes Taking all patient studies together, depth electrodes are proba-
during the evaluation. Upgrading of established systems and bly the most frequently used electrode types (Fig. 33.2), given
innovative setups has finally overcome this obstacle (35), and at that they were already in use in 1960s and 1970s in the medical
the time of the writing of this text, EEG systems with up to 256 field. They are inserted through Burr holes by stereotactic
channels are on the market. methods, using guidance from previously acquired brain
680 Part V ■ Complementary and Special Techniques

Figure 33.1 Coregistration of int -

racranial electrodes and the
patient’s individual MRI. Propo -
sition for a workflow aiming fusion
of the patient’s preimplantation
MRI with the postimplantation MRI
or postimplantation high-resolution
CT. Courtesy of L. Spinelli, PhD.
The preimplantation is normalized
and skull stripped, then fusioned
with the postimplantation MRI,
which underwent the same compu-
tation. In this example, particular
care is taken to compensate for the
brain shift due to compression
by subdural grids. This is less of
an issue in patients with depth
electrodes.

imaging. Depth electrodes are built as wires containing evenly
spaced electrode contacts of conductive material. The insertion
is done by a mandarin whose trajectory is determined by the
site of the target(s) through CT and/or MRI of the patient’s
brain. Sometimes additional arteriography or MR-angiography
is required in order to identify crucial vessels and diminish the
risk of hemorrhage. In many centers today, a stereotactic setup
is utilized using the patient’s MRI. More recently, a frameless
robot-guided insertion has been reported (48), and the future
will show if the application of such sophisticated methods will
decrease the risk of depth electrode placement and/or shorten
operation time significantly.
Electrode spacing is most often between 5 and 10 mm, but
any configuration, distance, and number of electrodes can be
chosen, that is, either at regular distances or only at sites that
cover the gray substance. Most electrodes contain 6 to 15
electrodes. The electrodes themselves can be rigid or flexible,
but most neurosurgeons prefer flexible electrodes, which may
lead to less precise electrode placement (due to involuntary
deviation from the initial target), but they have had a lower
risk of bleeding. Most companies that produce intracranial
electrodes allow the choice between platinum–iridium and
nickel–chrome alloys. Only the first combination is MR com-
patible, but they are somewhat more expensive. The elec-
trodes have multiple contacts (usually between 8 and 12) and
can be inserted orthogonally, that is, perpendicular to the lat-
eral surface or longitudinal-parasagittally. Oblique insertions
into the frontal lobe toward the basal ganglia have also been
employed. However, in all cases, there is only very limited
sampling of the neocortical–lateral cortex, which may require
additional subdural or scalp electrodes if the cortex needs
larger coverage.

Figure 33.2 Depth electrodes. Left: Coregistration of the

orthogonally inserted electrodes and the patient’s MRI.
The insertion sites of depth electrodes in both hemispheres
are marked in white. Upper right: Same patient as shown
on left, showing his depth electrode inserted in the left
amygdala. In this coregistration, each contact is identified
separately (light gray balls). Lower right: Longitudinal inser-
tion of depth electrodes, indicated in light gray.
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes
Given that this electrode type is the only one that goes into
the brain parenchyma, the major advantage of depth electrode
recordings is the opportunity to record from deep, buried cor-
tex, such as hippocampus, amygdala, or mesial frontal struc-
tures. Furthermore, the simultaneous sampling of mesial and
lateral cerebral structures, associated to an excellent EEG qual-
ity, is a favorable aspect of depth electrodes. They also permit
bilateral and multilobar sampling, but there is limited spatial
sampling in a given region, for example, identification of lan-
guage cortex and its extent is not possible with depth electrodes
alone. Another significant advantage is that they can be easily
removed, which can be accomplished without any anesthesia.
They can be removed even at bedside.
Risks
The major risk of using depth electrodes is intracranial or
intracerebral hemorrhage, which occurs in 1% to 4% (49). With
modern imaging, the rate seems to be lower. In a study of 50
patients with depth and subdural strip electrodes (50), no
deaths, no infections, and no new neurologic deficits were
reported. Ross et al. provided a review on previously reported
series of depth electrode patients, and when considering the
whole group of 1656 patients together, the relative risk was high-
est for intracranial hemorrhage (1.1%), followed by infection
(0.8%). In a recent review of 259 patients, a risk of symptomatic
intracranial hemorrhage of 1.2% and 0.7% for permanent neu-
rologic deficit is reported (51). The Montreal group reported the
absence of major complications in a total of 6415 electrode
implantations (491 patients) between 1976 and 2006, that is, a
time span before and after introduction of MRI. Hematoma and
infection were noted in 0.8% and 1.8%, respectively (52).
Overall, in this study, complications were more frequent in the
frontal lobe, perhaps due to the higher number of implanted
electrodes. Mortality was basically zero in the series after 1990.
Among the hemorrhages, only very few were life threatening. In
the authors’ experience, it should be less than 0.1%.
Patient counseling on the avoidance of antiplatelet drugs as
well as the implementation of preoperative coagulation param-
eters is recommended. There is no agreement on the use of pro-
phylactic antibiotic treatment in patients during depth
electrode investigations. If the contact between the intra- and
extracranial milieu is minimized with sterile bandages and the
patient is inhibited from scratching under the bandage, the risk
of infection is even less than 0.8%. Even years after depth elec-
trode investigations, discrete gliotic changes along the insertion
channel are seen in MRI images, but, to date, there is no evi-
dence that these changes become themselves epileptogenic or
are related to new neurologic or cognitive deficits.
SUBDURAL ELECTRODES
Technical Aspects
These electrodes were introduced somewhat later into the field
of epilepsy surgery than the depth electrodes and are used for
epidural recordings. When Penfield recorded epidurally from
the left parietotemporal region in a patient with posttraumatic
681
epilepsy in 1939 (53), he can be probably considered the first to
have used this electrode type. Subdural or epidural electrodes
consist of disk-shaped electrodes, embedded in transparent
silastic or Teflon material, arranged in rows or arrays of all
shapes (Fig. 33.3). As with depth electrodes, they are made of
platinum–iridium and nickel–chrome alloys, allowing MR
imaging only if they are made of platinum. The contacts are put
freehand on the cortex, subdurally on the pia mater. If neces-
sary, they can be reduced with sterile scissors or along perfo-
rated lines to smaller arrays directly in the OR. There are recent
reports that make use of neuronavigation and stereotactic
methods to improve the precision of placement, in particular
for strips, or for merging imaging data sets that help to take all
preoperative imaging information into consideration (54,55).
Similar to depth electrodes, determination of electrode posi-
tions by MRI or CT imaging is an important issue because
smaller or larger shifts may occur after the implantation, in par-
ticular with small arrays. The contacts are usually 5 to 10 mm
apart and can be arranged as one line (strip) or in a larger array,
for example, 8 8 contacts (grids of 64 electrodes) (Fig. 33.3).
Strip electrodes arranged in a single row can be inserted
through Burr holes. However, a large grid, for example, array of
4 4 or 8 8 electrodes, needs a full craniotomy. This is usu-
ally a more invasive procedure than the stereotactic procedure
employed for the placement of depth electrodes.
The main indication for subdural electrodes is the precise
localization of a seizure focus within a suspicious area. When
using grid arrays, two to three lobes can be covered with a single
grid, for example, frontal posterior, temporal, and parietal. This
requires that noninvasive evaluation narrows down the hypoth-
esis of the site of the focus to one or two lobes. Subdural elec-
trodes also allow for preoperative corticography, with the aim to
determine vital cortex. This procedure is described below in
section Special Considerations in Pediatric Patients and in a
study by Lesser et al. (118). By stimulation with short (2 to 5
Figure 33.3 Intraoperative subdural grid placement. Each contact
(platinum disk, 0.8 mm diameter) is recorded through a wire, which is
led out of the crane through a larger white cable (in the lower left of the
image; in this grid, bundles of eight wires each).
682 Part V ■ Complementary and Special Techniques
seconds) intermittent electrical currents, sensations or involun-
tary movements are produced, which supposedly reflect the
characteristics of the underlying cortex. Primary cortex can
thus be reliably identified. However, higher order cortical phe-
nomena, like language, are not always easily identified as shown
by Seeck et al. (56)
These electrodes do not necessarily require knowledge of
stereotactic procedures and/or a neuronavigation device and, as
such, can be performed in neurosurgic centers without experi-
ence in stereotactic methods. Subdural electrodes record from a
larger cortical surface and capture incompletely signals from
the sulci as well as from radial sources. It is a well-established
finding that only one third of the brain is accessible from the
surface, which also means only one third is accessible in terms
of recording. Thus, most brain tissue underneath a subdural
grid will not be directly recorded, but may be involved only after
propagation to these more superficial cortical regions has
occurred.
Risks
The morbidity is low for strip electrodes and includes infec-
tions, around 1%, or bleeding when bridge veins are damaged
(57). Larger grid arrays are more likely to be associated to more
significant complications. The infection rate is higher than with
depth or strip electrodes alone, and is reported to be around
8%. This risk can be diminished if the cables exit through a tun-
nel via a second incision a few centimeters away from the orig-
inal incision. As for depth electrodes, no consensus exists
whether or not prophylactic antibiotic therapy should be given
(57,63), albeit this issue is more relevant with subdural elec-
trodes. Based on personal experience, prophylactic antibiotic
treatment in patients with large arrays of subdural electrodes
was related to lower incidence of infectious problems, for exam-
ple, Cefazolinum 4 500 mg IV.
Most patients with grids experience local edema of the
underlying cortex, leading to headache and photophonopho-
bia. This complication occurs less often with strips. Rigorous
symptomatic treatment with antipain drugs as well as position-
ing of the head at a 30 to 45 elevation are helpful. Major or
near-fatal raised intracranial pressure is extremely rare, but if
discovered, may need immediate surgical decompression.
Consequently, a neurosurgeon on call nearby is a necessity.
In a recent study, delayed subdural hematoma was noted
to be the most frequent complication (8%) (58). They
occurred up to 3 days postimplantation. This “fragile” period
is probably somewhat longer and, as for most brain surgeries,
spans up to 5 days. A recent study in a pediatric group noted
postimplantation problems in 20% of their patients, with
predominance of hematoma, but this may be due to the fact
that 43% of their patients had already undergone a prior
epilepsy surgery procedure (59). Another pediatric series
reported a lower complication rate, including hematoma
(6%) (60). However, implantation of larger arrays of sub-
dural electrodes remains a relatively delicate procedure,
requesting unambiguous agreement on the indication, in
particular in children. Patients with a history of high-dose
radiotherapy and chemotherapy may develop significant
persistent neurologic deficits after subdural grid placement,
not explained by brain swelling (61). IEEG in this patient
group may be related to more severe complications than in
patients without this history.
Experience is also an important factor. The Cleveland Clinic
group reported 33% of minor and more serious complications
at the beginning of their epilepsy program, which dropped to
19% in more recent years. Overall, the use of more than 100
electrodes, longer recording sessions, that is, more than 2 weeks,
and more than one cable exit are related to an increase in com-
plication rates (62,63). It is probably safe to say that complica-
tions between 10% and 20% of the cases can be expected,
requiring close monitoring by specialized personnel and in a
center with 24 hours access to a scanner and an ICU. Regular
neurologic exams and monitoring for infectious diseases are
mandatory (Table 33.2).
If bleeding complications are present, it is safer to remove
the electrodes as soon as possible. If there are signs of local
infections, and if enough seizures are recorded, removal of
electrodes is recommended. The rising of C-reactive protein
(CRP), after attaining postoperatively normal or near-normal
values, is a valuable marker. If this is not the case, that is, no or
only too few seizures are recorded, introduction or reinforce-
ment of the antibiotic treatment with prolongation of the
recording period, perhaps with more rigorous drug and sleep
withdrawal, should be considered. Since intracranial electrode
implantation is a costly and intensive intervention, in terms of
manpower and, more importantly, physical and psychological
strain on the patients themselves, reimplantation at another
moment is not easily done or recommended. General guide-
lines are difficult to establish since every center has its own
resources, and patients differ in their willingness to cope with
unforeseen clinical situations.
Tabl e 3 3 . 2
Proposed Surveillance Algorithm for Patients
Admitted for Subdural Intracranial EEG Monitoring
• 24 hours supervision by nurses, technicians, and/or other
persons, for example, medical students or family members
• Postimplantation 24-hour surveillance in the ICU
• When admitted for EEG monitoring
• CRP, blood: every day or every second day
• Temperature: two times per day
• Brief neurologic exam (pupils, responsiveness, motor
functions) every 2–4 hours for the first 2 days, then
every 4–6 hours for the next 3–4 days
• Optional: prophylactic antibiotic treatment (e.g.,
cefazolin) starting from the day of the implantation
through the monitoring period
• Optional: dexamethasone during the first 3– 4 days
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes
Despite higher complication rates with subdural electrodes,
the rates of permanent neurologic deficits do not seem to differ
from the numbers for depth electrodes and are situated around
1% to 2% (63). The removal can be done at bedside, if strips
were used, but will require another craniotomy during which
the surgical resection of the focus is carried out as well, if indi-
cated, if grids are used.
FORAMEN OVALE ELECTRODES
AND EPIDURAL ELECTRODES
FOE, initially developed in 1985 by the Zürich group of Wieser
et al., are used somewhat less frequently in most centers (36).
FOE record epidurally and extracerebrally from the mesial tem-
poral lobe structures. They contain four to six contacts, and can
be inserted under local anesthesia (Fig. 33.4). They can also be
easily removed at bedside, similar to depth electrodes. Major
complications are rare, but they are not “semi-invasive” as ini-
tially advocated. Subarachnoidal hemorrhage and transient
brainstem symptoms, for example, trigeminal neuralgia, are
observed. Complication rates of 1% to 9% are reported, mostly
transient, that is, similar to rates reported for depth or subdural
electrodes. FOE yield good recordings from the mid and poste-
rior portions of the hippocampus, but less well from the ante-
rior aspect and the amygdala and/or small subcompartments of
the mesial temporal structures. The classical indication for FOE
is the lateralization of temporal lobe onset. They may be com-
bined with subdural or scalp electrodes for more extensive sam-
pling from lateral temporal or extratemporal structures, and
consequently the indication may be extended to other indica-
tions.
Epidural peg electrodes share configuration aspects with sub-
dural electrodes and are inserted through burr holes or twist
drill holes (37). They can be placed bilaterally, similar to scalp
electrodes, but cannot be used for cortical mapping. Electrical
stimulation through epidural electrodes is painful because the
rest of the dura is innervated and irritated by electrical current.
Precise sublobar focus localization is not possible with epidural
electrodes alone since their spatial sampling is not sufficient.
However, they have a low morbidity ( 0.5%), consisting
683
mainly of discomfort when the electrodes are inserted through
the temporalis muscle and some superficial wound infections.
Their main indication is the verification of the absence of
epileptogenic discharges and/or recording of selected sites
where scalp electrodes would be too contaminated by muscle
artifacts.
SURVEILLANCE
There is no consensus guidelines concerning the degree of med-
ical and paramedical staff surveillance and concomitant treat-
ment in patients admitted for EEG monitoring. However, a
“semi-intensive” care unit with a lower nurse–patient ratio
compared to normal care units appears the most appropriate
setting. There are still many centers that cannot offer profes-
sional surveillance at night or on weekends, leading to seizure-
related complications, for example, autonomous electrode
removal during the postictal confusional state.
The choice on whom to rely on during the patient’s monitor-
ing session is influenced by budget and legal considerations.
Although we know relatively well the risk of IEEG complications
such as bleeding or infection, there is disagreement across surgi-
cal epilepsy centers on who watches the patient: parent/close
family member; nonmedical staff, for example, out-of-work
people recruited on through a placement agency; EEG techni-
cians; or nurses. Patients who are recorded after drug with-
drawal carry a risk for more prolonged or more intense seizures
(i.e., more often generalized tonic–clonic seizures) resulting in
trauma and/or prolonged postictal confusion. Pulling out the
electrodes is also encountered. In a recent review, 3 out of 491
patients (0.6%) pulled out their electrodes during a seizure or
during the early postimplantation period (52). CSF leakage,
bleeding, or infection may result from such an act, particularly if
larger subdural electrodes arrays are removed.
Family members or nonmedical people may not be enough,
since they are often unable to judge the clinical context or sever-
ity of a given situation. There are no studies on the guidelines
on the size and composition of the surveillance staff necessary
to minimize these complications and to reduce the recording
time. For obvious reasons, professional surveillance by trained
Figure 33.4 Foramen ovale electrodes. Left: Insertion
procedure through the base of the skull (i.e., through the
foramen ovale). (Courtesy of CCND Winnipeg.) Right:
Foramen ovale electrodes. (Courtesy of Dixi electronics.)
684 Part V ■ Complementary and Special Techniques
nurses and technicians, in collaboration with an experienced
epileptologist or neurologist, is the optimal setting. Legal
aspects also need to be considered in order to be protected
against “under-surveillance,” and these differ markedly
between countries. In some countries, like the United States,
legal action against the hospital is more readily taken if compli-
cations are encountered, and therefore, different solutions and
their legal implications need to be more thoroughly studied.
The clinical observation of patients with IEEG should
include close monitoring of the neurologic status, cognitive
functions, and blood values depicting infections or bleeding.
The “algorithm” proposed in Table 33.2 stems from personal
experiences. Close monitoring for the first 5 days is recom-
mended by most of the neurosurgeons, due to an increased risk
of intracranial bleeding during this period. From then on, daily
neurologic surveillance may be sufficient, aiming particularly at
the detection of infectious complications.
Since the introduction of chronic IEEG, there is ongoing
debate if, when, and how long antibiotic therapy or corticos-
teroids are administered. Informal review of the protocols of
the different centers shows that changes of such protocols have
emerged. The most frequent pattern had the regular use of
steroids during the initial period. This pattern has been aban-
doned for most patients. In most centers, steroids and antibi-
otics are considered with subdural implants, but rarely
considered with depth electrodes. The placement of large grids
is almost always followed by focal cerebral edema of variable
degree and clinical symptoms, which usually improves sponta-
neously within 3 to 5 days postoperatively, or with steroid use.
Regarding the use of steroids, there are two studies on the
use of dexamethasone in patients undergoing monitoring with
subdural electrodes. The prophylactic use in children was felt
efficacious for brain swelling, as determined by CT changes
manifest by midline shift, decrease of ventricular size, and so
on, although it led to slightly longer (~1 day) seizure-monitor-
ing time (64). The benefit of steroids during a limited period of
3 days in adults seems to be less clear (65). In a randomized,
placebo-controlled study of 30 patients, less pain and nausea
were reported, but the effect was limited over time.
There is no study on the yield of antibiotics during IEEG.
Most centers consider prolonged antibiotics more with sub-
dural recordings as compared to depth electrodes. Other centers
give single or short-term period during and shortly after the
operation, while others use them also in patients with depth
electrodes. Up till now, no study supporting the view of better
protection against nosocomial infections with antibiotic ther-
apy exists.
COMPARISON OF YIELD BETWEEN
DIFFERENT ELECTRODE TYPES
Depth and subdural electrodes are the intracranial electrodes
that are most frequently used. However, each center has its own
preferences and experiences that may influence the choice of
one over the other.
Recordings from deep temporal structures, in particular from
the amygdala, yield better results with depth electrodes (66–68).
Subdural electrodes are also helpful in identifying a mesial
temporal focus. However, if they are placed or moved too later-
ally, in particular lateral to the collateral sulcus, false localization
and lateralization may occur due to propagation to orbitofrontal
or contralateral temporal cortex before recruitment of the
affected mesial temporal structure becomes visible (69).
Mesial extratemporal (i.e., interhemispheric) structures are
explored well with both depth and subdural electrodes.
Subdural electrodes are more easily placed given the larger cor-
tical surface compared to mesial temporal structures. However,
comparative studies do not exist.
In extratemporal lobe epilepsy, subdural electrodes seem to
be slightly better than depth electrodes (68), if convexity struc-
tures need to be recorded. Since they have the advantage to
allow cortical mapping, subdurals may be preferred if focus
localization includes mainly extratemporal cortex where there
is a question of the involvement and extent of eloquent cortex.
Combination of depth and subdural electrodes is also more and
more frequently used. However, comparative studies on the
yield of combined depth and subdural electrodes compared to
each electrode type alone are missing.
Comparisons between FOE and other electrode types have
also been performed. In seven patients with FOE and depth
electrodes, interictal and ictal discharges were picked up by FOE
if amygdala, hippocampus, and parahippocampal gyrus dis-
charged simultaneously, but less reliably if only a subcompart-
ment of the mesial temporal structures was active (70).
Intracranial electrodes served also as “gold standard” to
determine the accuracy of scalp and sphenoidal electrodes.
Compared to discharges retrieved in FOE, scalp electrodes over
the anterior temporal lobe (1 cm anterior to T1/T2) retrieved
95.6% of the discharges but midtemporal scalp electrodes
retrieved only 77.7% (T3/T4) (71). Rarely (3%), the interictal
discharges showed up in the anterior temporal scalp electrode
but not in the FOE. Discharges noted in sphenoidal electrodes
arose mainly in mesial temporal cortex, but not exclusively,
since in the group of 21 patients studied, 3 also had spikes ori-
ginating from the orbitofrontal and temporal neocortex (72).
INTERICTAL INTRACRANIAL EEG
All normal physiologic rhythms, for example, alpha, theta (dur-
ing drowsiness), and mu, which can be seen in scalp recordings,
appear also in IEEG at the corresponding cortical sites, that is,
mu over sensory–motor cortex. However, due to the absence of
scalp and skull, all EEG patterns look more “spiky.” Scalp and
skull filter out in particular high frequencies, including spikes,
whereas slow frequencies pass. Thus, only a small fraction of
intracranial spikes are visible in scalp EEG. If an area of less
than 6 cm 2 is active, no recognizable spike activity is found on
the scalp. Discharges from areas between 6 and 10 cm 2 resulted
in 10% spikes, and those from areas 10 cm 2 resulted in 90%
of spikes that were retrievable in scalp EEGs (24).
It has been reiterated that IEEG has the advantage of being
devoid of muscle artifacts; however, this might not be entirely
true as indicated by a recent case report (73). In this 17-year-old
boy, muscle artifacts appeared at inferolateral temporal
subdural electrodes, using another intracranial electrode as ref-
erence, while eating or during involuntary facial movement as
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes
part of the seizure semiology. Thus, muscle activity may be
picked up by intracranial recordings, similar to any other inter-
fering electrical activity, albeit its appearance may be different
from muscle artifacts in scalp recordings.
If high frequencies remain in the EEG, everything appears
“epileptogenic,” which occasionally has led to incorrect EEG
interpretation. This may be the reason why most authorities in
the field recommend relying mainly on the ictal data, and less on
the interictal EEG. Consequently, in intracranial recordings
interictal discharges appear more often multifocal and more fre-
quently than in surface recordings. To determine the relevance of
the interictal IEEG spikes in a given location, several approaches
are conceivable. First, the determination of the underlying area
(see above): if the spikes are seen over an area of 10 cm 2 or more
(as estimated by the number of simultaneously active elec-
trodes), they arise most likely from the epileptogenic zone (see
above). Second, a quantitative analysis, regarding the relative fre-
quency of spikes from different regions, may help to determine
their relevance. This could be done by visual analysis or through
automatic or semiautomatic detection algorithms. The latter
were used in a study of IEEG in 32 patients (24 TLE and 8
extratemporal lobe epilepsy), with mostly combined depth and
subdural electrode recordings (74). Spikes with highest fre-
quency, highest amplitude, and shorter duration originated
most often from the ictal onset zone, that is, 2 cm from the site
of seizure origin. Another study in 13 children, including also
patients with cortical dysplasia, came to similar conclusions. In
that study, subdural electrodes were implanted. Electrodes with
the highest spike frequency and amplitude were closest to the
area of seizure onset in most of the patients (75), and 11 out of
13 patients were seizure-free (Engel’s class I (76)). The yield of
careful interictal analysis was studied in patients with exclusively
extratemporal lobe epilepsy, including cortical dysplasia. No
quantification for different locations of interictal spikes was pro-
vided, except mapping of the extension of the interictal epilep-
togenic zone including depth and subdural electrodes. If the
resection volume included the area of dominating interictal
spikes, good surgical outcome was significantly more likely (77).
The definition of IEEG spikes is mostly qualitative and
adopted from scalp EEG. Criteria include a morphologic pat-
tern that has (a) triangular waveform, with or without a follow-
ing slow wave, (b) has duration of 70 msec, and (c) stands out
clearly from the background with its amplitude and steep com-
ponent. Given the difficulties to determine unambiguously the
presence or absence of spikes, spike detection programs for
scalp EEGs were developed and applied also for IEEG (78).
Three of those algorithms were tested for IEEG from seven
patients (six with TLE) and compared to analysis by visual
inspection from two human reviewers. The agreement between
the two EEG experts was relatively small with less than 50% of
all spikes identified by both reviewers, a number that compares
similarly to scalp EEG. Compared to the human reviewers, the
automatic algorithms were concordant in only 24% to 32%.
However, despite the low number of agreement between the
human experts and computer, this affected rather the number
of detected spikes but not the location: both reviewers and
automatic systems identified the spikes in the same anatomical
localizations.
685
The intracranial interictal EEG correlates of dysplastic tissue
have been investigated, since dysplasia and/or its extent are
often invisible in the MRI. Rhythmic spike-wave complexes in a
continuous or quasi-continuous fashion are most likely to be
associated with cortical dysplasia (79,80). In fact, in patients
with dysplasia, the analysis of “interictal” anomalies seemed to
be more promising than that of ictal intracranial recordings
(81). Other studies did not find any difference in interictal spik-
ing pattern between dysplastic and nondysplastic cortex (82).
Continuous discharge pattern are also found with severely gli-
otic tissue (83), but the difference between dysplasia and severe
gliosis should be obvious already from the patient’s MRI.
In conclusion, disregard for interictal discharges in the EEG
and concentration only on ictal recordings is an obsolete con-
cept if only “true” interictal spikes are considered. Careful
analysis of the localization and frequency of interictal spikes in
IEEG can significantly add to the detection of the epileptogenic
zone. Recently, studies were presented that aimed at the brain
“current source” reconstruction from intracranial recordings
(84), similar to algorithms for scalp EEG. Future studies will
determine their feasibility and correctness in a clinical environ-
ment (see also Chapter 57). If sophisticated analysis of interictal
IEEG provides sufficient data on the seizure-onset area and
resection leads to seizure freedom in most cases, recording time
could be significantly shorted and result in lower morbidity and
costs. Electric source localization of interictal scalp EEG corre-
sponds most often to the ictal onset zone, and its resection leads
to seizure freedom in the majority of patients (21,85).
ICTAL INTRACRANIAL EEG PATTERN
AND THEIR RELEVANCE TO OUTCOME
Despite promising results from interictal intracranial studies in
terms of localization of the epileptogenic zone, ictal recording
still remains the chief goal of IEEG. Till date, there is no consen-
sus of how many seizures are required to be confident enough to
recommend surgery. Most centers try to obtain a minimum of
five seizures or more if multifocal origin is suspected. The goal is
to determine a principal seizure-onset focus. However, strict
rules are probably not possible, given that intracranial record-
ings cannot be extended indefinitely due to medical concerns.
As a general rule, the chances of seizure control are markedly
better if the EEG onset precedes the clinical onset. The reasons
are obvious. If the seizure is already ongoing clinically, without
focal EEG changes, there is a strong risk that the electrodes do
not cover the zone of seizure generation, and any conclusion of
the true area of epilepsy onset remains obscure. Furthermore,
the chance of postoperative seizure control is higher if there is
only one type of seizure semiology. This is true for ictal scalp,
and is also true for ictal intracranial workups. However, since
the noninvasive “phase 1” workup usually precedes the invasive
“phase 2” workup, the information on how many different
seizure types are present is usually already provided.
A large body of studies looked into a possible relationship
between intracranial ictal EEG pattern, its location, and surgical
outcome. Although most early studies included patients with TLE
using depth electrode recordings, later observations were extended
to subdural electrode recordings and/or extratemporal epilepsies.
686 Part V ■ Complementary and Special Techniques
Concerning iEEG, the impact of focal versus regional or dif-
fuse onset has also been examined in several studies. A higher
chance of postoperative seizure freedom is associated with onset
in only few electrodes (86–88). Focal onset, defined as ictal activ-
ity in less than four to five electrode contacts, appears to be a
sign of proximity of the recording contacts to the epileptogenic
zone, so its resection is likely to result in seizure control or major
seizure reduction in up to 85% of patients with fast frequency
onset (88,89). This relationship was less evident in another study
of neocortical, mostly extratemporal epilepsy (90), which
reported no difference between focal versus nonfocal onset.
Concerning intracranial ictal EEG pattern, onset rhythms were
classified either in terms of frequency bands (92) or as certain
morphologic patterns (88). Since fast rhythms of cortical origin
are not filtered by skull and scalp, ictal onset, often characterized
by low-amplitude rapid beta or gamma frequencies, are seen eas-
ier but most importantly earlier than on scalp EEG. In fact, fast
EEG onset frequencies (in the beta or gamma range) are the most
frequently reported EEG patterns at onset (91,92), significantly
related to class I and II outcomes (89,93,94). According to Lee et
al., extratemporal onset is faster and found in the gamma range
( 30 Hz), whereas temporal lobe onset was seen in the beta-range
frequency (88). These initial fast EEG patterns are also often asso-
ciated with focal onset, and therefore, evocative of closeness of the
electrodes to the seizure-onset zone. Slower EEG onset frequencies
( 5 Hz), probably indicate that there is already propagation from
a focus elsewhere (88,95). During slow iEEG patterns, the patient
may show loss of responsiveness, automatisms, and other behav-
ioral signs of already ongoing seizure activity (95). Consequently,
better surgical outcome is reported with focal, faster ictal EEG
onset, although other evidence indicated that slow onset is not a
bad sign in all circumstances. If slow-onset ictal IEEG contains
poor prognosis, then it is associated with developmental pathol-
ogy, regional distribution, or extratemporal location (only 1 in 16
become seizure free) (88).
Other ictal IEEG patterns include sinusoidal pattern, preictal
spiking, or high-amplitude beta spiking. Sinusoidal pattern was
only seen with mature pathologic substrates. Its prognosis is
unclear. It was felt to be a sign of propagation (94), although
this was less clear in another report (88).
Next to low-voltage, fast beta activity, periodic spiking before
seizure onset was frequently observed. It represents another ictal
onset pattern and was found to be related to seizures of hip-
pocampal and amygdala onset (92,96–98). Ictal spiking at
seizure onset is correlated to the presence of hippocampal scle-
rosis and/or neuronal decrease in the CA1 region of the hip-
pocampus, that is, which is in itself a favorable sign with respect
to postoperative outcome (98). Consequently, ictal onset spiking
is related to good prognosis in terms of seizure control. Finally,
high-amplitude beta spike activity has also been described (99)
in TLE, but it is rare and its prognosis is still unclear.
Although the intracranial ictal onset pattern was not predic-
tive of the underlying pathology, low-voltage fast activity and
rhythmic alpha–theta spike-wave were more common in devel-
opmental substrates (e.g., dysplasia) (Table 33.3) (88).
Another study subject concerns the variability of ictal onset.
The rational behind it is that variable onset patterns could
Tabl e 3 3 . 3
Intracranial Ictal Onset Pattern
Low-voltage fast activity (LVFA; 13–20 Hz)
• Beta range (13–30 Hz)
• Gamma range ( 30 Hz)
Rhythmic spiking 1–2 Hz
Rhythmic sinusoidal pattern in alpha–theta range (around
5–9 Hz)
Rhythmic spiking in the alpha–theta range
Semirhythmic slow waves 5 Hz
High-amplitude beta spike activity
indicate multiple seizure origins. In a study of 18 patients with
TLE, more than one ictal EEG onset pattern was found in most
patients (87). In an earlier study, it has been underlined that the
variability of the onset frequency is a characteristic of TLE and
rarely found in extratemporal cases (92). However, in both
studies, surgical outcome was not reported, that is, it remains
unknown if variability of onset EEG pattern holds prognostic
significance.
Apart from onset IEEG pattern, the further evolution of the
seizure pattern and its prognostic significance has been addressed.
The speed of propagation represents an important factor. Fast
propagation, that is, within one to a few seconds, to remote
regions is related with a less favorable surgical result (90,100). This
appears to be equally true for temporal and extratemporal lobe
epilepsies. Interhemispheric propagation time in patients with
TLE was analyzed, and a delay of only 0.5 to 5 seconds was asso-
ciated with a poor prognosis (101). It is assumed that rapid or
even almost simultaneous EEG onset at two distant sites such as
both temporal lobes points to a pacemaker elsewhere with rapid
implication to the two sites. Alternatively, rapid propagation could
suggest diffuse damage of remote areas, each with significant
epileptogenic properties of its own.
Propagation may occur through different EEG patterns. Two
propagation patterns at distant sites were identified by Schiller
et al. in temporal and extratemporal lobe epilepsies (94): first,
rhythmic theta–delta activity correlating with the activity at the
seizure-onset site, and second, continuation of the seizure pat-
tern of the seizure-onset site. No significant correlation
between onset and propagation EEG pattern and outcome was
found. The visual analysis of iEEG during propagation is prob-
ably less relevant in the clinical context, and may even differ
within a patient from one seizure to the other. It remains to be
shown if coherence-based measurements are more informative,
that is, if they help to identify nonfocal epilepsies (102).
Finally, the end of the intracranial ictal EEG pattern (offset)
was studied in order to determine if offset outside the generat-
ing structures is related to worse outcome. Theoretically, it
could indicate the presence of epileptogenic tissue in remote
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes
sites, ready to pick up seizure activity from elsewhere. However,
there was no prognostic difference between patients with an
ipsilateral and contralateral offset (103), but this is challenged
by reports that find that patients with contralateral offset have
poorer surgical outcome (104,105). However, as for the onset,
the offset pattern can also vary within one patient, and more
studies are probably needed to determine the prognostic rele-
vance of the offset pattern.
Examples of ictal EEGs, exemplifying the different EEG
onset and offset patterns described above, are found at the end
of the main text.
Although there have been many studies on the value of ictal
IEEG, many questions remain on the prognostic significance of
a given EEG pattern. It appears that several variables influence
the creation of an ictal IEEG pattern, and these variables
include the affected lobe, underlying histopathology, the type of
electrode and sites of contacts, age of patient, and speed of
medication withdrawal. So many variables exist that most stud-
ies are underpowered. Multicenter studies making use of mod-
ern 3D localization of all contacts could help to better identify
the significance of ictal IEEG pattern.
HIGH-FREQUENCY OSCILLATIONS
In recent years, the research on the markers of epileptogenic tis-
sue in the interictal EEG have been restarted, given that the
new-generation recording systems with higher EEG sampling
rates can capture very fast rhythms (30 to 500 Hz). Probably the
first study on that subject was published by R. Fisher and his
colleagues (106) who found increased high-frequency activity,
40 Hz, in the seizure-onset zone. Many more studies followed,
supporting the notion that high-frequency oscillations (HFOs)
are more frequently colocalized with the seizure-onset zone,
albeit the exact shape and frequency band of what is truly
evocative of the epileptic focus still needs to be defined (107).
They are believed to reflect inhibitory field potentials, which
facilitate information transfer by synchronizing neuronal activ-
ity over long distances. These recordings require a high sam-
pling rate, that is, 2000 to 5000 Hz and a 500 to 1000 Hz
687
low-pass filter, instead of the usual 200 to 250 Hz sampling rate
and 70 to 100 Hz low-pass filter used for the routine iEEG.
HFOs are recorded with the commercially available subdural or
depth electrodes or with microwires that are inserted together
with macroelectrodes (Figs. 33.5 and 33.6).
HFOs are short bursts of high-frequency pattern that
become visible if the EEG is sampled with a wide bandwidth of
up to 5000 Hz. Bradin and coworkers differentiated ripples (60
to 160 Hz) and fast ripples (FRs) (250 to 500 Hz) (108), a
nomenclature that has been taken up by similar studies, albeit
the definition of lower and upper limits of the ripple frequen-
cies is slightly different from study to study. Both ripple types
increase during slow-wave sleep (109,110), which may explain
the propensity of seizures to occur during periods of decreased
vigilance or sleep. However, there is evidence that HFO of ripple
frequency is also found in the nonepileptogenic temporal lobe,
whereas FRs are equally predominant in wakefulness and sleep,
that is, they do not appear to be a physiologic event (see review
of subject in Chapter 37).
Across several studies and groups, it was claimed that FRs are
more often pinpointing to the ictal epileptogenic zone than
HFO of lower frequency (111,112,115); however, not all studies
came to the same conclusion (113). They may occur alone or
superposed on the spike (107). Overall, FRs appear to be a more
localized event compared to HFOs of lower ripple frequency
(112–114). They may also precede the seizure onset for up to
several seconds to 20 minutes (109,115), but are not present in
all patients (115). If HFOs indeed indicate the zone of seizure
onset, longer EEG recordings requiring the capture of several
seizures and confirming the presence of a unifocal ictal onset
zone would become obsolete. However, the issue is far from
resolved. Recordings from both macro- and microelectrodes
showed that the morphology of HFOs of both types, FRs and
ripples, also occur together, that is, the burst starts with ripples
and evolves to FRs, or inverse (112). FR HFOs were often
recorded from a single microwire (108,112), indicating a very
local event from a neuronal assembly. How and when such a
local and short discharge (20 to 50 msec) evolves to clinical
seizures of several minutes remains to be elucidated.
Figure 33.5 Microelectrode arrays.
Left: So-called Utah microarray
made of multiple silicon spikes of
1.0 or 1.5 mm length. A pressure
applicator is used to insert the sili-
con array into the cortex across the
pia. Right: Microelectrodes as a
bundle at the end of a depth elec-
trode. (From Turner DA, Patil PG,
Nicolelis MAL. Conceptua l a nd
Technical Approaches to Human
Neura l En semble Recordings.
Methods for Neura l Ensemble
Recordings, Frontiers in Neuro-
science. CRC Press; 2008.)
688 Part V ■ Complementary and Special Techniques
Research on human HFOs is relatively recent. Most of the
studies are done in mesial temporal lobe structures, given that
TLE patients are more frequent in the surgical patient popula-
tion and/or more frequently investigated with intracranial and
microwire electrodes together. The existence of ripples or FRs
in extratemporal lobe epilepsy is less well known up till now,
but preliminary data suggest that they are also present in
extratemporal areas (113). Thus, the most important question
to be resolved concerns which type of HFO is truly pathologic
and potentially serves as biomarker (107). It is of interest to
note that recent studies found paroxysmal fast activity even in
scalp recordings, mainly in pediatric patients in the beta and
gamma frequency bands (15 to 70 Hz) (116,117).
SPECIAL CONSIDERATIONS
IN PEDIATRIC PATIENTS
Invasive procedures are rare in infants younger than 3 years, but
if they are indicated, they should be done at specialized epilepsy
surgery centers. Close collaboration with the (neuro) pediatri-
cians is mandatory if the center is not already located in a pedi-
atric department. In contrast to previous attitudes, children
with epilepsy should be investigated as soon as pharmacoresis-
tance is documented. To wait until the epilepsy “grows” out is
no more appropriate unless the underlying syndrome is well
determined and surgical treatment is clearly not indicated, for
example, Dravet syndrome. Thorough workup helps to
determine the underlying syndrome and potential usefulness of
surgical therapy. Not all children with early-onset difficult
epilepsy are surgical candidates, but if they are, any year of per-
sistent seizures has deleterious effects on their development. If
the indication for IEEG is given, the same implantation strate-
gies, electrode types, and most importantly, the clinical reason-
ing regarding the goals apply as for adults.
Pediatric patients suffer more frequently from extratemporal
lobe epilepsy. Consequently, they are considered as candidate
Figure 33.6 High-frequency oscil-
lations recorded from macro- and
microelectrodes from the left mesial
temporal cortex, harboring also the
seizure-onset zone. Note the coinci-
dence of the spike in the macroelec-
trodes, and ripples recorded in
the microelectrodes. (From Worell
et al., Brain, 2008.)
for intracranial monitoring if the exact focus localization is not
yet evident from noninvasive monitoring. Like in adults, addi-
tional electrocorticography may be necessary to identify elo-
quent cortex.
In older children, the corticography procedure is similar to
that in adults (118). Trains of 50 Hz, alternating polarity
square-wave pulses of fixed pulse duration (0.3 msec) are most
often used, starting with low currents, for example, 0.5 or 1 mA.
The stimulation intensity is gradually augmented in 1-mA steps
until a sensation or movement is elicited or the maximal value
has been reached (up to a maximum of 10 to 15 mA, without
any perceptible sensation or movement). However, in very
young children, this approach is often ineffective, since motor
responses are virtually unobtainable under 1 year of age, and
with difficulties until 4 to 5 years of age (119). The Miami
group described a different stimulation procedure in children
younger than 5 years, which results more often in motor move-
ments: they propose dual increment of pulse duration and
stimulus intensity instead of increment of only stimulus inten-
sity (120). The minimal current intensity (in milliamperes)
required to elicit a response at very long pulse duration (in mil-
liseconds) is called rheobase. The double of rheobase is called
chronaxie, and corresponds to the pulse duration required to
elicit a response twice as intense as the rheobase. In nonmyeli-
nated fibers, the chronaxie is longer than that of myelinated
fibers. In very young children, myelination is still incomplete,
gradually increases with age and is also site dependent; that is,
myelination starts in the posterior cortices. Jayakar et al. inves-
tigated the dual-increment approach in six patients between 1
and 10 years using pulse duration between 0.3 and 1 msec and
increasing intensity starting at 1 mA up to 15 mA. With dual
increment, that is, alternating 1 to 2 mA and 0.1 to 0.2 msec
increments, motor responses were obtained in three patients of
1, 3, and 4.5 years, respectively, and afterdischarges in 12/14
electrodes, but no response with the adult paradigm. In three
older children, aged 5.5, 7, and 10 years, the adult paradigm,
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 689

that is, with fixed pulse duration of 0.3 msec, as well as the dual-
increment paradigm, both elicited responses and afterdis-
charges in 16/16 contacts.
An alternative approach in children (or adults) who are not
able to describe their feelings or collaborate with the examiner
is evoked potentials. In pre- or peroperative monitoring,
somatosensory or motor evoked potentials identified the local-
ization of the central sulcus (121,122).
OUTLOOK
Evaluation of patients with intracranial electrodes has under-
gone changes in terms of defining the indication for their use.
Most patients with mesial temporal epilepsy and hippocampal
sclerosis do not need IEEG monitoring if they have access to
modern imaging techniques. Even patients with nonlesional
epilepsy may not always need intracranial evaluation (123).
Despite advances in brain imaging, the rate of seizure-free
patients is still lower in patients with extratemporal compared to
TLE. The precision and yield of noninvasive imaging for patients
with extratemporal lobe epilepsy is still under evaluation and
will further improve. However, in the meantime patients with
conflicting results from the different exams in the noninvasive
phase (phase 1) are still candidates for intracranial monitoring.
As stated in the beginning of this chapter, each intracranial eval-
uation needs a solid hypothesis as to where the seizures might
have their origin. IEEG should not be a “fishing expedition,”
because even with EEG systems of 200 or more channels, not all
possible brain regions can be covered, and more importantly, the
complication risk increases and may outweigh the benefit of the
IEEG procedure. To obtain a valid hypothesis, coregistration and
multimodal imaging may help to increase the yield of the inva-
sive phase (124). Whole-head scalp EEG (and MEG) arrays of
200 or more electrodes/sensory, together with EEG source local-
ization algorithms, were successfully employed in pediatric and
adult epileptic patients with temporal and extratemporal lobe
epilepsy for focus localization (125,126).
IEEG may be used for other purposes in addition to determin-
ing the seizure focus. There are a number of studies, albeit limited,
that try to make use of IEEG recordings to determine neuropsy-
chological functionality of distinct brain structures. For example,
ongoing seizure activity was used as “electrical Wada test,” (127)
presence or absence of N400 components, as well as HFOs as indi-
cator of (healthy) mnestic processes (128–131) or event-related
gamma activity during naming to localize language cortex (132).
Intracranial monitoring imposes a certain burden on the
patient and his/her family, both on the psychological and phys-
ical levels, and the indication must be carefully weighted against
the possible benefit. Across most centers and studies, the num-
ber of patients who cannot be offered surgery, despite intracra-
nial monitoring, is around 25% to 30%. Future studies will
show if with the addition of sophisticated noninvasive brain
imaging this percentage can be lowered, either because non-
surgical candidates can be more easily identified or because
the electrodes are placed closer to the epileptogenic zone. See
Figures 33.7 to 33.21.

Figure 33.7 A: Dissociation between seizure-onset zone and symptomatic zone. Patient CA, 22-year-old male with left hip-
pocampal sclerosis but nonlocalizing ictal scalp pattern. Continuous discharges in the left hippocampus (depth electrodes;
HAG1, HPG1, HPG2) as interictal EEG. AD, right amygdala/ anterior temporal lobe; HD, right hippocampus/ midtemporal
lobe; G-contacts, subdural grid electrodes overlying the lateral temporoparietal cortex.
(Continued on next page)
Figure 33.7 (Continued) B: Position of depth elec-
trodes in the left and right temporal and frontal cortex,
and subdural grids over temporoparietal cortex.
Coregistration of high-resolution CT (with the elec-
trodes) and the patient’s MRI. C: The interictal rhythm
changes to rhythmic appearance of bursts of rapid 30 to
40 Hz pattern, appearing every 6 seconds. D: The hip-
pocampal bursts stop and are replaced by widespread 10
Hz rhythms in the lateral fronto-orbital cortex (FOG7)
and lateral temporal subdural contacts (contacts GA,
GB, GC, GD), coinciding with the onset of clinical symp-
toms (*), that is, lip smacking, manual automatisms,
loss of contact.
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 691

Figure 33.7 (Continued) E: The extrahippocampal discharges persist while the left hippocampus remains silent. This case
represents an example of a dissociation between areas of onset and the symptomatic propagation, which implicates rather
lateral neocortical contacts.

Figure 33.8 A: Slow ictal onset EEG pattern ( 5 Hz). Patient VA, 12-year-old female with nonlesional left hemispheric
epilepsy with dysesthesia in the right foot. Slow onset (2 Hz waves) in the left superior parietal region (→), corresponding
to the dark gray circled contacts. Five seconds later: rhythmic spikes in the adjacent region (**). IA and IP, anterior and
posterior temporal subdural strips (not shown).

(Continued on next page)

692 Part V ■ Complementary and Special Techniques

Figure 33.8 (Continued) B: Continuous rhythmic 5-Hz spiking in the contacts close to the sensory foot area, related to
the patient’s typical aura. However, persistent seizures postoperatively after a limited superior parietal resection suggest
that the site of ictal onset does not coincide with the epileptogenic zone. Slow-onset rhythms ( 5 Hz) are considered to
be less likely associated to the true epileptogenic zone. This is true even when the slow onset pattern evolves to rhythmic
discharges. Seizures continued after limited resection in the postcentral region.

Figure 33.9 A: Focal onset spiking. Patient AF, 18-year-old with nonlesional left frontal epilepsy. Localized onset as rhyth-
mic spiking (→) in four selective subdural contacts (C3, C4, D3, D4) over the lateral anterior left frontal cortex. Within
5 seconds spreads to other adjacent contacts.
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 693

Figure 33.9 (Continued) B: (Right) Localization of subdural electrodes coregistered with the patient MRI. Contacts of
seizure onset (see Figure 33.8) are shown as big, dark gray circles. The IEEG localization of the ictal focus was concordant
with the localization obtained by noninvasive electric source localization (left; see also Chapter 55). C: Persistent rhythmic
spike-wave pattern continued. No clinical symptoms were seen or noted by the patient. A topectomy of the cortex under-
lying the four contacts resulted in marked improvement of the epilepsy disorder (i.e., one to two GTCS per year if AEDs
were forgotten). Focal onset, that is, onset in less than five contacts is more likely related to good outcome, as it is the case
here. Independently, ictal onset spiking has been described as another favorable sign in hippocampal seizures, it may also
be true for extratemporal lobe epilepsy.
694 Part V ■ Complementary and Special Techniques

Figure 33.10 A: Low-voltage beta activity at ictal onset. Patient CR, 20-year-old, right-handed, with transmantle dysplasia
in the left posterior temporal cortex. Onset was characterized by rapid rhythms ( 25 Hz) in two distinct contacts (D4 C4).
Left: Schema of electrodes coregistered to the patient’s brain. Contacts corresponding to early ictal activity in black. B:
Propagation to other contact over other temporal lateral cortex is noted. Discharges remain relatively well localized, but
become associated with clinical symptoms (*).
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 695

Figure 33.10 (Continued) C: Discharges continue, implicating now the whole temporal lobe, and to a lesser degree also
frontal regions. The onset contacts remain the most active—low-amplitude beta activity onset is considered a favorable
sign in terms of seizure control. Indeed, resection of this area resulted in persistent seizure freedom ( 5 years).
Comprehension was found to be localized in the contralateral hemisphere.

Figure 33.11 A: Diffuse slow ictal onset. Patient FS, 42-year-old male, with right hippocampal sclerosis and multiple right
hemispheric heterotopia. Poorly defined ictal onset: diffuse onset with 1 to 2 Hz sharp slow waves in right frontal (CD7-8),
right lateral temporal (AD6-8), right parietal periventricular heterotopia (HPPD6-8). Left: Depth electrode position coregis-
tered with the patient’s brain.

(Continued on next page)

696 Part V ■ Complementary and Special Techniques

Figure 33.11 (Continued) B: No further localization, but synchronous 1 to 2 Hz discharges of a network comprising
fronto-orbital (FOD), right frontal lateral (CD), lateral temporal (AD), both right periventricular temporal (HPPTD), and
parietal (PAD) heterotopia. C: Recording of another seizure, with simultaneous onset of the temporal lobe, heterotopias,
and frontal structures, preceded by a brief hippocampal burst (HPD) of 6 Hz sharp waves. Resection of the hippocampus
and partially the anterior heterotopia brought only partial seizure relief (decrease of 50% ). Diffuse onset, even in the pre -
sence of structural lesions, is less likely related to good postoperative seizure control. Moreover, variable seizure-onset pat-
terns were noted, another unfavorable prognostic factor.
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 697

Figure 33.12 Gamma activity at onset. Patient MS, 15-year-old female with tuberous sclerosis. Largest tuber in the left
posterior temporal lobe. Seizure onset presents as rapid gamma activity ( 100 Hz; GG5-8; GH1-2). Incterictally (before
seizure onset) more widespread discharges over the left posterior cortex.

Figure 33.13 A: Patient MS: 80 to 100 Hz ripples at the site of seizure onset (GG3). Left: Coregistration of the patient’s
subdural electrodes position with her own brain. Site of ripple is indicated by dark gray circle.

(Continued on next page)

698 Part V ■ Complementary and Special Techniques

Figure 33.13 (Continued) B: Patient MS: Diffuse semirhythmic sharp waves. Close analysis reveals the presence of rapid
ripples riding on the discharges (max GH2). Resection of the tissues underlying these ripples resulted in seizure freedom
(follow-up 2 years), despite more diffuse spike and sharp wave discharges. Not only focal beta but also gamma activity at
seizure onset is a good prognostic factor, despite the presence of multifocal lesions.

Figure 33.14 A: Preictal spiking. Patient VT, 28-year-old, with left hippocampal sclerosis. Monitoring was advised due
to nonlateralizing seizure onset in ictal scalp EEGs. Depth electrodes were inserted orthogonally in both hemispheres.
(Right: FOD, fronto-orbital; CD, frontal lateral cortex/ anterior cingulate gyrus; AD, anterior temporal lobe/ amygdala;
HAD/ HPD, anterior/ posterior hippocampus. Left: FOG, CG, AG, HAG, HPG). Before the seizure, preictal spiking is noted
in the sclerotic left hippocampus (HAG2-3, HPG2-3).
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 699

Figure 33.14 (Continued) B: Inhabitual propagation pattern. Patient VT: Preictal spiking is replaced by rhythmic spiking
in the left hippocampus (HAG2, HPG2, →). Recruitment occurs first in the lateral frontal cortex (FOG7-8, *). Small ampli-
tude discharges in the left amydala become more rhythmic (AG1-2, 2-3, **). The smaller amplitude is explained by the closed
electrical field of the amygdala as compared to the hippocampus. C: Recruitment in the lateral frontal cortex (FOG7-8)
continues. Close analysis of the temporal pattern reveals that the discharges in the fronto-orbital cortex are 50 to 100 msec
later than the large hippocampal spikes, indicating that the hippocampus drives the fronto-orbital discharges. No propaga-
tion to lateral temporal contacts (HAG7,8; HPG7,8), explaining retrospectively why the scalp EEG was unrevealing.
Propagation pattern can show significant variability between patients and may involve even structures outside the lobe of
onset, including the contralateral hemisphere.
700 Part V ■ Complementary and Special Techniques

Figure 33.15 Low-amplitude gamma activity in extratemporal lobe epilepsy. Patient SC, 17-year-old, with nonlesional
right frontal epilepsy with daily nocturnal generalized tonic-clonic seizures (GTCS). Seizure onset is characterized by a very
focal beta activity (GA7), which diffuses to other subdural contacts. Resection of the epileptogenic zone including the cor-
tex underlying GA7 resulted in disappearance of GTCS (except once after sleep withdrawal), but persistence of short simple
partial seizures (sensory–motor symptoms of the tongue). Focal low-amplitude rapid ictal onset activity is favorable sign
also in extratemporal lobe epilepsy.

Figure 33.16 A: Significance of different onset patterns. Patient BE, 43-year-old, with bitemporal lobe epilepsy with left-
sided predominance, began after vaccination at the age of 7 years, MRI: bilateral atrophy. Onset is characterized by beta
activity in the left parahippocampal gyrus (LH4-5 and adjacent contacts, →). RF/ LF, right/ left fronto-orbital cortex; RA,
toward right amygdala (LA, left amygdala); RH/ LH, toward right/ left hippocampus.
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 701

Figure 33.16 (Continued) B: Offset pattern of this seizure in the same contacts as the seizure onset (parahippocampal
gyrus, LH3-4, LH4-5). C: Another seizure that also started with focal beta activity but is now in the left amygdala and hip-
pocampus (LH1-2, LA2-3, LA1-2, →), leading to recruitment of right mesial temporal structures (RA1-2, RA2-3, RH-1-2, *);
variable onset pattern are rather less favorable signs. Indeed, selective resection of left mesial temporal structures did not
lead to persistent seizure freedom, albeit reduction of approximately 70% .
702 Part V ■ Complementary and Special Techniques

Figure 33.17 Diffuse EEG onset. Patient SG, 20-year-old male, with epilepsy onset at the age of 3. The scalp EEG showed
almost continuous discharges over the left frontal lateral cortex. 8 8 subdural grid was implanted over the left frontal
and temporal cortex. Seizure onset (→) was noted simultaneously over frontal and temporal regions (B5–6, 6–7, 7–8;
H5–6, 6–7, 7–8). Limited resection of left frontal cortex due to the presence of language cortex (histopathology: severe cor-
tical dysplasia) did not control seizures persistently.

Figure 33.18 A: Ictal recording with foramen ovale electrodes (FOE). Patient ES: 21-year-old, left-handed, known for
seizures since the age of 9. Left hippocampal sclerosis. Scalp ictal EEG showed left, right, or bitemporal onset. Interictal
spikes were found over left and right temporal lobe with variable predominance. FOE were inserted bilaterally (left: L1-6;
right: R1-6), and recorded together with scalp electrodes. Ictal recording with first changes in the left FOE.
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 703

Figure 33.18 (Continued) B: The seizure evolves, with few changes in the corresponding scalp electrodes. C: A low-
amplitude beta activity starts (*).

(Continued on next page)

704 Part V ■ Complementary and Special Techniques

Figure 33.18 (Continued) D: Beta activity persists, with still few changes in the scalp EEG and no clinical symptoms.
Thus, FOE is capable of recording focal hippocampal ictal activity, with higher sensitivity as scalp EEG.

Figure 33.19 A: Subclinical seizures. Patient VT, 28-year-old, with left hippocampal sclerosis. Subclinical seizure in the
left sclerotic hippocampus ( 30 Hz beta activity) with very focal onset (in one contact, HPG2). No further recruitment is
observed. Clinical symptoms are limited to deficient verbal memory.
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 705

Figure 33.19 (Continued) B: Subclinical seizure pattern gradually stops and discharges are of lower frequency.
Multichannel IEEG needs to be carefully scrutinized, because they may involve only one channel. If frequent, they usually
indicate a high epileptogenicity of the recorded structure.

Figure 33.20 A: Ictal beta activity, less focal. Patient RP, 24-year-old, diagnosed with nonlesional right temporal epilepsy.
Noninvasive workup found right anterior and posterior temporal discharges. After ESI localizing the focus rather to poste-
rior temporal regions, careful review of the MRI showed focal polymicrogyria in the temporo-occipital junction. A subdural
grid was placed there, as well as depth electrodes in the left and right temporal (right: AD, HAD, HPD; left: AG, HMG)
and frontal regions (CD, FOD).

(Continued on next page)

706 Part V ■ Complementary and Special Techniques

Figure 33.20 (Continued) B: Beta activity persists, becoming more rhythmic and organized. Around 10 seconds after EEG
onset in the temporo-occipital cortex, recruitment of the right hippocampus (→). C: Persistence of right-sided seizure activ-
ity, with left hippocampal recruitment (→).
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 707

Figure 33.20 (Continued) D: Posterior and bihippocampal discharges persist. This recruitment pattern may explain why
the patient is amnesic of the entire event. Seizures came back 2 years after resection of the polymicropgyric cortex, but
subsided again after re-adaptation of drug treatment. Less focal ( 5 Hz) EEG onset, together with a less organized rhythm
at the beginning, may reflect more widespread epileptogenic zone.

Figure 33.21 A: Focal onset with late recruitment and contralateral offset. Patient SS, 37-year-old, right-handed, with non-
lesional right temporal lobe epilepsy. Depth electrodes were implanted in both temporal lobes and right frontal lobe. Onset
is characterized by high-amplitude spiking, mixed with beta activity. Onset is restricted to four contacts in the right mesial
temporal lobe (HAD1,2, HPD1,2; AG and HAG, left anterior and midtemporal lobe, respectively; FOD, fronto-orbital; CD,
right frontal lateral).

(Continued on next page)

708 Part V ■ Complementary and Special Techniques

Figure 33.21 (Continued) B: Focal discharges persist, somewhat more recruitment of the right amygdala (AD1-3). Please
note that each page contains 20 seconds of EEG. C: After 29 seconds recruitment of lateral temporal and fronto-orbital
contacts.
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes 709

Figure 33.21 (Continued) D: Only very late ( 40 seconds) recruitment of the contralateral hippocampus (*). E: While
the right-sided discharges subside, ictal spiking in the left hippocampus persists until the end of the seizure.

(Continued on next page)

710 Part V ■ Complementary and Special Techniques
Figure 33.21 (Continued) F: The patient received a selective right antero-mesial temporal resection and is seizure-free
since (histopathology: cortical dysplasia). Focal onset ( 5 contacts) and persistence focality is an advantageous finding.
However, offset in the contralateral hemisphere is obviously less predictive of postoperative outcome.
ACKNOWLEDGMENTS
Supported by the Swiss National Science Foundation (Grant
Nos. 320030-113766 and 33CM30-124089). MS is grateful to
Carlo Schaller for helpful comments on the neurosurgical
aspects of intracranial EEG. Dr Vitalo Chiosa and Dr Laurent
Spinelli provided invaluable help in the preparation of figures.
REFERENCES
1. Tonini C, Beghi E, Berg AT, et al. Predictors of epilepsy surgery
outcome: a meta-analysis. Epilepsy Res. 2004;62:75–87.
2. Cendes F, Li LM, Watson C, et al. Is ictal recording mandatory in
temporal lobe epilepsy? Arch Neurol. 2000;57:497–500.
3. Diehl B, Lüders HO. Temporal lobe epilepsy: when are invasive
recordings are needed? Epilepsia. 2000;41(suppl 3):S61–S74.
4. Reuber M, Kral T, Kurthen M, et al. New-onset psychogenic seizures
after intracranial neurosurgery. Acta Neurochir (Wien). 2002;144:
901–907.
5. Fourth level-epilepsy surgical center for epilepsy. Epilepsia. 1990;
31(suppl 1):S7–S8.
6. Harvey S, Cross JH, Shinnar S, et al. and ILAE Pediatric Epilepsy
Surgery Survey Taskforce. Defining the spectrum of internation
practice in pediatric epilepsy surgery patients. Epilepsia. 2008;49:
146–155.
7. Srikijvilaikul T, Locharernkul C, Deesudchit T, et al. The first inva-
sive EEG monitoring for surgical treatment of epilepsy in Thailand.
J Med Assoc Thai. 2006;89:527–532.
8. Seeck M, Lazeyras F, Michel CM, et al. Non-invasive epileptic focus
localization using EEG-triggered functional MRI and electromag-
netic tomography. Electroencephalogr Clin Neurophysiol. 1998;106:
508–512.
9. O’Brien TJ, Miles K, Ware R, et al. The cost-effective use of 18F-
FDG PET in the presurgical evaluation of medically refractory
focal epilepsy. J Nucl Med. 2008;49:931–937.
10. O’Brien TJ, So EL, Mullan BP, et al. Subtraction ictal SPECT co-
registered to MRI improves clinical usefulness of SPECT in local-
izing the surgical seizure focus. Neurology. 1998;50:445–454.
11. Laxer KD. Clinical applications of magnetic resonance spec-
troscopy. Epilepsia. 1997;38(suppl 4):S13–S17.
12. Knowlton RC, Elgavish RA, Limdi N, et al. Functional imaging: I.
Relative predictive value of intracranial electroencephalography.
Ann Neurol. 2008;64(1):25–34.
13. Duncan JS. The current status of neuroimaging for epilepsy. Curr
Opin Neurol. 2009;22:179–184.
14. Briellmann RS, Pell GS, Wellard RM, et al. MR imaging of epilepsy:
state of the art at 1.5 T and potential of 3 T. Epileptic Disord. 2003;
5:3–20.
15. Sirven JI, Malamut BL, Liporace JD, et al. Outcome after temporal
lobectomy in bilateral temporal lobe epilepsy. Ann Neurol. 1997;
42:873–878.
16. Holmes MD, Dodrill CB, Ojemann GA, et al. Outcome following
surgery in patients with bitemporal interictal epileptiform pat-
terns. Neurology. 1997;48:1037–1040.
17. Hirsch LJ, Spencer SS, Spencer DD, et al. Temporal lobectomy in
patients with bitemporal epilepsy defined by depth electroen-
cephalography. Ann Neurol. 1991;30:347–356.
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes
18. Boling W, Aghakhani Y, Andermann F, et al. Surgical treatment of
independent bitemporal lobe epilepsy defined by invasive record-
ings. J Neurol Neurosurg Psychiatr. 2009;80:533–538.
19. Mintzer S, Cendes F, Soss J, et al. Unilateral hippocampal sclerosis
with contralateral temporal scalp ictal onset. Epilepsia. 2004;45:
792–802.
20. Carne RP, O’Brien TJ, Kilpatrick CJ, et al. MRI-negative PET-
positive temporal lobe epilepsy: a distinct surgically remediable
syndrome. Brain. 2004;127:2276–2285.
21. Michel CM, Lantz G, Spinelli L, et al. 128-channel EEG source
imaging in epilepsy: clinical yield and localization precision. J Clin
Neurophysiol. 2004;21:71–83.
22. Brodbeck V, Lascano A, Spinelli L, et al. Accuracy of electrical source
imaging of epileptic spikes in patients with large brain lesions. J Clin
Neurophysiol. 2009;120:679–685.
23. Gavaret M, Trébuchon A, Bartolomei F, et al. Source localization of
scalp-EEG interictal spikes in posterior cortex epilepsies investi-
gated by HR-EEG and SEEG. Epilepsia. 2009;50:276–289.
24. Tao JX, Ray A, Hawes-Ebersole S, et al. Intracranial EEG substrates
of scalp EEG interictal spikes. Epilepsia. 2005;46:669–676.
25. Ray A, Tao JX, Hawes-Ebersole SM, et al. Localizing value of scalp
EEG spikes: a simultaneous scalp and intracranial study. Clin
Neurophysiol. 2007;118:69–79.
26. Ebersole JS, Hawes-Ebersole S. Clinical application of dipole mod-
els in the localization of epileptiform activity. J Clin Neurophysiol.
2007;24:120–129.
27. Rosenow F, Lüders H. Presurgical evaluation of epilepsy. Brain.
2001;124:11683–1700.
28. Wong CH, Birkett J, Byth K, et al. Risk factors for complications
during intracranial electrode recording in presurgical evaluation
of drug resistant partial epilepsy. Acta Neurochir. 2009;151:37–50.
29. Burneo JG, Steven DA, McLachlan RS, et al. Morbidity associated
with the use of intracranial electrodes for epilepsy surgery. Can J
Neurol Sci. 2006;33:223–227.
30. Berger H. Über das Elektrenkephalogramm des Menschen. Arch
Psychiatr. 1929;87:527–570.
31. Penfield W, Jasper H. Epilepsy and the Functional Anatomy of the
Human Brain. Boston: Little, Brown; 1954.
32. Talairach J, Bancaud J. Stereotaxic exploration and therapy in
epilepsy. In: Vinken PJ, Bruyn GW, eds. The Epilepsies. Handbook
of Clinical Neurology. Vol 15. Amsterdam: North Holland; 1974:
758–782.
33. Talairach J, Bancaud J. Stereotactic approach to epilepsy. Prog
Neurol Surg. 1973;5:297–354.
34. Wyler AR, Ojemann GA, Lettich E, et al. Subdural strip electrodes
for localizing epileptogenic foci. J Neurosurg. 1984;60:1195–1200.
35. Ives JR, Mainwaring NR, Gruber LJ, et al. 128-channel cable-
telemetry EEG recording system for long-term invasive monitor-
ing. Electroencephalogr Clin Neurophysiol. 1991;79:69–72.
36. Siegfried J, Wieser HG, Stodieck SR. Foramen ovale electrodes: a new
technique enabling presurgical evaluation of patients with mesiobasal
temporal lobe seizures. Appl Neurophysiol. 1985;48(1–6):408–417.
37. Barnett GH, Burgess RC, Awad IA, et al. Epidural peg electrodes
for the presurgical evaluation of intractable epilepsy. Neurosurgery.
1990;27:113–115.
38. Cordova JE, Rowe RE, Furman MD, et al. A method for imaging of
intracranial EEG electrodes using magnetic resonance imaging.
Comput Biomed Res. 1994;27:337–341.
39. Winkler PA, Vollmar C, Krishnan KG, et al. Usefulness of 3-D recon-
structed images of the human cerebral cortex for localization of sub-
dural electrodes in epilepsy surgery. Epilepsy Res. 2000;41:169–178.
711
40. Silberbusch MA, Rothman MI, Bergey GK, et al. Subdural grid
implantation for intracranial EEG recording: CT and MR appear-
ance. Am J Neuroradiol. 1998;19:1089–1093.
41. Kovalev D, Spreer J, Honegger J, et al. Rapid and fully automated
visualization of subdural electrodes in the presurgical evaluation
of epilepsy patients. Am J Neuroradiol. 2005;26:1078–1083.
42. Sebastiano F, Di Gennaro G, Esposito V, et al. A rapid and reliable
procedure to localize subdural electrodes in presurgical evaluation
of patients with drug-resistant focal epilepsy. Clin Neurophysiol.
2006;117:341–347.
43. Mahvash M, König R, Wellmer J, et al. Coregistration of digital
photography of the human cortex and cranial magnetic resonance
imaging for visualization of subdural electrodes in epilepsy sur-
gery. Neurosurgery. 2007;61:340–344.
44. Tao JX, Hawes-Ebersole S, Baldwin M, et al. The accuracy and reli-
ability of 3D CT/MRI co-registration in planning epilepsy surgery.
Clin Neurophysiol. 2009;120:748–753.
45. Seeck M, Spinelli L. Intracranial monitoring [Review]. Suppl Clin
Neurophysiol. 2004;57:485–493.
46. Mahvash M, König R, Urbach H, et al. FLAIR-/T1-/T2-co-
registration for image-guided diagnostic and resective epilepsy
surgery. Neurosurgery. 2008;62(suppl 2):482–488.
47. Maurer CR Jr, Fitzpatrick JM, Wang MY, et al. Registration of head
volume images using implantable fiducial markers. IEEE Trans
Med Imaging. 1997;16:447–462.
48. Spire WJ, Jobst BC, Thadani VM, et al. Robotic image-guided
depth electrode implantation in the evaluation of medically
intractable epilepsy. Neurosurg Focus. 2008;25(3):E19.
49. Van Buren JM. Complications of surgical procedures in the diag-
nosis and treatment of epilepsy. In: Engel J, ed. Surgical Treatment
of the Epilepsies. New York: Raven Press; 1987:465–475.
50. Ross DA, Brunberg JA, Drury I, et al. Intracerebral depth electrode
monitoring in partial epilepsy: the morbidity and efficacy of place-
ment using magnetic resonance image-guided stereotactic surgery.
Neurosurgery. 1996;39:327–334.
51. Sansur CA, Frysinger RC, Pouratian N, et al. Incidence of sympto-
matic hemorrhage after sterotactic electrode placement. JNeurosurg.
2007;107:998–1003.
52. Tanriverdi T, Ajlan A, Poulin N, et al. Morbidity in epilepsy sur-
gery: an experience based on 2449 epilepsy surgery procedures
from a single institution. J Neurosurg. 2009;110:1111–1123.
53. Almeida AN, Martinez V, Feindel W. The first case of invasive EEG
monitoring for the treatment of epilepsy: historical significance
and context. Epilepsia. 2005;46:1082–1085.
54. Eröss L, Bagó AG, Entz L, et al. Neuronavigation and fluoroscopy-
assisted subdural strip electrode positioning: a simple method to
increase intraoperative accuracy of strip localization in epilepsy
surgery. J Neurosurg. 2008. DOI: 10.3171/2008.6. JNS17611 [Epub
ahead of print].
55. Chamoun RB, Nayar VV, Yoshor D. Neuronavigation applied to
epilepsy monitoring with subdural electrodes. Neurosurg Focus.
2008;25:E21.
56. Seeck M, Pegna A, Ortigue S, et al. Speech arrest with stimulation
may not reliably predict language deficit after epilepsy surgery.
Neurology. 2006;66:592–594.
57. Wyler AR, Walker G, Somes G. The morbidity of long-term seizure
monitoring using subdural strip electrodes. J Neurosurg. 1991;74:
734–737.
58. Lee WS, Lee JK, Lee SA, et al. Complications and results of sub-
dural grid electrode implantation in epilepsy surgery. Surg Neurol.
2000;54:346–351.
712 Part V ■ Complementary and Special Techniques
59. Onal C, Otsubo H, Araki T, et al. Complications of invasive sub-
dural grid monitoring in children with epilepsy. J Neurosurg. 2003;
98:1017–1026.
60. Musleh W, Yassari R, Hecox K, et al. Low incidence of subdural
grid-related complications in prolonged pediatric EEG monitor-
ing. Pediatr Neurosurg. 2006;42:284–287.
61. Jobst BC, Williamson PD, Coughlin CT, et al. An unusual compli-
cation of intracranial electrodes. Epilepsia. 2000;41:898–902.
62. Hamer HM, Morris HH, Mascha EJ, et al. Complications of invasive
video-EEG monitoring with subdural grid electrodes. Neurology.
2002;58:97–103.
63. Wiggins GC, Elisevich K, Smith BJ. Morbidity and infection in
combined subdural grid and strip electrode investigation for
intractable epilepsy. Epilepsy Res. 1999;37:73–80.
64. Araki T, Otsubo H, Makino Y, et al. Efficacy if dexamethasone on
cerebral swelling and seizures during subdural grid EEG recording
in children. Epilepsia. 2006;47:176–180.
65. Sahjpaul RL, Mahon J, Wiebe S. Dexamethasone for morbidity
after subdural electrode insertion—a randomized controlled trial.
Can J Neurol Sci. 2003;30:340–348.
66. Sperling MR, O’Connor MJ. Comparison of depth and subdural elec-
trodes in recording temporal lobe seizures. Neurology. 1989;39:1497.
67. Spencer SS, Spencer DD, Williamson PD, et al. Combined depth
and subdural electrode investigation in uncontrolled epilepsy.
Neurology. 1990;40:74–79.
68. Brekelmans GJF, Van Emde Boas W, Velis DN, et al. Comparison of
combined versus subdural or intracerebral electrodes alone in
presurgical focus localization. Epilepsia. 1998;39:1290–1301.
69. Eisenschenk S, Gilmore RL, Cibula JE, et al. Lateralization of tem-
poral lobe foci: depth versus subdural electrodes. Clin Neurophysiol.
2001;112:836–844.
70. Wieser HG, Elger CE, Stodieck SR. The “foramen ovale electrode”:
a new recording method for the preoperative evaluation of
patients suffering from mesio-basal temporal lobe epilepsy.
Electroencephalogr Clin Neurophysiol. 1985;61:314–322.
71. Fernandez Torre JL, Alarcon G, Binnie CD, et al. Comparison of
sphenoidal, foramen ovale and anterior temporal placements for
detecting interictal epileptiform discharges in presurgical assess-
ment for temporal lobe epilepsy. Clin Neurophysiol. 1999;110:
895–904.
72. Marks DA, Katz A, Booke J, et al. Comparison and correlation of
surface and sphenoidal electrodes with simultaneous intracranial
recording: an interictal study. Electroencephalogr Clin Neurophysiol.
1992;82:23–29.
73. Otsubo H, Ochi A, Imai K, et al. High-frequency oscillations of
ictal muscle activity and epileptogenic discharges on intracranial
EEG in a temporal lobe epilepsy patient. Clin Neurophysiol. 2008;
119:862–868.
74. Hufnagel A, Dümpelmann M, Zentner J, et al. Clinical relevance of
quantified intracranial interictal spike activity in presurgical eval-
uation of epilepsy. Epilepsia. 2000;41:467–478.
75. Asano E, Muzik O, Shah A, et al. Quantitative interictal subdural
EEG analyses in children with neocortical epilepsy. Epilepsia. 2003;
44:425–434.
76. Engel J. Outcome with respect to epileptic seizures. In: Engel J Jr,
ed. Surgical Treatment of the Epilepsies. New York, NY: Raven Press;
1987:553–571.
77. Bautista RE, Cobbs MA, Spencer DD, et al. Prediction of surgical
outcome by interictal epileptiform abnormalities during intracra-
nial EEG monitoring in patients with extrahippocampal seizures.
Epilepsia. 1999;40:880–890.
78. Dümpelmann M, Elger CE. Visual and automatic investigation of
epileptiform spikes in intracranial EEG recordings. Epilepsia.
1999;40:275–285.
79. Palmini A, Gambardella A, Andermann F, et al. Intrinsic epilepto-
genicity of human dysplastic cortex as suggested by corticography
and surgical results. Ann Neurol. 1995;37:476–487.
80. Gambardella A, Palmini A, Andermann F, et al. Usefulness of focal
rhythmic discharges on scalp EEG of patients with focal cortical dys-
plasia and intractable epilepsy. Electroencephalogr Clin Neurophysiol.
1996;98:243–249.
81. Hirabayashi S, Binnie CD, Janota I, et al. Surgical treatment of
epilepsy due to cortical dysplasia: clinical and EEG findings.
J Neurol Neurosurg Psychiatr. 1993;56:765–770.
82. Turkdogan D, Duchowny M, Resnick T, et al. Subdural EEG pat-
terns in children with Taylor-type cortical dysplasia: comparison
with nondysplastic lesions. Clin Neurophysiol. 2005;22:37–42.
83. Guerreiro MM, Quesney LF, Salanova V, et al. Continuous elec-
trocorticogram epileptiform discharges due to brain gliosis.
J Clin Neurophysiol. 2003;20:239–242.
84. Zhang Y, van Drongelen W, Kohrman M, et al. Three-
dimensional brain current source reconstruction from intra-
cranial ECoG recordings. Neuroimage. 2008;42:683–695.
85. Sperli F, Spinelli L, Seeck M, et al. EEG source imaging in pedi-
atric epilepsy surgery: a new perspective in presurgical workup.
Epilepsia. 2006;47:981–990.
86. Jung WY, Pacia SV, Devinsky O. Neocortical temporal lobe
epilepsy: intracranial EEG features and surgical outcome. J Clin
Neurophysiol. 1999;16:419–425.
87. Wennberg R, Arruda F, Quesney LF, et al. Preeminence of
extrahippocampal structures in the generation of mesial tempo-
ral seizures: evidence from human depth electrode recordings.
Epilepsia. 2002;43:716–726.
88. Lee SA, Spencer DD, Spencer SS. Intracranial EEG seizure-onset
patterns in neocortical epilepsy. Epilepsia. 2000;41:297–307.
89. Wetjen N, Marsh WR, Myer FB, et al. Intracranial electroen-
cephalography seizure onset patterns and surgical outcomes in
non-lesional extratemporal epilepsy. J Neurosurg. 2009;110:
1147–1152.
90. Kutsy RL, Farrell DF, Ojemann GA. Ictal patterns of neocortical
seizures monitored with intracranial electrodes: correlation with
surgical outcome. Epilepsia. 1999;40:257–266.
91. McGonigal A, Bartolomei F, Régis J, et al. Stereoelectro -
encephalography in presurgical assessment of MRI-negative
epilepsy. Brain. 2007;130:3169–3183.
92. Spencer SS, Guimaraes P, Katz A, et al. Morphological patterns of
seizures recorded intracranially. Epilepsia. 1992;33:537–545.
93. Fisher R, Weber W, Lesser R, et al. High-frequency EEG activity
at the start of seizures. J Clin Neurophysiol. 1992;83:229–235.
94. Schiller Y, Cascino GD, Busacker NE, et al. Characterization and
comparison of local onset and remote propagated electrographic
seizures recorded with intracranial electrodes. Epilepsia. 1998;
39:380–388.
95. Blumenfeld H, Rivera M, McNally KA, et al. Ictal neocortical slow-
ing in temporal lobe epilepsy. Neurology. 2004;63:1015–1021.
96. Spencer SS, Kim J, Spencer DD. Ictal spikes: a marker of specific
hippocampal cell loss. Electroencephalogr Clin Neurophysiol.
1992;83:104–111.
97. Townsend JB, Engel J Jr. Clinicopathological correlation of
low voltage fast and high amplitude spike and wave mesial
temporal stereoencephalographic ictal onsets. Epilepsia. 1991;
32(suppl 3):21.
 Chapter 33 ■ Intracranial Monitoring: Depth, Subdural, and Foramen Ovale Electrodes
98. Schuh LA, Henry TR, Ross DA, et al. Ictal spiking patterns
recorded from temporal depth electrodes predict good outcome
after anterior temporal lobectomy. Epilepsia. 2000;41:316–319.
99. Faught E, Kuzniecky RI, Hurst DC. Ictal EEG wave forms from
epidural electrodes predictive of seizure control after temporal
lobectomy. Electroencephalogr Clin Neurophysiol. 1992;83:229–235.
100. Adam C, Saint-Hilaire JM, Richer F. Temporal and spatial charac-
teristics of intracerebral seizure propagation: predictive value in
surgery for temporal lobe epilepsy. Epilepsia. 1994;35:1065–1072.
101. Lieb JP, Engel J Jr, Babb TL. Interhemispheric propagation time
of human hippocampal seizures. I. Relationship to surgical out-
come. Epilepsia. 1986;27:286–293.
102. Zaveri HP, Pincus SM, Goncharova II, et al. Localization-related
epilepsy exhibits significant connectivity away from the seizure-
onset area. Neuroreport. 2009;20:891–895.
103. Brekelmans GJF, Velis DN, van Veelen CWM, et al. Intracranial
EEG seizure off-set termination patterns: relation to outcome of
epilepsy surgery in temporal lobe epilepsy. Epilepsia. 1998;39:
259–266.
104. Spencer SS, Spencer DD. Implication of seizure termination loca-
tion in temporal lobe epilepsy. Epilepsia. 1996;37:455–458.
105. Verma A, Lewis D, Van Landingham KE, et al. Lateralized seizure
termination: relationship to outcome following anterior tempo-
ral lobectomy. Epilepsy Res. 2001;47:9–15.
106. Fisher RS, Webber WR, Lesser RP, et al. High-frequency EEG activ-
ity at the start of seizures. J Clin Neurophysiol. 1992;9:441–448.
107. Engel J Jr, Bragin A, Staba R, et al. High-frequency oscillations:
what is normal and what is not? Epilepsia. 2009;50:598–604.
108. Bragin A, Wilson CL, Staba RJ, et al. Interictal high-frequency
oscillations (80–500 Hz) in the human epileptic brain: entorhinal
cortex. Ann Neurol. 2002;52(4):407–415.
109. Staba RJ, Wilson CL, Bragin A, et al. High-frequency oscillations
recorded in human medial temporal lobe during sleep. Ann
Neurol. 2004;56:108–115.
110. Worrell GA, Parish L, Cranstoun SD, et al. High-frequency oscil-
lations and seizure generation in neocortical epilepsy. Brain.
2004;127(pt 7):1496–1506.
111. Urrestarazu E, Chander R, Dubeau F, et al. Interictal high-
frequency oscillations (100–500 Hz) in the intracerebral EEG of
epileptic patients. Brain. 2007;130:2354–2366.
112. Worrell GA, Gardner AB, Stead SM, et al. High-frequency oscil-
lations in human temporal lobe: simultaneous microwire and
clinical macroelectrode recordings. Brain. 2008;131:928–937.
113. Jacobs J, Levan P, Chander R, et al. Interictal high-frequency
oscillations (80–500 Hz) are an indicator of seizure onset areas
independent of spikes in the human epileptic brain. Epilepsia.
2008;49(11):1893–1907
114. Jirsch JD, Urrestarazu E, LeVan P, et al. High-frequency oscilla-
tions during human focal seizures. Brain. 2006;129:1593–1608.
115. Khosravani H, Mehrotra N, Rigby M, et al. Spatial localization
and time-dependant changes of electrographic high frequency
oscillations in human temporal lobe epilepsy. Epilepsia. 2008
[Epub ahead of print].
713
116. Yamazaki M, Chan D, Tovar-Spinoza Z, et al. Interictal epilepto-
genic fast oscillations on neonatal and infantile EEGs in
hemimegalencephaly. Epilepsy Res. 2008 [Epub ahead of print].
117. Wu JY, Koh S, Sankar R, et al. Paroxysmal fast activity: an interic-
tal scalp EEG marker of epileptogenesis in children. Epilepsy Res.
2008;82:99–106.
118. Lesser RP, Lüders H, Klem G, et al. Extraoperative cortical func-
tional localization in patients with epilepsy. J Clin Neurophysiol.
1987;4:27–53.
119. Duchowny M, Jayakar P. Functional cortical mapping in chil-
dren. In: Devinsky O, Beric A, Dogali M, eds. Electrical and
Magnetic Stimulation of the Brain and Spinal Cord. New York, NY:
Raven Press, Ltd; 1993:149–154.
120. Jayakar P, Alvarez LA, Duchowny MS, et al. A safe and effective
paradigm to functionally map the cortex in childhood. J Clin
Neurophysiol. 1992;9:288–293.
121. Allison T, McCarthy G, Luby M, et al. Localization of functional
regions of human mesial cortex by somatosensory evoked poten-
tial recording and by cortical stimulation. Electroencephalogr Clin
Neurophysiol. 1996;100:126–140.
122. Neuloh G, Schramm J. Motor evoked potential monitoring for
the surgery of brain tumours and vascular malformations. Adv
Tech Stand Neurosurg. 2004;29:171–228.
123. Carne RP, O’Brien TJ, Kilpatrick CJ, et al. MRI-negative PET-
positive temporal lobe epilepsy: a distinct surgically remediable
syndrome. Brain. 2004;127:2276–2285.
124. Kurian M, Spinelli L, Delavelle J, et al. Multimodality imaging for
focus localization in pediatric pharmacoresistant epilepsy.
Epileptic Disord. 2007;9:20–31.
125. Michel CM, Grave de Peralta R, Lantz G, et al. Spatiotemporal
EEG analysis and distributed source estimation in presurgical
epilepsy evaluation. J Clin Neurophysiol. 1999;16:239–266.
126. Lantz G, Grave de Peralta R, Spinelli L, et al. Epileptic source
localization with high density EEG: how many electrodes are
needed? Clin Neurophysiol. 2003;114:63–69.
127. Regard M, Cook ND, Wieser HG, et al. The dynamics of cerebral
dominance during unilateral limbic seizures. Brain. 1994;117:
91–104.
128. Seeck M, Mainwaring N, Cosgrove R, et al. Neurophysiologic cor-
relates of implicit face memory in intracranial visual evoked
potentials. Neurology. 1997;49:1312–1316.
129. Grunwald T, Lehnertz K, Heinze HJ, et al. Verbal novelty detec-
tion within the human hippocampus proper. Proc Natl Acad Sci
U S A. 1998;95:3193–3197.
130. Axmacher N, Elger CE, Fell J. Ripples in the medial temporal lobe
are relevant for human memory consolidation. Brain. 2008;131:
1806–1817.
131. Vannucci M, Pezer N, Helmstaedter C, et al. Hippocampal
response to visual objects is related to visual memory functioning.
Neuroreport. 2008;19:965–968.
132. Sinai A, Bowers CW, Crainiceanu CM, et al. Eletrocorticographic
high gamma activity versus electrical cortical stimulation map-
ping of naming. Brain. 2005;128:1556–1570.

Electrocorticography
E
lectrocorticography (ECoG) is intraoperative electroen-
cephalography (EEG) recorded directly from the exposed
cerebral cortex. The technique is used for localizing
epileptic discharges and regions of impaired cortex prior to
resection. Together with direct cortical stimulation, the tech-
niques are used to identify motor, language, and sensory cortex
during tumor resection. The goal of ECoG is to separate regions
suitable for resection from those that should be left intact.
ECoG was the first application of intraoperative neurophys-
iology, a field that has since evolved to include many other tech-
niques (1). Even as far back as the 1940s, ECoG has been used
to localize epileptogenic tissue prior in the neurosurgical treat-
ment of partial epilepsy (2–7). The equipment for ECoG has
evolved since then, but the clinical goal remains relatively
unchanged.
The French group (8–10) developed a method called stere-
oencephalography (SEEG) to record from deep structures and
buried cortex that are not readily accessible by subdural or cor-
tical methods of recording. In this way, ECoG can be obtained
simultaneously from superficial and deep brain structures. A
review (11) of the SEEG technique in relation to other methods
of ECoG notes that SEEG exploration is well tolerated by the
majority of patients and overall complication rates of SEEG are
reported as being of the order of 5% (12,13). This is similar to
complication rates of subdural strips (14) and is somewhat less
than those of subdural grids (15). Although acute depth elec-
trodes appear to be “more invasive” than subdural recordings,
the complication rate is not higher.
ELECTRODES AND METHODS
OF RECORDING DIRECTLY FROM
CEREBRAL CORTEX
ECoG recording from the exposed cerebral cortex most often
uses strip and grid electrodes. These electrodes are arrays of
3-mm metal discs embedded in a flexible silicon rubber matrix.
Each disc is 1 cm apart measured center to center. Discs are
stainless steel or platinum alloy. Fine flexible wires connect to
each disc. These wires course through the silicon rubber matrix
and are gathered together to exit at one end of the strip or grid.
Together they form a ribbon cable or braided wire connector
cable leading to a multipin connector. A matched multipin con-
nector with grounded heavier wiring transmits to the standard
EEG machine electrode junction box. Strips and grids come in
many different sizes. Sizes used more often in surgery are 1 8
strips and 4 5 grids. For patients in whom these are left in
CHAPTER
MARC R. NUWER 34
place chronically in the subdural space for days or weeks to
catch epileptic seizures in an epilepsy monitoring unit, grid
sizes up to 8 8 are available.
When used in surgery, ECoG grids and strips tend to curl up
or pull slightly away from the cortex especially in sulci. This
results in poor electrical connection to some contacts. To pre-
vent corners pulling up, the surgeon should place saline-wetted
surgical Cottonoid over each strip or grid to weigh it down.
These strips and grids can be moved around to record from var-
ious regions of exposed cortex, or even under the dura or skull
edges, to record from as many regions as needed.
Montages are either bipolar among the adjacent contacts or
referential to a distant site. A clip electrode on muscle or a disc
scalp electrode can serve as a distant neutral reference. Some
users prefer the mastoid site. Because the amplitude of ECoG
is several-fold higher in amplitude than EEG picked up by
electrodes on the muscle or scalp, referential recordings are
often used without excessive channel contamination by an
active reference.
A low-frequency filter of 0.5 Hz and a high-frequency filter
of 70 Hz ensure adequate recording of epileptiform discharges
and background activity. The amplitude of EEG recorded
directly from the cortex is several-fold higher than EEG
recorded from the scalp. This allows for a relatively low sensitiv-
ity of 30 to 50 µV/mm to be used for ECoG recordings. A col-
lection of 20 to 40 cortical electrodes is recommended, which
often are composed of signals from more than one grid or strip.
Acute multicontact depth electrodes are used by some cen-
ters. This has been reported as very helpful for an acute explo-
ration of deep limbic and neocortical structures during
temporal lobectomy (16,17).
DIRECT CORTICAL STIMULATION
TECHNIQUE
Direct cortical stimulation is used to localize functional cortex.
This technique requires special instrumentation and precau-
tions. Generally equipment specifically designed for this pur-
pose is used. Contact with the cortical surface usually is made
using a bipolar stimulation wand or probe held against the cor-
tex. Alternately, the stimulation may be delivered through the
grid or strip electrodes; the latter stimulation technique is used
more often for the subacute subdural studies on epilepsy mon-
itoring units. The wand stimulation technique is more conven-
ient during surgery where the cortex is directly exposed. The
typical stimulating wand or probe is 30 cm long, with two
715
716 Part V ■ Complementary and Special Techniques
round metal ball tips separated by a few millimeters. Such a
wand can move among cortical locations quickly. This allows
for testing many different locations in a short time.
Stimuli are bipolar and biphasic. Bipolar means that electric
current runs between the two adjacent active contacts. Biphasic
means that the current switches polarity halfway through each
stimulus pulse. Biphasic stimulation is preferred in the ECoG
setting so as not to leave a net electrical polarization at either
metal–cortex contacts. Polarization could allow metal ions to
move onto the brain, which is undesirable especially for iron
ions that would serve as a source of future cortical irritation.
Polarization alters the electrical sensitivity of the underlying
cortex to subsequent stimulus pulses. Bipolar biphasic stimula-
tion equipment is available, but most stimulation equipment
for routine intraoperative monitoring is not biphasic. The latter
routine stimulation equipment is unsuitable for direct cortical
stimulation studies.
Typical direct cortical stimuli are 100 to 300 sec (0.1 to
0.3 msec) pulses delivered in trains of 50 pulses per second for
several seconds. A 5- to 7-second stimulation train is used for
language testing, whereas shorter trains are used for motor test-
ing. Stimulus pulses are delivered using currents up to 14 mil-
liamperes (mA) for the usual wand or probe designs. This is
sufficient to disrupt function at the cortex underlying the wand’s
location. Often one starts with a lower intensity, for example, at
4 to 6 mA, increasing gradually to higher intensities if no desired
or adverse effects are seen at the lower intensities at each site.
Either constant current or constant voltage stimulation may be
used. Duration longer than 8 seconds should be avoided.
During direct cortical stimulation, ECoG is recorded from
nearby sites. This is carried out to check whether epileptic-like
discharges occur with the stimulations. When epileptic-like dis-
charges last longer than the stimulation, they are referred to as
afterdischarges. Afterdischarges appear as rhythmic repetitive
high-voltage spikes or polyspike–wave complexes lasting sec-
onds or occasionally minutes. Such prolonged afterdischarges
are usually subclinical or associated with subtle clinical behav-
ioral manifestations (18). When afterdischarges occur, stimula-
tion may have spread to other cortical regions. Therefore those
stimulation trials should not be used to localize cortical func-
tions because they include activation of areas away from the
wand’s tip. Afterdischarges may herald epileptic seizures occur-
ring with further stimulation at that site or intensity. This may
prompt moving the stimulating wand to a different site or to
reducing the stimulus intensity.
Cortical stimulation can provoke epileptic seizures. Cold
Ringer’s lactate solution can treat seizures provoked by cortical
stimulation. Such irrigation rapidly stops partial seizures (19).
INTERPRETATION OF ECOG AND DIRECT
CORTICAL STIMULATION
The goal of ECoG is to determine locations of epileptic or oth-
erwise impaired cortex. The recording usually shows EEG fea-
tures similar to those seen with scalp EEG. Frontocentral fast
activity, low-amplitude generalized slowing, and frontopolar
triangular slow waves are seen during many types of anesthesia.
Recordings above and below the sylvian fissure would show
more fast activity above and less below. ECoG background
activity is more sharply contoured than typically seen on scalp
EEG recordings; this sharply contoured EEG is similar to the
scalp EEG over a skull breach. This sharply contoured back-
ground needs to be distinguished from epileptic activity during
interpretation. Some ECoG epileptiform spikes may not be
apparent on concurrent scalp-recorded EEG (20). Areas of
impaired cortex may show relatively decreased fast activity,
increased slow activity, or epileptic spikes. A knowledgeable and
experienced physician neurophysiologist identifies these
changes and thereby identifies regions with impaired function.
ECoG is used for certain patients during epilepsy surgery.
Most often this is when the epileptogenic zone was inadequately
localized by other methods. ECoG is used to further define the
resection limits. For patients with very frequent seizures during
either scalp or invasive EEG monitoring, intraoperative ictal
ECoG may be useful to demarcate the electrographic ictal onset
zone more precisely. An interictal ECoG is much more com-
monly obtained. Interictal spike discharges usually arise from
the epileptogenic zone but also extend to cortex beyond it. Those
regions are defined and discussed with the surgeon to tailor the
resection to the needs of the individual patient.
Sometimes ECoG is performed after resection. Epileptiform
discharges observed at the resection margins may just represent
“injury potentials” or “injury spikes,” not the location of addi-
tional epileptic tissue. Epileptic spikes distant from the resection
margin are more concerning, and frontal lobe epilepsy patients
with epileptiform discharges three or more gyri away from the
resection site indicated likelihood of seizure recurrence.
To avoid anesthetic effects, the anesthesia is lightened or
minimized during the ECoG. Many anesthetics can alter the
presence of epileptiform discharges sought in ECoG. The
effects of the anesthetic on the ECoG are clearly less important
than the anesthetic’s clearance time. Propofol clears adequately
to allow good ECoG in most patients. Nitrous oxide and nar-
cotics may be used during the ECoG to avoid pain or con-
sciousness during the procedure, although these may affect the
epileptiform discharges seen on the recording.
Direct cortical stimulation testing is interpreted by the pres-
ence of movement or sensation or the disruption of language.
Neuromuscular blockade needs to be avoided. For motor test-
ing, anesthesia cannot be deep. Observations of responses can
be made visually, or EMG electrodes can be placed in various
muscles to record movements under the surgical drapes.
Stimulations then can identify various face, arms, and leg
regions. Finding stimulation effects at some cortical areas
shows that the anesthetic depth is adequate and the technique
is working properly. Considerable individual variability exists
in location of eloquent cortex regions for language, motor, and
sensory function (21–23). Anatomic locations need to be deter-
mined for each eloquent function for each individual patient.
Stimulation mapping allows for best resection of pathologic tis-
sue while preserving function (24,25). Functional neuroimag-
ing navigation methods are usually combined with stimulation
mapping; such a dual approach is the current best guide for the
limits of a resection (26,27).

The patient is awake during surgery for sensory and language
testing. The conscious patient is extubated while restrained on
the table (28). The primary sensory cortex can be mapped using
such stimulation in the awake patient. For some awake epilepsy
patients, stimulation near a seizure focus will cause the patient’s
typical aura.
For language testing, a neuropsychologist or other trained lan-
guage specialist administers language tasks for different aspects of
auditory and visual language processing and short-term memory.
The clinical neurophysiologist delivers the stimulations and
observes for afterdischarges while the surgeon holds the wand and
the language specialist administers oral and visual testing. Most
adult patients tolerate this awake craniotomy language testing
adequately. Stimulation at a series of cortical locations can iden-
tify functional language areas. Regions without noticeable disrup-
tion at 12 to 14 mA are considered candidates for resection.
The asleep–awake–asleep technique requires cooperative
patients (28). Initial anesthesia is often propofol or sodium
thiopental with endotracheal intubation. Local anesthetic is infil-
trated around the operative site margins. The patient is awakened
and extubated once the craniotomy exposure is achieved.
Consistent intraoperative language performance baseline is
needed before stimulation testing. Without baseline consistency,
any observed language disruption might be due to inconsistent
performance rather than specific anatomic disruption due to
stimulation. After mapping is completed, the patient is reintu-
bated and remains under general anesthesia for the remainder of
the procedure.
Language mapping includes several types of tasks. Visual
object naming involves the naming of objects presented as line
drawings from the Boston Naming Test (29). Drawings are pre-
sented that the patient could name quickly in preoperative test-
ing (21). Naming is a useful screening task for language during
cortical stimulation mapping because naming deficits are fre-
quently part of aphasia syndromes. Speech arrest or anomia may
occur upon stimulating a cortical region important for lan-
guage, although not all cortical language sites produce naming
deficits (30,31). Different cortical sites may reveal stimulation
deficits in sentence reading (32). Another easily accomplished
stimulation language test is word generation in which patients
generate lists of words that begin with a certain letter or are of a
certain category (e.g., animals). In auditory responsive naming,
the patient names objects, that is, the answer for “tall pink bird”
would be “flamingo.” Auditory responsive naming and word
generation tasks activate frontal language areas (33,34). Visual
responsive naming differs from auditory responsive naming
because the question is written on a flash card.
Stimulation mapping effects are well localized. Moving the
stimulating electrodes 0.5 to 2 cm, even along the same gyrus,
yields different functional effects (32,35). A site with consistent
language function can be 2 cm from a site on the same gyrus
without language disruption (21). Language function often is
localized to small 1 to 2 cm 2 regions (32,36). Postoperative lan-
guage deficits occur when resection margins are within 1.5 to
2.0 cm of language function sites. No permanent language
deficits were reported if the surgical margins remain 2.0 cm
from language sites (32,36,37).
Chapter 34 ■ Electrocorticography 717
GENERAL ANESTHESIA AND CORTICAL
EEG ACTIVITY
Anesthetic drugs tend to either excite or depress the EEG and
follow a pattern summarized by Winters (38). The activating or
exciting anesthetic agents are propofol and etomidate. They
inhibit brainstem arousal mechanisms through GABAA receptor
mechanisms (39,40). The resulting EEG patterns resemble slow-
wave sleep including vertex sharp waves and spindles. Despite
early reports, propofol does not produce EEG epileptiform
activity (41–43). Propofol is a popular agent for ECoG recording
and before direct cortical stimulation. Because of its rapid
metabolism, propofol can induce unconsciousness during por-
tions of the craniotomy. Yet it is eliminated rapidly enough to
allow intraoperative awakening, a major advantage for ECoG
(44,45). Etomidate is another activating or exciting anesthetic
agent. Like propofol, it enhances epileptic activity at low doses
(0.1 mg/kg). It may produce seizures in patients with epilepsy
(46). The paradoxical excitatory effect of low-dose propofol may
result from a membrane-level interaction between the GABAA
current and an intrinsic membrane slow potassium current (M-
current) according to a proposal by McCarthy et al. (47). In that
theory, a potential network mechanism underlying the genera-
tion of propofol-induced paradoxical excitation consists of an
interneuron antisynchrony, one that also produces an EEG beta
rhythm at low drug concentrations.
The commonly used volatile inhalation anesthetic agents are
sedating anesthetics. They include halothane, desflurane, isoflu-
rane, and sevoflurane. They cause immobility through multiple
molecular targets predominantly in the spinal cord, including
GABAA receptors, glycine receptors, glutamate receptors, and
two-pore-domain potassium channels (48). These agents cause
rhythmic frontocentral alpha activity. This rhythmic fast activ-
ity can be abolished by nitrous oxide (49). Hyperventilation
with high concentrations of these agents, except for desflurane,
can produce epileptiform spikes or even electrographic non-
convulsive seizures. These agents can reduce (50) or increase
(51) epileptiform discharges at standard anesthetic levels. They
have little effect at low levels (52).
Barbiturates are similar to inhalational agents in producing
activation and fast activity at low doses and a depressant effect
leading to burst suppression at higher doses. These bursts may
have very sharp components that could be misinterpreted as
epileptiform. At low dose, the short-acting barbiturate metho-
hexital (0.5 mg/kg) does activate epileptic spike activity during
ECoG, and so it has been used deliberately to identify seizure
foci (46,53–55). Methohexital also can produce seizures.
Benzodiazepines produce frontal beta activity with a decrease in
alpha activity at low doses. Some patients show epileptiform
spike activity in that frontal beta activity. Slowing of higher fre-
quencies through inhibitory mechanisms may produce the bar-
biturate beta activity (56). Higher benzodiazepine doses
produce generalized slowing. As an anticonvulsant, benzodi-
azepines are avoided if ECoG is planned for identification of
seizure foci. Nitrous oxide produces a frontocentral fast activity
when used by itself. When used with inhalational agents, it pro-
duces variable results. Nitrous oxide is proconvulsant when
718 Part V ■ Complementary and Special Techniques

combined with some agents. It may increase or suppress epilep-
tic spike activity during ECoG depending on which other
agents are used (57,58). Droperidol and dexmedetomidine are
other adjunctive agents sometimes used during ECoG proce-
dures. These seem compatible with ECoG. Narcotic drugs can
increase epileptiform spiking, including spiking or seizures
from sites outside the epileptogenic area (59–61). Remifentanil
may increase epileptiform discharges at higher doses (62) but
not with lower-dose continuous infusions (63).
Overall, light anesthesia enhances the ability to find epilep-
tiform spiking during ECoG. Propofol is an example of an
anesthetic agent that can be reduced sufficiently quickly to
achieve a light anesthesia, and which does not suppress intra-
operative epileptiform abnormalities. Methohexital or other
drugs sometimes are used to enhance epileptiform abnormali-
ties during ECoG.
APPLICATION IN EPILEPSY SURGERY
Spikes occur not only in the epileptogenic zone but also in
other cortical areas. Even the most active spiking area may not
coincide with the location of the lesion (6). The goal in ECoG
is to localize the region to be resected. Not all ECoG epileptic
spike locations are to be resected. Interictal epileptic distur-
bances recorded at ECoG in patients with temporal lobe
epilepsy (TLE) often involve (Fig. 34.1) adjacent regions
(18,64–77). Some reports suggest that interictal epileptic dis-
turbance recorded intraoperatively should guide surgical resec-
tion (18,68,78–83)). Others have emphasized that standardized
lobectomy could be performed without ECoG control for
patients with typical mesial temporal epilepsy (84–86). ECoG
recording after a temporal resection is controversial. Some
authors believe that the persistence of spike activity after surgi-
cal excision is of little or no prognostic value (2,87–89). Other
investigators advocate that persistent spike activity after surgi-
cal removal predicts an unsatisfactory outcome (16,79,90,91).
Recent reports (92) showed that intraoperative ECoG activ-
ities can be analyzed with respect to more complex spike pat-
terns, particularly in TLEs; namely, in addition to focal spiking
with or without propagation, focal slowing in the theta or delta
range and ictaform ECoG patterns were found. Furthermore,
high-frequency oscillations (HFOs; (93)) in the range of 250 to
600 Hz (fast ripples) have been identified in the ECoGs
recorded from the hippocampus and parahippocampal areas of
patients with TLE, which may be used to identify brain epilep-
togenic areas. Rates and durations of HFOs were found (94) to

Figure 34.1 ECoG recording showing spontaneous multifocal interictal spiking involving the second temporal gyrus ante-
riorly (1–2), electrode 6 in the first temporal gyrus, and the mesial temporal lobe structures synchronously (A1–2, P1–2).
In addition, widespread interictal epileptiform bursts involving the mesial temporal lobe structures as well as the tempo-
ral neocortex were recorded. (Reproduced from Stefan, H., Quesney, L.F., Abou-Khalil, B., and Olivier, A. 1991.
Electrocorticography in temporal lobe epilepsy surgery. Acta Neurol Scand. 83:65–72. )
 Chapter 34 ■ Electrocorticography 719

Figure 34.2 Preexisting electrocorticogram in a patient with a right frontocentral cortical dysplastic lesion on MRI. The
tracing shows electrographic seizure activity recorded multifocally. A 11>2 to 2 Hz rhythmic activity was seen at electrode 3;
rhythmic polyspike activity at 10 Hz recorded maximally at electrodes 1 and 7; rhythmic 2 to 3 Hz spike-and-wave activ-
ity recorded at electrode 10. (Reproduced from Palmini, A., Gambardella, A., Andermann, F., et al. 1995. Intrinsic epilep-
togenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol. 37:476–487.)

be significantly higher in the seizure onset zone than outside,
and occurred to a large extent independently of spikes.
In extratemporal surgery, interictal epileptiform spiking is
often widespread (95–100) even when the patient has a circum-
scribed lesion. In neocortical epilepsy, ECoG mapping of the
interictal epileptogenic area is useful to tailor the surgical resec-
tion (101). The distribution and abundance of interictal spiking
recorded at ECoG is a prognostic indicator of postsurgical
seizure control outcome (100). Both preexcision epileptic
abnormality recorded from less than two gyri away and absence
of postexcision epileptic activity distant to the resection border
strongly predicted a favorable outcome. Residual spiking lim-
ited to the resection border did not significantly correlate with
outcome. The presence of a circumscribed frontal lobe lesion
was also significantly correlated with favorable outcome. In
frontal lobe epilepsy, postexcision residual spiking indicates a
poor outcome in terms of seizure control (102).
ECoG is useful for localizing cortical dysplasia. The region of
the lesion shows rhythmic electrographic seizure discharges
(Fig. 34.2) corresponding to MRI evidence for cortical dysplasia.
A good surgical outcome in these patients is proportional to the
completeness of the lesion’s surgical removal (103) (Fig. 34.3).
Afterdischarge locations during direct cortical stimulation
do not correlate with impaired function or pathology and often
arise at normal cortex. Stimulation at certain anatomical sites,
for example, hippocampus, is more predisposed to elicit after-
discharges (104–106). Electrical stimulation thresholds are
often increased in damaged areas (107,108), so that impaired
regions tend not to have afterdischarges (16,107).
SUMMARY
ECoG is a method to detect cortical regions with substantial
epileptiform interictal discharges during surgery. Direct corti-
cal stimulation studies provide a means to identify language,
motor, and sensory regions during a craniotomy. These tech-
niques can help to determine the limits for a resection. They are
used during epilepsy and tumor surgery. They can be adversely
720 Part V ■ Complementary and Special Techniques

Figure 34.3 A: Baseline preexcision ECoG prior to methohexital injection showing active limbic epileptiform activity and
a single independent temporal neocortical spike. B: Preexcision ECoG 1 minute after injection of 40 mg methohexital show-
ing activation of epileptiform activity recorded independently from limbic and temporal neocortical structures. A slight
burst-suppression pattern developing over neocortex.

Figure 34.3 (continued) C: Postexcision ECoG showing activation of temporal neocortical epileptiform activity and burst-
suppression pattern. (Reproduced from Wennberg, R., Quesney, F., Olivier, A., and Dubeau, F. 1997b. Induction of burst-
suppression and activation of epileptiform activity after methohexital and selective amygdalo-hippocampectomy.
Electroencephalogr Clin Neurophysiol. 102:443–451.)
affected or limited by the anesthetic agents used. Amount of
spiking from diffuse or remote cortical regions ECoG can pre-
dict postoperative seizure control.
REFERENCES
1. Nuwer, MR. Intraoperative Monitoring of Neural Function:
Handbook of Clinical Neurophysiology. Amsterdam: Elsevier; 2008.
2. Ajmone Marsan C, Baldwin M. Electrocorticography. In: Baldwin
M, Bailey P, eds. Temporal Lobe Epilepsy Springfield, IL: Charles C
Thomas; 1958: 368–395.
3. Jasper, HH. Electrocorticogram in man. Electroencephalogr Clin
Neurophysiol. 1949;2(suppl):16–29.
4. Marshall, C, Walker, AE. Electrocorticography. Bull John Hopkins
Hosp. 1949;85:344–359.
5. Meyer, HR., Knott R., Hayne RA et al and Sweeney, D.B. The sur-
gery of epilepsy. Limitations of the concept of the corticoelectro-
graphic “spike” as an index of the epileptogenic focus.
J Neurosurg. 1950;7:337–346.
6. Penfield W, Jasper HH Epilepsy and the Functional Anatomy of the
Human Brain. Boston, MA: Little, Brown; 1954.
7. Walker AE, Marshall C, Beresford EM Electrocorticographic char-
acteristics of the cerebrum posttraumatic epilepsy. Assoc Res Nerv
Men Dis. 1947;26:502–515.
8. Bancaud J, Talairach J, Bonis A, et al. La stéréoélectroencephalo-
graphie dans l’épilepsie: informations neurophysiopathologiques
apportées par l’investigation fonctionelle stéréotaxique. Paris:
Masson & Cie; 1965.
9. Talairach J, Bancaud J, Szickla G, et al. Approche nouvelle de la
chirurgie de l’épilepsie: méthodologie stéréotaxique et résultats
thérapeutiques. Neurochirurgie. 1974;20:1–240.
Chapter 34 ■ Electrocorticography 721
10. Chauvel P, Buser P, Badier J, et al. La “zone épileptogène” chez
l’homme: représentation des événements intercritiques par
cartes spatio-temporelles. Rev Neurol (Paris). 143:443–550.
11. McGonigal A, Bartolomei F, Régis J, et al. Stereoelectro -
encephalography in presurgical assessment of MRI-negative
epilepsy. Brain. 2007; 130(pt 12):3169–3183. [Epub September 12,
2007].
12. Guenot M, Isnard J, Ryvlin P, et al. Neurophysiological monitoring
for epilepsy surgery: the Talairach SEEG method. Stereo -
electroencephalography Indications, results, complications and
therapeutic applications in a series of 100 consecutive cases.
Stereotact Funct Neurosurg. 77:29–32.
13. Cossu M, Cardinale F, Castana L, et al. Stereoelectro -
encephalography in the presurgical evaluation of focal epilepsy: a
retrospective analysis of 215 procedures. Neurosurgery. 2005;57:
706–718.
14. Alarcon G, Valentin A, Watt C, et al. Is it worth pursuing surgery
for epilepsy in patients with normal neuroimaging? J Neurol
Neurosurg Psychiatry. 2006;77:474–480.
15. Hamer HM, Morris HH, Mascha MS, et al. Complications of inva-
sive video-EEG monitoring with subdural grid electrodes.
Neurology. 2002;58:97–103.
16. Gloor P. ntributions of electroencephalography and electrocor-
ticography to the neurosurgical treatment of the epilepsies. In:
Purpura DP, Penry JK, Walker RD, eds. Neurosurgical
Management of Epilepsies. New York: Raven Press; 1975: 59–105.
17. Wennberg R, Quesney F, Olivier A et al. Mesial temporal versus lat-
eral temporal interictal epileptiform activity: comparison of
chronic and acute intracranial recordings. Electroencephalogr Clin
Neurophysiol. 1997;102:486–494.
18. Stefan H, Quesney LF, Abou-Khalil, B., et al Electrocorticography in
temporal lobe epilepsy surgery. Acta Neurol Scand. 1991;83:65–72.
722 Part V ■ Complementary and Special Techniques
19. Sartorius CJ, Berger MS. Rapid termination of intraoperative
stimulation-evoked seizures with application of cold Ringer’s lac-
tate to the cortex. Technical note. J Neurosurg. 1998;88:349–351.
20. Tao JX, Ray, A., Hawes-Ebersole, S, et al Intracranial EEG sub-
strates of scalp EEG interictal spikes. Epilepsia. 2005;46:669–676.
21. Ojemann GA. Individual variability in cortical localization of lan-
guage. J Neurosurg. 1979;50:164–169.
22. Uematsu S, Lesser R, Fisher RS, et al. Motor and sensory cortex in
humans: topography studied with chronic subdural stimulation.
Neurosurgery. 1992;31:59–71; discussion 71–72.
23. Branco DM, Coelho TM, Branco BM, et al. Functional variability
of the human cortical motor map: electrical stimulation findings
in perirolandic epilepsy surgery. J Clin Neurophysiol. 2003;20:
17–25.
24. Duffau, H, Lopes M, Arthuis F, et al. Contribution of intraopera-
tive electrical stimulations in surgery of low grade gliomas: a com-
parative study between two series without (1985–96) and with
(1996–2003) functional mapping in the same institution. J Neurol
Neurosurg Psychiatry. 2005;76:845–851.
25. Duffau H, Capelle L, Sichez J, et al. Intra-operative direct electrical
stimulations of the central nervous system: the Salpetriere experi-
ence with 60 patients. Acta Neurochir (Wien). 1999;141:
1157–1167.
26. Duffau H. Intraoperative direct subcortical stimulation for identi-
fication of the internal capsule, combined with an image-guided
stereotactic system during surgery for basal ganglia lesions. Surg
Neurol. 2000;53:250–254.
27. Kamada K, Todo T, Masutani Y, et al. Combined use of tractogra-
phy-integrated functional neuronavigation and direct fiber stimu-
lation. J Neurosurg. 2005;102:664–672.
28. Huncke K, Van De Wiele B, Fried I et al The asleep-awake-asleep
anesthetic technique for intraoperative language mapping.
Neurosurgery. 1998;42:1312–1316; discussion 1316–1317.
29. Kaplan E, Goodglass, H Weintraub, S. Boston Naming Test.
Philadelphia, PA: Lea and Febiger; 1976.
30. Ojemann G, Mateer, C. Human language cortex: localization of
memory, syntax, and sequential motorphoneme identification sys-
tems. Science. 1979;205:1401–1403.
31. Ojemann GA. Functional mapping of cortical language areas in
adults. Intraoperative approaches. Adv Neurol. 1993;63:155–163.
32. Ojemann GA. Brain organization for language from the perspec-
tive of electrical stimulation mapping. Behav Brain Sci. 1983;2:
189–230.
33. Bookheimer SY, Zeffiro TA, Blaxton, et al T Regional cerebral
blood flow during auditory responsive naming: evidence for cross-
modality neural activation. Neuroreport. 1998;9:2409–2413.
34. Cuenod CA, Bookheimer SY, Hertz-Pannier L, et al Functional
MRI during word generation, using conventional equipment: a
potential tool for language localization in the clinical environ-
ment. Neurology. 1995;45:1821–1827.
35. Ojemann, GA, Whitaker HA. Language localization and variabil-
ity. Brain Lang. 1978;6:239–260.
36. Ojemann A, Cortical organization of language. J Neurosci.
1991;11:2281–2287.
37. Haglund MM, Berger MS, Shamseldin M, et al Cortical localiza-
tion of temporal lobe language sites in patients with gliomas.
Neurosurgery. 1994;34:567–576; discussion 576.
38. Winters WD. Effects of drugs on the electrical activity of the brain:
anesthetics. Annu Rev Pharmacol Toxicol. 1976;16:413–426.
39. Nelson LE, Guo TZ, Lu, J., JCBNP et al The sedative component of
anesthesia is mediated by GABA(A) receptors in an endogenous
sleep pathway. Nat Neurosci. 2002;5:979–984.
40. Rudolph U, Antkowiak B. Molecular and neuronal substrates for
general anaesthetics. Nat Rev Neurosci. 2004;5:709–720.
41. Ebrahim ZY, Schubert, Van Ness et al The effect of propofol on the
electroencephalogram of patients with epilepsy. Anesth Analg.
1994;78:275–279.
42. Samra SK, Sneyd JR, Ross DA, et al and Henry, T.R. Effects of
propofol sedation on seizures and intracranially recorded epilepti-
form activity in patients with partial epilepsy. Anesthesiology.
1995;82:843–851.
43. Cheng MA, Tempelhoff R, Silbergeld DL, et al. Large-dose propo-
fol alone in adult epileptic patients: electrocorticographic results.
Anesth Analg. 1996;83(1):169–174.
44. Skucas, AP Artru A. Anesthetic complications of awake cran-
iotomies for epilepsy surgery. Anesth Analg. 2006;102:882–887.
45. Soriano SG, Eldredge EA, Wang FK, et al. The effect of propofol on
intraoperative Electrocorticography and cortical stimulation during
awake craniotomies in children. Paediatr Anaesth. 2000;10: 29–34.
46. Rampil IJ. Electroencephalogram. In: Albin MA, ed. Textbook of
Neuroanesthesia with Neurosurgical and Neuroscience Perspectives .
New York: McGraw-Hill; 1997: 193–220.
47. McCarthy MM, Brown, EN Kopell N. Potential network mecha-
nisms mediating electroencephalographic beta rhythm changes
during propofol-induced paradoxical excitation. J Neurosci.
2008;28:13488–13504.
48. Grasshoff C, Rudolph U Antkowiak, B. Molecular and systemic
mechanisms of general anaesthesia: the ‘multi-site and multiple
mechanisms’ concept. Curr Opin Anaesthesiol. 2005;18:386–391.
49. Yli-Hankala A. The effect of nitrous oxide on EEG spectral power
during halothane and isoflurane anaesthesia. Acta Anaesthesiol
Scand. 1990;34:579–584.
50. Asano E, Benedek K, Shah A, et al. Is intraoperative electrocor-
ticography reliable in children with intractable neocortical
epilepsy? Epilepsia. 2004;45(9):1091–1099.
51. Endo T, Sato K, Shamato H, Effects of sevoflurane on electrocor-
ticography in patients with intractable temporal lobe epilepsy. J
Neurosurg Anesthesiol. 14:59–62.
52. Fiol ME, Boening JA, Cruz-Rodriguez R, Effect of isoflurane
(Forane) on intraoperative electrocorticogram. Epilepsia. 1993;34:
897–900.
53. Ajmone Marsan C, O’Connor M. Electrocorticography. In:
Remond A, ed. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol. 10C. Amsterdam: Elsevier; 1973: 3–49.
54. Musella L, Wilder BJ, Schmidt RP. Electroencephalographic activa-
tion with intravenous methohexital in psycho-motor epilepsy.
Neurology (Minneapolis). 1971:21:594–602.
55. Wennberg R, Quesney F, Olivier A, et al Induction of burst-sup-
pression and activation of epileptiform activity after methohexital
and selective amygdalo-hippocampectomy. Electroencephalogr
Clin Neurophysiol. 1997;102:443–451.
56. Jensen O, Goel P, Kopell N, et al. On the human sensorimotor-cor-
tex beta rhythm: sources and modeling. Neuroimage. 2005;26:
347–355.
57. Artru AA, Lettich E, Colley PS, et al. Nitrous oxide: suppression of
focal epileptiform activity during inhalation, and spreading of
seizure activity following withdrawal. J Neurosurg Anesthesiol.
1990;2(3):189–193.
58. Hosain S, Nagarajan L, Fraser R, et al. Effects of nitrous oxide on
electrocorticography during epilepsy surgery. Electroencephalogr
Clin Neurophysiol. 1996;102:340–342.
59. Tempelhoff R, Modica PA, Bernardo KL, et al.I Fentanylinduced
electrocorticographic seizures in patients with complex partial
epilepsy. J Neurosurg. 1992;77:201–208.

60. Cascino GD, So EL, Sharbrough FW, et al. Alfentanil induced
epileptiform activity in patients with partial epilepsy. J Clin
Neurophysiol. 1993;10(4):520–525.
61. Manninen PH, Burke SJ, Wennberg R, et al. Intraoperative local-
ization of an epileptogenic focus with alfentanil and fentanyl.
Anesth Analg. 1999;88:1101–1106.
62. Wass CT, Grady RE, Fessler AJ, et al. The effects of remifentanil on
epileptiform discharges during intraoperative electrocorticogra-
phy in patients undergoing epilepsy surgery. Epilepsia.
2001;42:1340–1344.
63. Herrick IA, Craen RA, Blume WT, et al. Sedative doses of remifen-
tanil have minimal effect on ECoG spike activity during awake
epilepsy surgery. J Neurosurg Anesthesiol. 2002;14(1):55–58.
64. Cendes F, Dubeau F, Olivier A, et al. Increased neocortical spiking
and surgical outcome after selective amygdalo- hippocampectomy.
Epilepsy Res. 1993;16:195–206.
65. Devinsky O, Canevini MP, Sato S, et al. Quantitative electrocor-
ticography in patients undergoing temporal lobectomy.
J Epilepsy. 1992;5:178–185.
66. Engel J Jr, Driver MV, Falconer MA. Electrophysiological correlates
of pathology and surgical results in temporal lobe epilepsy. Brain.
1975;98:129–156.
67. Gastaut H, Naquet R, Vigouroux R, et al. Étude électrographique
chez l’homme et chez l’animal des décharges épileptiques dites
“psychomotrices.” Rev Neurol. 1953;88:310–354.
68. Graf M, Niedermeyer E, Schiemann J, et al. Electrocorticography:
information derived from intraoperative recordings during seizure
surgery. Clin Electroencephalogr. 1984;15:83–91.
69. Kajtor F, Hullay J, Farago L, et al. Electrical activity of the hip-
pocampus of patients with temporal lobe epilepsy. AMA Arch
Neurol Psychiatry. 1958;80:25–38.
70. Niedermeyer, E. 1982. Electrocorticography. In Electro-
encephalography, E. Niedermeyer and F. Lopes da Silva, Eds., pp.
537–541. Munich: Urban & Schwarzenberg.
71. Niedermeyer E. Electrocorticography. In: Niedermeyer E, Lopes da
Silva F, eds. Electroencephalography, Basic Principles, Clinical
Applications and Related Fields, 2nd ed. Baltimore, MD: Urban and
Schwarzenberg; 1987:613–617
72. Quesney LF, Niedermeyer E. Electrocorticography. In:
Niedermeyer E, Lopes da Silva F, eds. Electroencephalography: Basic
Principles, Clinical Applications, and Related Fields, 3rd ed.
Baltimore, MD: Williams & Wilkins; 1993:695–699.
73. Rasmussen TB. Surgical treatment of complex partial seizures:
results, lessons and problems. Epilepsia. 1983;24(suppl 1):S65–S76.
74. Tuunainen A, Nousiainen U, Mervaala E, et al. Postoperative EEG
and electrocorticography: relation to clinical outcome in patients
with temporal lobe surgery. Epilepsia. 1994;35:1165–1173.
75. Walker AE. Temporal lobectomy. J Neurosurg. 1967;26: 642–649.
76. Walker AE. Surgery for epilepsy. In: Magnus O, Lorentz de Haas
AM, eds. Handbook of Clinical Neurology, Vol. 15: The Epilepsies.
Amsterdam: North Holland; 1974:739–757
77. Quesney LF, Abou-Khalil B, Cole A, et al. Pre-operative extracra-
nial and intracerebral EEG investigation in patients with temporal
lobe epilepsy: trends, results and review of pathophysiologic
mechanisms. Acta Neurol Scand. 1988;78:52–61.
78. Ajmone Marsan C. Depth electrography and electrocorticography.
In: Aminoff MJ, ed. Electrodiagnosis in Clinical Neurology. New
York: Churchill Livingstone; 1980:167–196.
79. Bengzon ARA, Rasmussen T, Gloor P, et al. Prognostic factors in
surgical treatment of temporal lobe epilepsy. Neurology
(Minneapolis). 1968;18:717–731.
Chapter 34 ■ Electrocorticography 723
80. Fiol ME, Gates JR, Torres F, et al. The prognostic value of residual
spikes in the postexcision electrocorticogram after temporal lobec-
tomy. Neurology. 1991;41:512–516.
81. Jennum P, Dhuna A, Davies K, et al. Outcome of resective surgery
for intractable partial epilepsy guided by subdural electrode
arrays. Acta Neurol Scand. 1993;87:434–437.
82. Ojeman GA. Basic mechanisms implicated in surgical treatments
of epilepsy. In: Lockard JS, Ward AA Jr, eds. Window to Brain
Mechanisms. New York: Raven Press; 1980:261–277.
83. Primrose DC, Ojeman GA. Outcome of resective surgery for tem-
poral lobe epilepsy. In: Lüders H, ed. Epilepsy Surgery. New York:
Raven Press; 1992:601–611.
84. Engel J Jr, ed. Approaches to the localization of the epileptogenic
lesion. Surgical Treatment of the Epilepsies. New York: Raven Press;
1987:75–100.
85. Spencer DD, Inserni J. Temporal lobectomy. In: Lüders HO, ed.
Epilepsy Surgery. New York: Raven Press; 1992:533–545
86. Doyle WK, Spencer DD. Anterior temporal resections. In: Engel
J Jr, Pedley T, eds. Epilepsy: A Comprehensive Textbook. Vol. 2.
Philadelphia, PA: Lippincott–Raven; 1997:1807–1817
87. Chatrian GE, Quesney LF. Intraoperative electrocorticography. In:
Engel J Jr, Pedley TA, eds. Epilepsy: A Comprehensive Textbook. New
York: Raven Press; 1997
88. Gibbs FA, Amador L, Rich C. Electro -encephalographic findings
and therapeutic results in surgical treatment of psychomotor
epilepsy. In: Baldwin M, Bailey P, eds. Temporal Lobe Epilepsy.
Springfield, IL: Charles C Thomas; 1958:358–367
89. Walker AE, Lichtenstein S, Marshall C. A critical analysis of elec-
trocorticography in temporal lobe epilepsy. Arch Neurol.
1960;2:172–182.
90. Jasper HH, Arfel-Capdevielle G, Rasmussen T. Evaluation of EEG
and cortical electrographic studies for prognosis of seizures fol-
lowing surgical excision of epileptogenic lesions. Epilepsia.
1961;2:130–137.
91. McBride MC, Binnie CD, Janota I, et al. Predictive value of intra-
operative electrocorticograms. Electroencephalogr. Clin Neurol.
1991;30:526–532.
92. Stefan H, Hopfengärtner R, Kreiselmeyer G, et al. Interictal triple
ECoG characteristics of temporal lobe epilepsies: an intraoperative
ECoG analysis correlated with surgical outcome. Clin
Neurophysiol. 2008;119:642–652.
93. Engel J Jr, Bragin A, Staba R, et al. High-frequency oscillations:
what is normal and what is not? Epilepsia. 2009;50(4):598-604.
94. Jacobs J, LeVan P, Chander R, et al. Interictal high-frequency oscil-
lations (80–500 Hz) are an indicator of seizure onset areas inde-
pendent of spikes in the human epileptic brain. Epilepsia.
2008;49:1893–1907.
95. Quesney LF, Constain M, Fish DR, et al. Frontal lobe epilepsy—
field of recent emphasis. Am J EEG Technol. 1990;30:177–193.
96. Quesney LF, Constain M, Rasmussen T, et al. How large are frontal
lobe epileptogenic zones? EEG, ECoG and SEEG evidence. In:
Chauvel P, Delgado-Escueta AV, Halgren E, et al., eds. Advances in
Neurology, Vol. 57. New York: Raven Press; 1992
97. Salanova V, Andermann F, Olivier A, et al. Occipital lobe epilepsy:
electroclinical manifestations, electrocorticography, cortical stimu-
lation and outcome in 42 patients treated between 1930 and 1991.
Brain. 1992;115:1655–1680.
98. Salanova V, Andermann F, Rasmussen T, et al. Parietal lobe
epilepsy. Clinical manifestations and outcome in 82 patients
treated surgically between 1929 and 1988. Brain. 1995;118:
607–627.
724 Part V ■ Complementary and Special Techniques
99. Salanova V, Andermann F, Rasmussen T, et al. Tumoural parietal
lobe epilepsy. Clinical manifestations and outcome in 34 patients
treated between 1934 and 1988. Brain. 1995;118:1289–1304.
100. Wennberg R, Quesney LF, Villemure JG. Epileptiform and non-
epileptiform paroxysmal activity from isolated cortex after func-
tional hemispherectomy. Electroencephalogr Clin Neurophysiol.
1997;102:437–442
101. Quesney LF, Wennberg R, Olivier A, et al., eds. EcoG findings in
extratemporal epilepsy: the MNI experience. Electrocorticography:
Current Trends and Future Perspectives, (EEG Suppl. 48).
Amsterdam: Elsevier Science; 1998.
102. Wennberg R, Quesney LF, Olivier A, et al. Electrocorticography
and outcome in frontal lobe epilepsy. Electroencephalogr Clin
Neurophysiol. 1998;106:357–368.
103. Palmini A, Gambardella A, Andermann F, et al. Intrinsic epilep-
togenicity of human dysplastic cortex as suggested by corticogra-
phy and surgical results. Ann Neurol. 1995;37:476–487.
104. Jasper HH. Electrocorticography. In: Penfield W, Jasper HH, eds.
Epilepsy and the Functional Anatomy of the Human Brain. Boston,
MA: Little, Brown; 1954:692–738
105. Walker AE. Electrocorticography in epilepsy. A surgeon’s
appraisal. In: Fischgold H, ed. Electroencephalographie et
Electrocorticographie de l’epilepsie, Vol. 2. 1949:30–37.
106. Ajmone Marsan C. Focal electrical stimulation. In: Purpura DP,
Penry JK, Tower DB, et al., eds. Experimental Models of Epilepsy:
A Manual for the Laboratory Worker. New York: Raven Press;
1972:147–172
107. Bernier GP, Richer F, Giard N. Electrical stimulation of the
human brain in epilepsy. Epilepsia. 1990;31:513–520.
108. Cherlow DG, Dymond AM, Crandall PH, et al. Evoked response
and after-discharge thresholds to electrical stimulation in tempo-
ral lobe epileptics. Arch Neurol. 1977;34:527–531.

Principles and Techniques for Long-Term
EEG Recording (EMU, ICU, Ambulatory)
JEAN GOTMAN, MARC NUWER, AND RONALD G. EMERSON
INTRODUCTION
Most standard EEG recordings last 20 to 30 minutes and some
are extended to 1 or 2 hours. It is also possible to record the EEG
over long periods of time, from a few hours to days or weeks.
This is the domain of “long-term monitoring” (LTM) and this
chapter will discuss when such long recordings should be per-
formed and in particular the special recording and analysis tech-
niques that are available to facilitate the collection and review of
these data in the most efficient way, while providing the patient
with minimum discomfort. Most such LTM is performed in the
context of episodic disorders and the characterization of the
behavioral component of these disorders is an integral part of
the diagnosis process. Video recording of the patient, synchro-
nized with the EEG, is therefore an integral part of LTM.
Long-term EEG and video recording were born in the 1970s
and became a mature field in the early 1980s (1). The technol-
ogy has considerably improved since then but fundamental
aspects have not really changed, with the possible exception of
LTM in the intensive care unit (ICU), which has only developed
in the last few years. It is interesting to note that the combined
recording of EEG and behavior has a long history in the context
of epilepsy (2); at that time the “long-term” part was missing and
only the “EEG video” (albeit on film) was possible for a few min-
utes. The recording of behavior only became practically feasible
over extended periods in the 1970s with time-lapse video
recorders, recording one or just a few frames per second. Major
leaps occurred with the availability of VCRs and more recently
with the digital recording of video signals. With respect to the
ability to record EEG signals over the long term, the storage
capacity of computers has steadily increased, such that it
was possible to record discontinuous samples in the 1970s and
1980s and then it became possible to record continuously for
24 hours in the 1990s, at the standard sampling rate of 200 Hz.
Today the storage requirements for the long-term EEG signal
have become trivial, except in the context of large numbers of
channels (100 or more) and very high sampling rates (2000 Hz).
LONG-TERM MONITORING FOR EPILEPSY
Indications
The International League Against Epilepsy has published guide-
lines for the use of LTM in epilepsy (3). In order to make a solid
diagnosis of epilepsy and of the type of epilepsy, a description
of the behavioral and cognitive manifestations of the seizures is
necessary. This is most often obtained from the patients and
their relatives. In many cases, the description is incomplete or
CHAPTER
35
even erroneous: patients are often not aware of their seizures or
of parts of their symptoms and relatives are frequently poor
historians and poor observers. In some cases, it is important to
know the EEG manifestations of the seizures in order to make a
correct diagnosis. For example, behavioral arrest can occur with
primary generalized epilepsy and generalized spike and wave
and with a temporal lobe seizure. One can refine the indications
for LTM as follows (3):
• The diagnosis of epilepsy needs to be made or rejected, as the
events cannot be unambiguously classified from historical
information. Some patients have epileptic and nonepileptic
seizures.
• The diagnosis of epilepsy has been made but the type of
seizure or syndrome is not clear.
• Epilepsy is medically intractable and surgery is considered.
This requires a full characterization of the electroclinical
manifestations of seizures, at first with scalp EEG, and possi-
bly subsequently with intracranial electrodes in case the scalp
EEG does not provide sufficient information.
• Seizures need to be quantified to assess response to medical
treatment, as may be the case in brief but frequent seizures.
In addition, LTM allows recording throughout the
sleep–wake cycle. Some patients have predominantly or exclu-
sively seizures during sleep, others predominantly on awaken-
ing. Interictal activity in focal epilepsy is activated by non-REM
sleep and in some patients is only present during that stage.
Equipment and Personnel
The basic components of a LTM system are the electrodes and
electrode wires, the amplifier, the EEG transmission system, the
camera, and the recording computer. We will examine the char-
acteristics of these different components.
Electrodes
Electrode attachment is a particularly important aspect of
obtaining good-quality EEG during LTM and while the patient
experiences the many movements caused by the activities (e.g.,
eating) present during a recording that last several days and
while the patient has seizures, which can include violent move-
ments. Electrode problems are the most common cause of poor
quality of a recording. The scalp electrodes must be placed in
such a way that they provide a good contact for a long period,
typically at least 24 hours, sometimes longer if technicians are
not available every day or during weekends. Collodion or other
durable electrode-scalp adhesive is therefore recommended. It
is important to minimize artifacts caused by the movement of
725
726 Part V ■ Complementary and Special Techniques
electrode wires, which are particularly common during
seizures. This can be achieved by wrapping a bandage or plac-
ing elastic netting over the patient’s head, thus preventing the
wires from moving. This also limits the patient’s access to elec-
trodes (e.g., to prevent dislodging electrodes during scratch-
ing). All wires can be bundled together until the input of the
amplifier: the bundle has more rigidity than individual wires
and will also limit movement artifact.
If a patient undergoing LTM requires an MRI, the electrodes
must be removed and then replaced after the study. This can
require a considerable time. Electrodes that can be kept during
MRI studies have been developed recently. They include MR-
compatible material for the electrodes and the wires and an
MR-compatible connector system. Such electrodes include
gold, silver–silver chloride, subdermal silver wire, and conduc-
tive plastic electrodes (4,5). Some of these electrodes are also
CT-compatible, greatly facilitating EEG monitoring of patients
in the ICU.
It has been shown in multiple studies that electrodes record-
ing from regions lower than the standard 10/20 electrodes are
important in the characterization of temporal and inferior
frontal discharges. A minimum configuration for LTM should
therefore include 25 electrodes: the 19 electrodes of the 10/20
system plus three inferior electrodes on each side (e.g., F9, T9,
P9, F10, T10, P10 of the 10/10 system). Recordings are most
often referential and it is therefore particularly important to
have a reference of good quality. Display with any other refer-
ence can be subsequently calculated. The EKG must also be
recorded since cardiac changes can be an important component
of seizures or seizure-like events. It may be useful in some situ-
ations to use additional electrodes in order to obtain better spa-
tial sampling of a region. We will not discuss here the various
types of intracerebral electrodes as this is addressed elsewhere
in this volume.
EEG Amplification and Transmission
It is most convenient if the amplifier is sufficiently small to be
carried by the patient, usually in a pouch placed on the chest or
belt. After the amplifier, the transmitted signal is subject to
much less artifact because it has been amplified. The amplifier
should have minimal characteristics of a standard EEG ampli-
fier. These are typically a low filter at 0.3 Hz and high filter at 60
or 70 Hz and sampling at 200 or 250 Hz. There is recent evi-
dence that high frequencies in the intracerebral EEG (6–8; see
also Chapter 37) and low frequencies in newborns (9) may also
be of interest. Optimum characteristics may include low fre-
quencies (low filter at 0.1 Hz) and high frequencies (high filter
at 600 Hz and sampling at 2000 Hz). Digital filters can be used
when reviewing the EEG to eliminate unwanted low frequen-
cies if they have been recorded. Recording with a high sampling
rate has the disadvantage of increasing memory requirements
and slowing down display. Twelve-bit digitizing is common and
most likely to be sufficient, although 16-bit digitizing is also
available and covers more than sufficiently the dynamic range
of the intracerebral EEG.
Since the patient is monitored for extended periods, the cable
system connecting the patient-worn amplifier to the computer
or to a connector in the patient room must be light and comfort-
able, as well as very robust to withstand the stress of large
seizures. The functions of the cable are to transmit the EEG sig-
nal and to provide power to the amplifier. Recent technology
allows wireless transmission of the EEG, which gives much free-
dom of movement and added comfort to the patient. The
drawback of wireless transmission is that the equipment carried
by the patient must have a battery to power the amplifier as well
as the wireless transmission. This significantly increases the
weight of the patient-worn system and runs the risk of battery
depletion, thus interrupting monitoring. Wireless transmission
must therefore be considered for relatively short periods, with
cable transmission used most of the time.
Video Recording
Digital video cameras are most commonly used. The usual
color camera can be combined with an infrared illuminator and
an automatic switching to black and white for low light condi-
tions, thus providing an excellent quality image even in dark-
ness. Allowing darkness increases patient comfort at night. The
digital video signal is recorded continuously on the computer,
synchronized with the EEG. The EEG amplifier and the video
camera have independent clocks for sampling the EEG and for
timing video frames. This makes the synchronization of the two
signals a complex operation and it is important to verify that
this synchronization is maintained.
Recording Computer, Display, and Network
The EEG and the video signals are transmitted to the computer,
which may be at a distance from the patient room, either by
dedicated cable or more and more commonly through a stan-
dard computer network (either the hospital network or a net-
work dedicated to the LTM and EEG laboratories). Given the
storage capacity of modern computer disks, there is no problem
in storing several days of EEG recording, even sampled at high
rates, and several days of video recording. Large monitors allow
the easy display of 15 or 20 seconds of EEG on the screen.
Characteristics of the EEG display, gain, time scale, and filters,
can easily be changed. Since EEGs are usually recorded referen-
tially, alternate montages can be viewed (bipolar, changed refer-
ence, average reference, etc.). Network facilities can be used to
view the EEG from any computer connected to the computer
on which the EEG is stored, either during the recording or after
it is completed. This includes viewing from locations outside
the hospital, such as at home, provided security barriers are
respected.
Archiving the data is a challenge because it is not practical to
archive all recorded data on permanent storage media such as
DVDs, particularly the video, because of size. It is likely that
complete archiving will become possible in a few years, but at
present, the data that are to be archived must be selected by a
person familiar with EEG interpretation to make sure all
important information is retained.
Personnel and EEG Interpretation
Trained observers need to be near the patients to provide close
clinical observation and interaction with the patient during
 Chapter 35 ■ Principles and Techniques for Long-Term EEG Recording (EMU, ICU, Ambulatory)
seizures, as well for ensuring patient safety. To assist with con-
stantly observing patients on monitoring, patients can press an
alarm button to report any auras or seizures they notice.
Observation for seizures may be enhanced with automatic
seizure detection systems (see below). It is also useful if a rela-
tive or companion can stay in the room with the patient and
alert staff when seizures occur.
Experienced EEG technologists should be used so as to
obtain the best information from an LTM session. They are crit-
ical to secure electrodes well and in the proper locations. They
review the preliminary recorded data and seek to achieve suit-
able recording quality. In some centers they make a first review
of the 24-hour session, marking the seizures and any section of
interest. Only the marked sections are reviewed subsequently by
the clinical neurophysiology physician. In other centers, the full
recording is reviewed by the physician, but this is often consid-
ered too time consuming. Automatic seizure and spike detection
may be helpful in selecting the sections to review (see below). In
addition to selecting what data need review, it is necessary to
decide what subset of the original recording needs to be perma-
nently archived. (The full original recording is usually too large
for storage, particularly the video.) The selection of what needs
to be retained results from review and interpretation of the
record. It typically consists of all seizures and some interictal
spikes, as well as some sample of the EEG background.
Duration of Monitoring and Drug Withdrawal
The duration of monitoring depends on the purpose. In gen-
eral, it is best to try to be generous with the duration of moni-
toring because it is frequent that unexpected events are
discovered. Patients may report one seizure type but not be
aware that they also have brief seizures of which they are not
aware or nocturnal seizures nobody has witnessed. Some
patients have more than one type of seizure, so a sample of only
two to four seizures may be too few to use for major clinical
decisions such as surgery. The clinical description reported by
the patient or family is often incomplete or erroneous and it
may require the recording of several seizures to fully document
a seizure disorder that is different from what was expected. In
the context of presurgical evaluation, it has been estimated that
five seizures with the same region of onset and no contradictory
information provide a reasonable sample to decide that the
onset is indeed in that region (10).
Antiepileptic medications are frequently reduced or discon-
tinued during LTM in order to allow seizures to occur more fre-
quently. In the context of presurgical evaluation, the question of
whether seizures occurring after medication withdrawal are
similar to habitual seizures has been raised. Symptomatology
and the EEG pattern at onset are similar after medication with-
drawal (11–14). There is no evidence that seizures spread more
rapidly from one hemisphere to the other after drug with-
drawal. Seizures are more often secondarily generalized after
drug reduction (15,16), so monitoring unit staff should be pre-
pared to respond in case of repeated generalized seizures.
It is also often said that the reduction in antiepileptic med-
ication results in increased spiking activity. Careful studies of
the timing of drug withdrawal, seizure occurrence, and changes
727
in spiking (17,18) have shown that spiking does not increase
following drug withdrawal per se. Spiking does increase follow-
ing seizures. When drugs are reduced, seizures occur and spik-
ing increases as a result of increased seizures. It may therefore
appear that drug reduction results in increased spiking, but this
is an indirect effect mediated by seizures. In the absence of
seizures, spiking does not change. These observations were con-
firmed in a recent study using more sophisticated analysis tools
(19). This study also demonstrated that spiking could even
decrease as a result of drug withdrawal in the absence of
seizures.
Considerations for Children
or Infants and for Adults
LTM can be performed at all ages including during the new-
born period (20). In newborns and young infants, a subset of
the standard 10/20 electrodes can be used although one has to
be careful because seizures can be very focal. Collodion should
not be used in young children, as their skin is very delicate. As
seizures and interictal activity are more frequent in young
patients than in adults, the duration of monitoring is usually
shorter, often limited to 1 or 2 days. In children, a full syn-
dromic definition often requires that different seizure patterns
be recorded, as well as the interictal activity (3). In adults but
more particularly for children, it is important to have a parent
in the room during monitoring to help identify the events being
investigated.
Automatic Seizure and Spike Detection
The size of computer disks is such that it is simple to perform a
continuous 32-channel, 7-day recording on a standard com-
puter, storing EEG and video. A computer-based recording
allows a flexible review procedure, with random access to any
part of the recording and the availability of numerous methods
of data manipulation and analysis. In most 24-hour recordings,
95% to 99% of the recording offers little useful information in
the evaluation of an epileptic disorder. Automatic detection
methods, despite their imperfections, can facilitate the marking
and thus the review of valuable information.
Automatic Seizure Detection
During LTM, patients are usually asked to press a button when
they feel a seizure coming. If patients could always feel their
seizure coming, there would be no need for automatic seizure
detection. It is frequent, however, that patients are not aware of
their seizures, and observers are not always present or able to
recognize a seizure. Furthermore, some seizures have very min-
imal clinical signs—and sometimes no visible signs at all.
Finding all seizures may require the review of days and days of
EEG, a process that is long and tedious—and not always effec-
tive in as much as short seizures can be missed, particularly if
there are many channels as in intracerebral recordings.
Automatic seizure detection can be helpful in this context, since
it provides a means for marking the EEG sections that are likely
to include seizure patterns. For this purpose, the detection can
take place at any time during the seizure (not necessarily early)
because the EEG will be reviewed a posteriori.
728 Part V ■ Complementary and Special Techniques
Seizure detection could also be used in real time, to warn the
patient or observer that a seizure has just started. In this case,
the detection must occur early to allow the patient to take pro-
tective measures (if he or she is still able to do so), to assist an
observer in providing protective measures in a timely fashion,
and to facilitate the observer’s attempts to question the patient
so as to improve our understanding of clinical signs (e.g., can
the patient follow commands, talk, remember?). Such an early
seizure detection system can be used during monitoring in the
clinic (21,22), but could also be incorporated in an implantable
device that would provide a warning (e.g., auditory signal) to
the patient during normal everyday activities.
Seizure prediction is an area of active investigation but,
unlike seizure detection and the warning provided by the onset
of the seizure, it is an area that has proven difficult and without
clear results. Despite initially promising results, it is not yet
clear if it is possible to predict seizures (23; see also Chapter 30).
Detection Methods
Our current understanding is that seizures consist of paroxys-
mal rhythmic activity generated by a large number of synchro-
nized neurons. This activity usually evolves in its spatial
distribution, frequency, and amplitude. The patterns that are
actually recorded depend not only on the discharge itself, but
Cz-C3
C3-T3
T3-Zy1
Zy1-Zy2
Zy2-T4
T4-C4
C4-Cz
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
150 uV
1s
Figure 35.1 Scalp-recorded right temporal seizure occurring in the context of muscle and eye movement artifacts. This
early part of the seizure consists of a discharge in the theta range of frequencies, commonly seen in temporal lobe seizures.
also on the respective position of the electrodes and the gener-
ator. In some cases, no seizure discharge may be visible because
the electrodes are too far from the generator. Similarly, the
seizure discharge may appear only as a mild modification to the
ongoing EEG. It is therefore important to distinguish the
seizure itself from its manifestations (i.e., from the vantage
point of the recording electrodes). Since seizure detection is
performed from analyzing the EEG, it is only the seizures that
have a clear manifestation in the EEG that can be detected.
Figures 35.1 and 35.2 show examples of seizures recorded from
the scalp, and Figure 35.3 shows a seizure recorded from intrac-
erebral electrodes. In Figures 35.1 and 35.3, the specific
moment of onset is not easily identified. In Figure 35.2, the
onset is clear, consisting of a few seconds of rhythmic discharge
followed by irregular slow waves.
There is no formal definition of what a seizure consist of in
the EEG. Qualitative definitions usually include the terms
“paroxysmal, rhythmic, evolving”— but an unambiguous defini-
tion has not been formulated. Automatic detection methods have
therefore relied on trying to encode these qualitative characteris-
tics into algorithms, and improving the algorithms by trial and
error in an effort to increase detection sensitivity and decrease
the false detection rate. The first general-purpose seizure detec-
tion method validated on a large data sample, characterized each
---
---
---
---
---
---
---
---
---
---
---
---
---
---
---
 Chapter 35 ■ Principles and Techniques for Long-Term EEG Recording (EMU, ICU, Ambulatory)
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
100 uV
1s
Figure 35.2 Scalp-recorded seizure with a generalized and abrupt onset. A few seconds of rhythmic theta frequency
activity are followed by irregular slow waves in the delta range.
2-second epoch by measures of frequency, rhythmicity, and
amplitude change, and compared these measures to a constantly
updated background (24,25). A set of empirical rules combined
the findings in multiple epochs and multiple channels to reach a
decision on detection. The concept of a “constantly updated
background” plays an important role in the analysis of long-term
data, since EEG is nonstationary, that is, the background content
changes considerably due to factors such as states of vigilance. It
is also important that the results not be dependent on the selec-
tion of a particular background epoch, a dependence that would
introduce arbitrariness into the process. More recent methods
that have followed a similar empirical approach include those of
Grewal and Gotman (26), and Saab and Gotman (27).
Artificial neural networks (ANNs) are an interesting
approach to seizure detection (28–30). The ANN is presented
with a large number of training examples that comprise both
seizure and nonseizure data, and gradually learns to differenti-
ate seizures from nonseizure EEG. A set of features characteriz-
ing the EEG needs to be specified in advance (e.g., amplitude,
rhythmicity, dominant frequency, etc.), but it is not necessary to
explicitly describe seizure patterns. The ANN automatically
determines which combinations of features are characteristic of
seizures by adapting its internal structure to reflect the training
data. The performance of ANN methods depends on an appro-
priate set of features and on exposure to a sufficiently large
training set containing a variety of seizure patterns.
729
---
---
---
---
---
---
---
---
---
---
---
---
---
---
---
---
An extension to the above concept of context is to incorpo-
rate a particular patient’s seizure characteristics in the context.
In this case, the method is aimed at detecting seizures in that
particular patient, and only seizures that look like the sample
seizures (that have been incorporated in the context) are
detected. The method can then be labeled “patient-specific”—
or more correctly, “seizure-specific”—since a patient can have
several types of seizure. Better detection performance can thus
be obtained if it is acceptable to detect only a specific seizure
pattern in a given patient (21,22,31,32).
Seizure patterns are quite different in the newborn EEG than
in older children and adults: the discharges are often much
slower and sometimes very focal (limited to one electrode). For
these reasons, methods specific to the newborn have been
developed, with the particular aim of detecting the very slow
discharges often present during newborn seizures (33–35).
Validation of Detection Methods
A method must be tested on a data set independent from that
on which it was developed. The testing data must be represen-
tative of the context in which the method will be used: they
should include a large variety of seizures, and a much larger
amount of nonseizure data than of seizure data (in order to
ensure that all types of interictal data are well represented—
quiet and active wakefulness, different sleep stages, different
types of artifacts common during LTM). The best way to avoid
730 Part V ■ Complementary and Special Techniques

RH2-RH3
---

RH4-RH5
---

RH5-RH6
---

RH6-RH7
---

RH7-RH8
---

RH8-RH9
---

RH9-RH10
---

RE1-RE2 ---

RE2-RE3 ---

RE3-RE4 ---

RE5-RE6 ---

RE6-RE7 ---

500 uV

5s

Figure 35.3 Seizure recorded from intracerebral electrodes placed on the temporal lobe. The RH contacts are along an
orthogonal multicontact depth electrode having its deepest contacts (lower numbers) in the hippocampus and most superfi-
cial contacts in the second temporal convolution. The RE electrodes are epidural electrodes placed along the first temporal
convolution. The seizure onset consists of mixed frequency activity, followed by a more rhythmic discharge. Note compressed
time scale. For the purpose of clarity, this EEG illustrates only a fraction of the electrodes implanted in this patient’s brain.

selection bias is probably to select all data between two given
dates. An important issue in validating a method is the defini-
tion of a seizure itself. Whereas there is no ambiguity with a
clear seizure lasting 30 seconds or 1 minute, significant ques-
tions arise for short events: What is the minimum duration of a
seizure? Should a seizure having clinical manifestations, but no
change in the EEG, be counted as an event to detect? What if the
EEG changes are very minor, such as a brief flattening or a short
burst of slow waves? When presenting validation results, these
issues should be addressed.
Methods that were evaluated on extensive data sets (hun-
dreds of hours, tens of patients) report detection sensitivities
from 75% to 90%. Sensitivity is often lower in newborns than
in adults, as some newborn seizure patterns are very subtle. The
false detection rate, the most common measure of failure, is in
the range of 0.3 to 1 per hour in recent studies. Few studies
report results with intracerebral electrodes, but these recordings
have important distinguishing characteristics: scalp artifacts
(electromyogram [EMG], movement, eye blinks) are absent,
thus reducing false detections; the dynamic range of sponta-
neous fluctuations in background is much larger in intracere-
bral than in scalp EEG—thus resulting in more false detections;
the intracerebral EEG often includes many paroxysmal bursts of
uncertain significance, which can also contribute to false detec-
tions; and seizure patterns are often more prominent in intrac-
erebral EEG, thus resulting in better sensitivity (26).
Automatic Spike Detection
Many epileptic patients present with interictal abnormalities in
their EEG. These consist of abnormalities of the background,
which will not be discussed here, and paroxysmal events such as
spikes, polyspikes, sharp waves, and bursts of spike and wave.
These paroxysmal events occur most often unpredictably and
sometimes rarely. They play an important role in defining some
epileptic syndromes and they contribute to localizing the
source of seizure discharges. Although used primarily for these
purposes, it is important to remember that they, by themselves,
disturb normal brain function (36,37). Automatic spike detec-
tion methods started to be developed when computer analysis
of the EEG was in its infancy, in the 1960s. Despite this long his-
tory and despite the tremendous increase in the power of com-
puters over this period, automatic detection today is far from
perfect. There are probably two reasons for this: the first is that
spikes are not a well-defined pattern, as illustrated by the low
 Chapter 35 ■ Principles and Techniques for Long-Term EEG Recording (EMU, ICU, Ambulatory)
inter-rater agreement demonstrated in the study of Wilson et
al. (38); it is an almost impossible task to detect a pattern for
which there is so little agreement. The second reason may be
that, despite its apparent simplicity, spike detection is quite a
difficult task that the human brain performs rapidly but that
takes into account a complex analysis of temporal and spatial
features as well as context, an analysis that is not easy to repro-
duce in a computer program.
During routine EEG examinations or short recordings, auto-
matic spike detection is of little use for an expert because it is
easy to find all the spikes visually. It is sometimes said, however,
that nonexperts overinterpret the EEG, defining as epileptiform
transients that are artifacts or normal variants (39). This can
result in overtreatment of patients. In such a context a reliable
automatic detection method would be useful. Generally, auto-
matic detection becomes valuable mostly when analyzing long
recordings, particularly during long-term epilepsy monitoring,
when hours or days of recording have to be examined. It is also
in this context that automatic detection is most difficult
because the recording includes multiple transients that are dif-
ficult to differentiate from epileptiform transients: eye blinks,
movement and muscle artifacts, electrode artifacts, sleep tran-
sients (vertex waves, spindles, K complexes).
An excellent review of spike detection methods for scalp
EEG has recently been published by Halford (40) and we refer
the reader to this comprehensive review. No spike detection
method has addressed specifically the problem of detecting
spikes in the intracerebral EEG. Although in appearance it is an
easier problem than detection in the scalp EEG because of the
absence of most of the artifacts, it remains a difficult problem
because the issue mentioned above regarding the definition of
spikes is worse in the intracerebral EEG. Some channels show
almost continuous sharp activity and it is very difficult to iden-
tify isolated transients. The morphology of sharp transients is
also highly variable. In other words, the problem is even more
ill-defined than for scalp spikes.
LONG-TERM ICU MONITORING
ICU EEG monitoring is used for several clinical purposes. In
coma EEG can search for possible presence of seizures, some of
which are nonconvulsive. For patients at risk for deterioration,
EEG can monitor continuously for adverse signs. For patients
in a deliberate burst suppression state, EEG can monitor that
the induced coma is sufficiently deep.
A primary goal of ICU EEG monitoring is identification of
acute adverse events early enough to intervene to prevent or ame-
liorate long-term adverse clinical sequellae. Examples of acute
events in neurologic critical care patients include subarachnoid
hemorrhage (SAH), intracranial bleeding, vascular thrombosis
or embolus, convulsive or nonconvulsive seizures, cerebral
edema, or vasospasm. Some are beyond meaningful intervention.
Many are amenable to various medical or surgical therapies.
Typical Indications for ICU EEG Monitoring
Monitoring can identify subtle or nonconvulsive seizures.
Seizures can manifest in various ways and circumstances. EEG
731
can determine whether a critically ill patient is having seizures,
including nonconvulsive seizures, that cause a decreased level of
consciousness (41,42). Nonconvulsive seizure EEG patterns
may be frequent to constant epileptic spiking, or more discrete
cycles of ictal discharges separated by interictal attenuation and
slowing. Spike morphology can be the traditional single spike
or sharp wave events, or more organized trains of spike wave
seen in focal or generalized epilepsy (43–52). EEG monitoring
of critically ill patients is indicated for patients with decreased
level of consciousness not otherwise explained.
ICU EEG monitoring can separate movements from seizures.
For example, pseudoperiodic lateralized epileptiform discharges
(PLEDs) can be seen at the site of focal lesions, such as ischemic
stroke (53,54). They may be accompanied by myoclonic jerks,
usually seen as temporally separate rather than synchronous
with the PLED discharges. EEG recordings with EMG channels
accompanying the myoclonic jerks can demonstrate the lack of
temporal relationship of jerks to EEG discharges, evidence that
observed movements are myoclonic rather than epileptic. Some
medications, for example, tacrolimus, are associated with tics or
other repetitive movements for which EEG with an EMG chan-
nel or video-EEG ICU monitoring can help determine whether
movements are epileptic.
Monitoring can grade the degree of encephalopathy. Slowing
in routine EEGs can be either continuous or intermittent. A
fixed structural deficit or anoxic injury tends to produce con-
tinuous slowing. A metabolic encephalopathy tends to produce
waxing and waning of background, that is, intermittent slow-
ing. For both the continuous and intermittent slowing, the
degree of EEG slow activity corresponds in a general way with
the degree of impairment. For example, alpha coma from
hypoxic injury is characterized by monotonously continuous
frontocentral alpha activity (55–61) that is nonreactive, that is,
does not change when the patient is stimulated. EEG monitor-
ing can evaluate the EEG not only for reactivity to stimulation
and moment to moment monotony, but also for the sponta-
neous variability that occurs more normally over many minutes
to hours. Presence of such variability is a favorable sign, and
absence is unfavorable. Shorter routine EEGs may miss these
longer duration fluctuations in EEG content because the sam-
ple is too short. ICU EEG monitoring can identify and measure
the variability in long-term EEGs.
Burst suppression carries a grave prognosis when associated
with anoxia or diffuse brain injury. It also is the desired state
when a patient is in a deliberately heavily sedated state. EEG
monitoring can monitor the extent of burst suppression. That
can be used as a gauge of prognosis in hypoxic–ischemic burst
suppression. Or, it can be a feedback measurement for adequacy
of therapy in deliberate burst suppression. ICU EEG monitor-
ing can measure and trend the level of burst suppression.
Incidence and Impact of Seizures and Status Epilepticus
Seizures occurred in 29% to 35% of ICU patients who had
stroke, intracerebral hemorrhage, seizures, metabolic coma,
brain tumors, and brain trauma as reasons for hospitalization
(41,62,63). Table 35-1 gives additional rates of ICU seizures. Two
thirds to three fourths of patients had nonconvulsive seizures or
732 Part V ■ Complementary and Special Techniques

Tabl e 3 5 . 1

The Incidence of Seizures Using ICU EEG Monitoring in a Neurologic Critical Care Illness Population

Percentage with
ICU EEG Studies N Seizures Principal Diagnosis
Jordan (62) 124 35 Neurologic critical illness
Vespa et al. (64) 91 22 Brain trauma
Vespa et al. (144) 300 21 Subarachnoid
hemorrhage/brain trauma
Vespa et al. (65) 65 28 Primary intracerebral hemorrhage
Claassen et al. (66) 204 17 Subarachnoid hemorrhage
Pandian et al. (67) 105 42 Neurologic critical illness
Young and Doig (145) 55 20 Neurologic critical illness

nonconvulsive status epilepticus. Such seizures are not readily seizures (78). The increase in glutamate concentration exceeds
diagnosed by motor manifestations. The diagnosis required EEG the concentration known to induce cell death and swelling, and
monitoring. Clinical manifestations of the seizures included persists for days. This excitotoxic secondary injury in turn
persistent stupor and coma. Many EEG patterns were typical
ictal events, whereas others had various kinds of continuous
repetitive sharp waves, polyspikes, or sharply contoured rhyth-
mic slowing. The variety of EEG presentations in critically ill
patients differs from the presentation of ictal events in an
epilepsy monitoring unit. Among these patients, 21% were
initially misdiagnosed with nonepileptic conditions. EEG mon-
itoring sometimes was not initiated at the beginning of the ICU
stay. An average of 43 hours passed before EEG monitoring was
used, the correct diagnosis made, and appropriate therapy
undertaken. Other reports are similar (63,68–70).
Changes in patient care commonly included initiating or
modifying anticonvulsants, obtaining neuroimaging, or adjust-
ing blood pressure. EEG monitoring was decisive in initiating
or prompting changes in patient care in 51% of patients and
made a significant contribution in 31%. Several studies assessed
the incidence of nonconvulsive seizures and status in ICU
patients (66,71–73). The incidence ranged from 11% (74) to
55% (66). After moderate to severe traumatic brain injury, 22%
of 91 patients had seizures, a majority (57%) of which were Figure 35.4 EEG detection of four nonconvulsive seizures. EEG mon-
nonconvulsive (Fig. 35.4) (64). The incidence of overt clinical itoring over 5 hours in a 50-year-old hypertensive man 3 days after a
seizures after an ischemia stroke was 5% to 17% (75–77), 3-cm left thalamic hemorrhagic stroke. There were no outward signs of
whereas the incidence of subtle or electrographic nonconvul- seizures on examination of this comatose patient. EEG monitoring
sive seizures is twice as high (62). After intracerebral hemor- detected three nonconvulsive seizures during this 5-hour period.
rhage the incidence of seizures was 28% (65), many of which Monitoring trends from six EEG channels are shown. Time is displayed
were nonconvulsive. Compared to studies of overt clinical along the horizontal axis, from 11 AM to 4 PM. For each channel, the
seizures, ICU EEG monitoring detects more than twice as many amount of alpha activity is displayed as vertical bars. Alpha activity is
seizures, with the increase being mainly nonconvulsive seizures. expressed as a percentage of the overall EEG activity from 1 to 30 Hz.
Some convulsive seizures also were found that had been over- The percent-alpha values range here from approximately 150% to 50% .
looked by nursing and medical staff. Seizures are suspected when an abrupt jump in percent-alpha is seen,
followed by a suppression of alpha activity. The three suspected events
Seizures in the ICU Damage the Brain are noted here with arrows. Review of the stored polygraph EEG
Seizures after brain injury or hemorrhage provoke a pathophys- showed that these were generalized epileptic EEG discharges. These
iologic response at a time when the brain is very vulnerable. repetitive nonconvulsive seizures were then treated with phenytoin.
Levels of extracellular glutamate increase after posttraumatic (Figure courtesy of UCLA EEG Lab.)
 Chapter 35 ■ Principles and Techniques for Long-Term EEG Recording (EMU, ICU, Ambulatory)
causes lipid peroxidation, membrane disruption, and increases
in extracellular lipid breakdown product (e.g., glycerol) levels
(79). Nonconvulsive seizures in critically ill patients increase
glucose metabolism and intracranial pressure. These mecha-
nisms most likely underlie the observation that nonconvulsive
seizures after brain hemorrhage lead to increased cerebral
edema and midline shift (65). Compared with brain hemor-
rhage patients without seizures, those with seizures had a sub-
stantial increase in midline shift and increased mortality.
Seizures were an independent factor predicting mortality. These
lines of evidence show that nonconvulsive seizures should not
be considered benign in critically ill patients, and that they
should be detected and treated.
Variability, Reactivity, and Sleep Cycles During Coma
Certain EEG features are associated with better prognoses. In
the comatose or deeply encephalopathic patient, these favorable
features include reactivity, variability, and sleep cycles.
Unfavorable findings are monotonous slow activity without
reactivity or variability (80–82). Reactivity is seen over short
times in response to verbal, visual, or tactile stimulation.
Variability is a spontaneous change in the EEG seen over longer
periods. Sleep cycles are one form of variability. EEG monitor-
ing is the only way to detect these changes that occur over long
periods. Sleep–wake cycles are particularly favorable. The find-
ing of a less regulated variability, cycling, or alternating EEG
pattern also carries a favorable prognosis. In postanoxic and
traumatic coma, 95% of patients had unfavorable outcomes
when their EEG showed a slow monotonous pattern, whereas
only 30% of patients had unfavorable neurologic outcomes
when their EEG showed variable or sleep–wake cycle EEG pat-
terns (83). In a further 500 patients with clinical evidence of
upper brainstem impairment, the mortality rate was 93%
among patients whose EEG showed a slow monotonous pat-
tern, 46% among those with a variable pattern, and 28% among
those patients with a sleep cycle pattern (84). Among patients
with a cortical or subcortical injury from head trauma, the
mortality rates were 63% for a slow monotonous EEG pattern,
40% for a variable pattern, and 9% for a sleep cycle pattern
(84). EEG monitoring can detect and measure these different
patterns, but a routine EEG in the ICU may miss the long-term
changes of variability and sleep cycles. These EEG changes are
not noticeable on physical examination.
Cerebral Ischemia
EEG is altered by cerebral blood flow (CBF) below 18 to
25 mL/100 g/min, whereas ischemic injury occurs at lower CBF
levels of 12 to 18 mL/100 g/min (85–87). EEG changes begin
with loss of fast activity and progress at lower CBF to polymor-
phic delta and eventually to isoelectric (88). EEG may demon-
strate brain function recovery before the clinical exam (89).
These observations make the EEG a useful tool to monitor
brain perfusion.
Acute ischemic stroke is best visualized on diffusion- and
perfusion-weighted MRI. Those techniques are not well suited
for continuous ICU monitoring. EEG can serve a monitoring
role, screening for changes suspicious for new acute ischemic
733
stroke. Often the patient is sent for an MRI or CT to evaluate
the acute change, since EEG is not specific enough to localize or
confirm the cause of a change. The amount of EEG slowing
parallels perfusion-weighted MRI lesion volume (90). EEG also
correlates well with clinical control of hypertension and
hypotension. Six of 11 patients showed EEG improvement dur-
ing hypertension, and 12 of 18 patients showed EEG worsening
during periods of hypotension (91). An acute EEG index
reflected the therapeutic effects of tPA (90,92). The progression
of adverse change to isoelectric can be regional rather than gen-
eralized, a phenomenon labeled by Schneider and Jordan (93)
as regional attenuation without delta (RAWOD).
Vasospasm After Subarachnoid Hemorrhage
SAH causes EEG slowing (94) and disrupts the posterior dom-
inant rhythm (95). Among 116 consecutive SAH patients with
EEG monitoring (96), 16% showed periodic epileptiform dis-
charges, 6% nonconvulsive seizures, and 4% nonconvulsive sta-
tus epilepticus. The EEG was nonreactive to external stimuli in
14%, showed no state changes in 14%, and stage II sleep tran-
sients were absent in 85%. Other EEG findings included frontal
rhythmic delta activity (FRDA; 8%) and stimulus-induced
rhythmic, periodic, or ictal discharges (SIRPIDs; 8%) (96,97).
Clinical deterioration or delayed cerebral ischemia after SAH is
a major complication. Angiographic vasospasm is detected in
50% to 70% of SAH patients (98) and delayed cerebral ischemia
in 19% to 46% (99–104) and up to 54% with worse SAH grades
(105). Because treatments for vasospasm may prevent infarc-
tion if started early, timely warning is the key. Angiography and
transcranial Doppler are not monitoring techniques, whereas
EEG monitoring is continuous. EEG also may help guide effec-
tiveness of therapy. Several EEG changes herald the onset of
vasospasm: total power trends (106), relative alpha variability
(107), and poststimulation alpha/delta ratio (66). EEG in one
study was 100% sensitive for angiographically defined
vasospasm (107). Focal ischemia from vasospasm may produce
bilateral EEG changes (66,107). EEG correctly predicted a
vasospasm-free course among 73% of 151 SAH patients.
Neonatal Applications
The simple amplitude integrated trending technique has
received attention in the neonatal ICU setting (108). The tech-
nique has been applied in recent years as a method for identify-
ing or monitoring a variety of critical care neonatal conditions.
A greater number of total power bursts per hour were favorable
for preterm infants with intraventricular hemorrhage
(109–111). Features could differentiate preterm infants with
normal neurologic status from those with brain injury
(112–116). Criteria included voltage, continuity, variability,
bursts, and sleep cycles.
Limitations arise in the use of power trends without EEG
tracings. There is considerable expertise and literature built over
decades on the implications of spikes, sharp waves, symmetry,
and content of premature neonatal EEG. That literature and
clinical application depends on interpretation of EEG tracings
by knowledgeable electroencephalographers. The use of trends
by themselves lacks the detail to allow clinical application of that
734 Part V ■ Complementary and Special Techniques
extensive knowledge. As a result, the field of neonatal ICU EEG
finds itself where adult EEG was several decades ago when com-
pressed spectral array (CSA) (117,118) was the commonly used
technique for adult ICU EEG monitoring.
Methods
The principal goal is rapid identification of new complications
in time to intervene, thereby lessening or avoiding long-term
neurologic complications. ICU monitoring must be automated
because nurses cannot devote much time to EEG problem solv-
ing. Monitoring should identify quickly abrupt changes as well
as gradual changes.
Methods for ICU EEG monitoring are borrowed from other
EEG applications. Continuous recording is taken from epilepsy
monitoring. Observing for complications, deterioration, or
improvement is taken from intraoperative monitoring.
Trending of features is taken from polysomnography and ear-
lier generation CSA techniques.
Prior to 1985, ICU EEG monitoring favored two- or four-
channel CSAs (117,118). CSA monitors were too difficult to
understand and use. In 1985, modern generation EEG monitor-
ing was introduced using innovations that went beyond CSA
(119). Those innovations include recording from large number
of EEG channels and storage of the EEG for review. Simpler
trend displays are easier to understand. Remote access allows
for expert review as needed. Both frequency trends and EEG
tracings are continuously displayed at the bedside. Automated
seizure detection is borrowed from epilepsy monitoring tools.
Nurses are taught simple rules for calling attention to EEG
changes. More recently, video-EEG monitoring has been added
for seizures (67).
ICU EEG requires a trained electroencephalographer to
interpret and to review changes frequently. Nurses can be
taught the general principles of when changes need attention,
but it is difficult to teach most ICU nurses sufficient skills to
understand many findings. For that reason, a clinical neuro-
physiology physician’s involvement usually is needed several
times daily and on request to interpret the EEG when changes
occur. There are two clinical paradigms for this involvement. In
one, a critical care neurologist is trained as an electroencephalo-
grapher and includes the EEG among the tests monitored
throughout the ICU stay. In the second paradigm, an electroen-
cephalographer interprets the EEG repeatedly during each day
and is available for further reviews as needed, and conveys the
findings to the patient’s critical care physician.
Staff and physician availability are needed to connect
patients, to maintain and interpret recordings. Frequency
analysis trending can improve identification of change and
reduce reading time but trends should never be interpreted in
isolation. The original EEG tracings should be the basis for
definitive interpretation, which requires training in electroen-
cephalography. A 24-hour coverage by an electroencephalog-
rapher in neurologic ICUs is impractical, so trained ICU staff
or specially trained technicians akin to cardiac telemetry tech-
nicians could screen the trends and tracings and refer ques-
tionable events for definitive interpretation remotely by a
trained electroencephalographer. Remote access to real-time
EEG is feasible (120), as is standard for intraoperative moni-
toring. Since hook ups are needed at any time 24 hours daily,
a process for applying electrodes and initiating monitoring
requires either a knowledgeable technologist or an allied
health professional. Such health professionals, such as ICU
nurses, might use preconfigured head nets or caps to locate
electrode sites.
Some have suggested that single-channel frequency analysis
systems, such as the BIS monitor (Aspect Medical, Newton,
MA) or the SNAP monitor (Everest Biomedical, Chesterfield,
MO), can be used to monitor patients with status epilepticus.
However, these limited quantitative methods have great diffi-
culty monitoring seizure activity because changes are not spe-
cific enough. These limitations are similar to the drawbacks that
led to replacement of CSA by modern generation ICU EEG
technology.
Conclusion
ICU EEG monitoring is a method for detecting acute problems
in critical care patients in time to intervene and prevent serious
long-term sequellae. It allows for identification of complica-
tions so as to reduce ICU stay and improve clinical outcome.
Beginning in 1985 the modern generation ICU EEG monitor-
ing technique has used large numbers of EEG channels, remote
access to read the tracings, saving of entire EEG files for review,
and simplified trending for ease of nurses and nonspecialists.
Trending also helps to identify gradual changes over hours and
days. These and other tools have moved the field well beyond
the initial CSA techniques of the 1970s.
Now EEG ICU monitoring is used to identify seizures and
status epilepticus, focal impairment, and measure gradual
improvement or deterioration. Those changes can suggest the
onset of such important clinical complications as hemorrhage
or vasospasm. Monitoring helps to show the adequacy of treat-
ment with blood pressure control to achieve adequate cerebral
perfusion, or medications to induce deliberate burst suppres-
sion. The tools can help with prognosis. The field is better
defined and advanced in the adult ICU. In the neonatal ICU,
monitoring tools still try to apply trending alone.
In much of the same way that electrocardiographic monitor-
ing is a required element of caring for the myocardial ischemia
patient, ICU EEG monitoring is becoming the standard of care
required for neurologic critical patient care and others with
coma of uncertain cause.
AMBULATORY MONITORING
Ambulatory EEG (AEEG) fills a useful niche between routine
laboratory-based EEGs and inpatient video-EEG monitoring.
AEEG provides the ability to extend the duration of EEG sur-
veillance from the 30 minutes typical of a routine EEG to sev-
eral days without requiring hospitalization. In the past, AEEG
devices offered limited fidelity and flexibility, and could
record relatively few channels. Those limitations no longer
apply, and current AEEG systems can produce recordings
equivalent in quality to recordings obtained using standard
laboratory systems.
 Chapter 35 ■ Principles and Techniques for Long-Term EEG Recording (EMU, ICU, Ambulatory)
Clinical Utility of Ambulatory EEG
The ability to record EEG over extended periods increases the
likelihood of capturing interictal discharges and seizures. This
information is often critical for accurate diagnosis and clini-
cal management. In one study, the yield of AEEG monitoring
for providing evidence to support the diagnosis of epilepsy
was 2.5 times that of a standard EEG, and 83% that of inpa-
tient video-EEG monitoring (121). Importantly, the chance of
capturing seizures is far greater on AEEG than standard EEG;
in a study comparing sleep-deprived EEGs and AEEGs in
patients with normal or nondiagnostic standard EEGs, sleep-
deprived EEGs captured no seizures while seizures were
recorded in 15% of patients on AEEG, including in several
patients who were unaware of them (122). Underreporting of
seizures is well recognized, in part because amnesia for the
ictus is common; in one study, patients were unaware of 63%
of all seizures occurring in an inpatient epilepsy monitoring
unit (123).
While the ability to record EEG as the patient goes about
normal activity is an advantage, daytime recordings are likely
to contain substantially more movement- and muscle-related
artifacts than recordings made in an EEG laboratory.
Nighttime recordings are more likely to be artifact free, and
indeed the ability to obtain prolonged sleep recording is an
attractive feature of AEEG. Perhaps the greatest limitation of
AEEG is that it is not possible for a technician to monitor
recording quality and electrode integrity, and to adjust elec-
trodes if necessary. Failure of electrode contacts and occasion-
ally of ambulatory equipment compromises the quality of
AEEG recordings.
Although AEEG can provide days of continuous EEG mon-
itoring, it is generally not a substitute for inpatient video-EEG,
because simultaneous time-locked video provides substantial
diagnostic information not provided by EEG alone. Most com-
mercial AEEG systems are not designed to record time-locked
video along with EEG. Although some AEEG systems do sup-
port video, the requirement that the patient remains “on-
camera” defeats one advantage of AEEG, the capability of
recording EEG as the patient goes about routine activities.
Further, inpatient video-EEG monitoring allows trained per-
sonnel to observe, to interact with, and to test patients during
seizures or seizure-like episodes, acquiring important informa-
tion regarding semiology that is not captured by unsupervised
video recording. The inpatient setting makes it possible to safely
withdraw medication to increase the likelihood of recording
seizures, and, of course, the inpatient setting is uniquely suited
to invasive EEG recording (3).
These limitations notwithstanding, one group has reported
the successful use of outpatient monitoring as part of the
presurgical evaluation of a group of patients with temporal
lobe epilepsy (124). Further, seizure frequency is often
reported to decrease on hospitalization (125,126), perhaps
related to medication compliance or environment stressors
(124). Some have suggested a role for AEEG in conjunction
with inpatient video-EEG monitoring, either prior to or
following hospitalization (127).
735
Indications for Ambulatory EEG
Given these considerations, the principal indications for AEEG
monitoring are (128):
1. Diagnosis: In patients with suspected epilepsy, AEEG may
help to establish the diagnosis of epilepsy by recording inter-
ictal discharges and seizures, and may help to distinguish
epilepsy from nonepileptic seizures and other behavior dis-
orders, syncope, cardiac arrhythmias, and disturbances of
sleep.
2. Classification: The electrographic patterns and scalp distri-
butions of interictal discharges and seizures, their temporal
patterns of occurrence including precipitating events and
stimuli, as well as the EEG background activity may provide
useful information for refining the definition of seizure type
and the epilepsy diagnosis.
3. Quantification: AEEG may provide important information
regarding the frequency of discharges and seizures, as well as
documentation of response to medication. This may be par-
ticularly important in patients with frequent seizures, of
which they may not be aware.
Ambulatory EEG Equipment
Since the early days of galvanometers and photographic record-
ing devices, clinical neurophysiology has progressed in lockstep
with technology. The first commercial devices recorded four
channels of EEG (or three EEG plus a time code), and used a
four-channel analog 1/8-in. cassette tape recorder designed for
general industrial applications (129). A standard tape cassette
could store 24 hours of EEG. EEG signals were amplified using
specially designed small single-channel preamplifiers that were
affixed, along with electrodes, to the scalp with collodion. A
video display reader station permitted rapid visual review, as
fast as 1 hour of EEG/minute.
In the early 1980s, a nine-channel analog device was intro-
duced that used standard 1/4-in. recording tape. An important
advantage was that eight-channel EEG proved easier to inter-
pret than three- or four-channel EEG; with eight channels,
montages could be constructed that more closely resembled
traditional montages. While sensitivity for identification of
epileptiform abnormalities was similar given three- or eight-
channel data, review of eight-channel data afforded a lower
false-positive rate (130). Additionally, these devices offered
greater flexibility for adjustment of filters and gain, printouts of
selectable segments, and capability to easily locate the EEG
recorded at specific times during the study. A useful new feature
was the ability to convert one or several EEG channels to audio,
allowing seizures to be identified by listening to high-speed
playback (128).
Attachment of eight on-head preamplifiers to scalp with col-
lodion proved cumbersome and time consuming. Further,
although in principle on-head amplification can reduce artifact
related to electrode lead length and movement, it became
apparent that on-head amplification could not reduce artifact
from some important sources, including EMG and eye move-
ment. Further, fixing a standard EEG lead wire to the scalp with
collodion next to the electrode cup effectively mitigates
736 Part V ■ Complementary and Special Techniques
electrode artifacts produced by movement at the electrode
cup–scalp interface (128). With the introduction of eight-
channel AEEG devices, off-head amplifiers for use with stan-
dard EEG electrodes became available. The trend toward
off-head amplification has continued; nonetheless, on-head
amplification may be superior under conditions of vigorous
movement, for example, during a convulsive seizure. Direct
bipolar recordings can provide signal quality superior to that
produced by referential recordings reformatted into bipolar
montages (128,131).
During the 1980s, 16-channel digital AEEG devices were
introduced, still using tape for storage. Initially, the trade-off for
16 channels was limited storage capabilities; instead of retaining
the complete recording, intermittent samples, for example, 15
seconds every 10 minutes, as well as sections before and after
manual activation of a push button by the patient, were stored.
Subsequently hybrid “computer-assisted” AEEG systems com-
bined a 16-channel waist-worn device with a portable com-
puter, connected by a wire tether, that processed EEG in real
time using automated spike and seizure detection software
(132,133). Data available for clinical review consisted of inter-
mittent samples, manually recognized events, and automatic
detections, in much the same manner as the inpatient
video/monitoring systems in use at the time.
Current AEEG devices represent an important departure
from past designs. Present-day digital technology makes possible
portable EEG devices that are not much larger than a PDA or a
cell phone and are able to record EEG with fidelity and number
of channels equal to standard laboratory EEG machines. Indeed,
the designs of AEEG systems and standard EEG systems have
largely merged. In a typical EEG machine, amplification and
analog to digital conversion are performed in the “headbox”; the
computer performs control, data storage, and display functions.
An AEEG system simply adds battery power and solid-state
memory to the headbox. Since some headboxes, particularly
those intended for use in inpatient monitoring units, often con-
tain memory and rechargeable batteries, permitting the patient
to disconnect from the wired system for several hours, real dif-
ferences between AEEG systems and other EEG systems relate to
packaging, quantity of storage, battery life, and marketing.
The typical modern AEEG recorder provides for 21 or more
referential channels of EEG that can be reformatted in the same
manner as standard laboratory EEG. It samples EEG with 16-bit
resolution at typically 200 or 256 Hz/channel. It also has several
channels for inputs such as pulse oximetry and other
polysomnographic devices, EMG and EKG. It connects to a desk-
top computer using USB or Ethernet for startup, and for down-
load and review of data. It weighs just over 1 lb, is powered by a
several AA or rechargeable batteries, and has 2 to 4 GB of flash
memory capable of storing 2 to 4 days of continuous data. AEEG
systems are also available that support time-synchronized video.
Real-time processing of EEG for spike and seizure detection
was important to the design of earlier computer-assisted AEEG
systems that had insufficient storage capacity to retain entire 16-
channel day-long EEG recordings, and instead stored only sam-
ples and automatic detections. Storage capacity is no longer a
limiting factor, and present-day systems retain all EEG recorded.
When the study is complete, data are easily transferred to a desk-
top computer for review, and if desired, processing by automatic
spike and seizure detection software (27,30). As with inpatient
LTM recordings, the physician has the option of exhaustively
reviewing the entire recording, or relying on periodic samples,
“push button events,” and automated detections.
Future Directions
AEEG has the potential to provide accurate information about
the seizure frequency that, in some patients, may be difficult to
obtain otherwise (134). Underreporting of seizures by patients
is well recognized (123), in part because reporting requires the
patient to remember an event that itself commonly produces
amnesia. When seizures go undetected and underreported,
medical management may be compromised because the effi-
cacy of treatment can be difficult to ascertain. The apparent
relationship between seizure frequency and the risk of sudden
unexplained death in patients with epilepsy underscores the
importance of accurate measurement (135).
Proof of concept for chronic AEEG has been effectively pro-
vided by Neuropace Inc.’s investigational responsive neurostim-
ulation (RNS) device (136,137). Implanted within the skull and
connected to subdural or depth electrodes, the device is
designed to detect seizures and deliver electrical stimulation
intended to abort the seizure (138). The device also contains
memory sufficient to store about 30 minutes of four-channel
EEG corresponding to seizure detections, allowing it to func-
tion as a chronic AEEG recorder.
Technology is presently within reach that should make the
design of a device optimized for chronic AEEG recording and
monitoring feasible. High-capacity, dense, solid-state storage is
now in widespread use, in standard AEEG systems as well as
consumer electronic devices. Newly developed power harvesting
techniques may provide a solution to the problems of power
consumption and battery size and life that have, until now,
plagued implantable devices (139). In contrast to standard near-
field inductive coupling, used to recharge consumer devices and
which requires that the power source and powered device be in
close proximity, power can be transmitted using radio frequency
(RF) energy over somewhat greater distances. Holleman et al.
recently described a low-power (20 A at 1.9 V) prototype bat-
tery-free neural interface, which counted action potentials/sec-
ond, and was powered by a 900-MHz radio signal at a distance
of 1 m (140). Proprietary RF harvesting techniques claim to be
capable of safely providing milliwatt-range power over even
greater distances (141). It is easy to envision a device implanted
in the skull that records full-fidelity EEG through epidural peg
electrodes. Epidural electrodes would provide high-quality sta-
ble EEG signals, relatively immune from artifacts. At night, while
the patient sleeps, the device would recharge while data were
automatically downloaded to a computer that would run seizure
detection software, logging the results or transmitting them,
along with raw data if desired, back to the physician.
An attractive alternative to wired connections between mul-
tiple transcranial passive peg electrodes and an implanted AEEG
 Chapter 35 ■ Principles and Techniques for Long-Term EEG Recording (EMU, ICU, Ambulatory)
recording device could be a wireless local network, consisting of
small active peg electrodes (142) and the AEEG recorder, per-
haps implanted in the chest or abdominal wall. Each peg would
record a channel of EEG and transmit it back to the recorder,
where the data would be stored for subsequent download. The
network could use the Medical Implant Communication Service
band, 402 to 405 kHz, recently established by the FCC for radio
transmission to implanted medical devices (143). This concept
is readily expanded to include continuous recording of other
physiologic signals, such as EKG, respiration, and blood oxygen
saturation. The relatively liberal power budget afforded by RF
power harvesting might permit the functionality of implanted
AEEG systems to include not only data storage, but also real-
time seizure detection and, if effective methodologies emerge,
seizure prediction and warning (23).
REFERENCES
1. Gotman J. Seizure recognition and analysis. In: Gotman J, Ives JR,
Gloor P, eds. Long-Term Monitoring in Epilepsy. Amsterdam: Elsevier;
1985:133–145.
2. Hunter J, Jasper HH. A method of analysis of seizure pattern and elec-
troencephalogram; a cinematographic technique. Electroencephalogr
Clin Neurophysiol. 1949;1(1):113.
3. Velis D, Plouin P, Gotman J, et al. ILAE DMC Subcommittee on
Neurophysiology. Recommendations regarding the requirements
and applications for long-term recordings in epilepsy. Epilepsia.
2007;48(2):379–384.
4. Mirsattari SM, Tapsell LM, Ives JR, et al. MR imaging-compatible
electroencephalography electrode system for an epilepsy monitor-
ing unit. Am J Neuroradiol. 2008;29(9):1649–1651.
5. Mirsattari SM, Davies-Schinkel C, Young GB, et al. Usefulness of a
1.5 T MRI-compatible EEG electrode system for routine use in the
intensive care unit of a tertiary care hospital. Epilepsy Res. 2009;
84(1):28–32.
6. Jirsch JD, Urrestarazu E, LeVan P, et al. High-frequency oscillations
during human focal seizures. Brain. 2006;129:1593–1608.
7. Urrestarazu E, Chander R, Dubeau F, et al. Interictal high-
frequency oscillations (100–500 Hz) in the intracerebral EEG of
epileptic patients. Brain. 2007;130:2354–2366.
8. Jacobs J, LeVan P, Chander R, et al. Interictal high-frequency oscil-
lations (80–500 Hz) are an indicator of seizure onset areas inde-
pendent of spikes in the human epileptic brain. Epilepsia.
2008;49:1893–1907.
9. Vanhatalo S, Voipio J, Kaila K. Full-band EEG (fbEEG): a new stan-
dard for clinical electroencephalography. Clin EEG Neurosci.
2005;36(4):311–317.
10. Blum D. Prevalence of bilateral partial seizure foci and implica-
tions for electroencephalographic telemetry monitoring and
epilepsy surgery. Electroencephalogr Clin Neurophysiol. 1994;
91(5):329–336.
11. Marciani MG, Gotman J. Effects of drug withdrawal on location of
seizure onset. Epilepsia. 1986;27(4):423–431.
12. So N, Gotman J. Quantitative EEG analysis of carbamazepine
effects on amygdaloid kindled seizures in cats. Electroencephalogr
Clin Neurophysiol. 1990;76(1):63–72.
13. Zhou D, Wang Y, Hopp P, et al. Influence on ictal seizure semiol-
ogy of rapid withdrawal of carbamazepine and valproate in
monotherapy. Epilepsia. 2002;43(4):386–393.
737
14. Theodore WH, Porter RJ, Albert P, et al. The secondarily generalized
tonic–clonic seizure: a videotape analysis. Neurology. 1994;44(8):
1403–1407.
15. Spencer SS, Spencer DD, Williamson PD, et al. Ictal effects of anti-
convulsant medication withdrawal in epileptic patients. Epilepsia.
1981;22(3):297–307.
16. Marciani MG, Gotman J, Andermann F, et al. Patterns of seizure
activation after withdrawal of antiepileptic medication. Neurology.
1985;35(11):1537–1543.
17. Gotman J, Marciani MG. Electroencephalographic spiking activ-
ity, drug levels, and seizure occurrence in epileptic patients. Ann
Neurol. 1985;17(6):597–603.
18. Gotman J, Koffler DJ. Interictal spiking increases after seizures but
does not after decrease in medication. Electroencephalogr Clin
Neurophysiol. 1989;72(1):7–15.
19. Spencer SS, Goncharova II, Duckrow RB, et al. Interictal spikes on
intracranial recording: behavior, physiology, and implications.
Epilepsia. 2008;49(11):1881–1892.
20. Mizrahi EM. Pediatric electroencephalographic video monitoring.
J Clin Neurophysiol. 1999;16(2):100–110.
21. Qu H, Gotman J. A patient-specific algorithm for the detection of
seizure onset in long-term EEG monitoring: possible use as a
warning device. IEEE Trans Biomed Eng. 1997;44(2):115–122.
22. Shoeb A, Edwards H, Connolly J, et al. Patient-specific seizure
onset detection. Epilepsy Behav. 2004;5(4):483–498.
23. Mormann F, Andrzejak RG, Elger CE, et al. Seizure prediction: the
long and winding road. Brain. 2007;130:314–333.
24. Gotman J. Interhemispheric relations during bilateral spike-and-
wave activity. Epilepsia. 1981;22(4):453–466.
25. Gotman J. Automatic seizure detection: improvements and
evaluation. Electroencephalogr Clin Neurophysiol. 1990;76(4):
317–324.
26. Grewal S, Gotman J. An automatic warning system for epileptic
seizures recorded on intracerebral EEGs. Clin Neurophysiol.
2005;116(10):2460–2472.
27. Saab ME, Gotman J. A system to detect the onset of epileptic
seizures in scalp EEG. Clin Neurophysiol. 2005;116(2):427–442.
28. Gabor AJ, Leach RR, Dowla FU. Automated seizure detection using a
self-organising neural network. Electroencephalogr Clin Neurophysiol.
1996;99:257–266.
29. Webber WRS, Lesser RP, Richardson RT, et al. An approach to
seizure detection using an artificial neural network. Electro-
encephalogr Clin Neurophysiol. 1996;98:250–272.
30. Wilson SB, Scheuer ML, Emerson RG, et al. Seizure detection:
evaluation of the Reveal algorithm. Clin Neurophysiol. 2004;
115(10):2280–2291.
31. Qu H, Gotman J. A seizure warning system for long-term epilepsy
monitoring. Neurology. 1995;45(12):2250–2254.
32. Wilson SB. A neural network method for automatic and incre-
mental learning applied to patient-dependent seizure detection.
Clin Neurophysiol. 2005;116(8):1785–1795.
33. Gotman J, Flanagan D, Rosenblatt B, et al. Evaluation of an auto-
matic seizure detection method for the newborn EEG. Electro-
encephalogr Clin Neurophysiol. 1997;103(3):363–369.
34. Celka P, Colditz P. A computer-aided detection of EEG seizures in
infants: a singular-spectrum approach and performance compari-
son. IEEE Trans Biomed Eng. 2002;49(5):455–462.
35. Aarabi A, Wallois F, Grebe R. Automated neonatal seizure detec-
tion: a multistage classification system through feature selection
based on relevance and redundancy analysis. Clin Neurophysiol.
2006;117(2):328–340.
738 Part V ■ Complementary and Special Techniques
36. Aarts JH, Binnie CD, Smit AM, et al. Selective cognitive impair-
ment during focal and generalized epileptiform EEG activity.
Brain. 1984;107(pt 1):293–308.
37. Shewmon DA, Erwin RJ. Transient impairment of visual percep-
tion induced by single interictal occipital spikes. J Clin Exp
Neuropsychol. 1989;11(5):675–691.
38. Wilson SB, Harner RN, Duffy FH, et al. Spike detection. I.
Correlation and reliability of human experts. Electroencephalogr
Clin Neurophysiol. 1996;98(3):186–198.
39. Benbadis SR, Lin K. Errors in EEG interpretation and misdiagno-
sis of epilepsy. Which EEG patterns are overread? Eur Neurol.
2008;59:267–271.
40. Halford JJ. Computerized epileptiform transient detection in the
scalp electroencephalogram: obstacles to progress and the example
of computerized ECG interpretation. Clin Neurophysiol. 2009;
120(11):1909–1915.
41. Jordan KG. Continuous EEG and evoked potential monitoring in
the neuroscience intensive care unit. J Clin Neurophysiol. 1993;
10:445–475.
42. Jordan KG. Nonconvulsive seizures and nonconvulsive status
epilepticus detected by continuous EEG monitoring in the neuro-
ICU. Neurology. 1992;42(suppl 3):194.
43. Ballenger CE III, King OW, Gallagher B. Partial complex status
epilepticus. Neurology. 1983;33:1545–1552.
44. Bauer G, Aichner F, Mayr U. Nonconvulsive status epilepticus fol-
lowing generalized tonic–clonic seizures. Eur Neurol. 1982;21:
411–419.
45. Doose H. Nonconvulsive status epilepticus in childhood: clinical
aspects and classification. Adv Neurol. 1984;34:115–125.
46. Ellis JM, Lee SI. Acute prolonged confusion in later life as an ictal
state. Epilepsia. 1978;19:119–128.
47. Engel J Jr, ed. Seizures and Epilepsy. Philadelphia, PA: F.A. Davis;
1989.
48. Escueta AV, Boxley J, Stubbs N, et al. Prolonged twilight state and
automatisms: a case report. Neurology. 1974;24:331–339.
49. Porter RJ, Penry JK. Petit mal status. Adv Neurol. 1983;34:61–67.
50. Pritchard PB III, O’Neal DB. Nonconvulsive status epilepticus
following metrizamide myelography. Ann Neurol. 1984;16:
252–254.
51. Rumpl E, Hinterhuber H. Unusual “spike-wave stupor” in a
patient with manic-depressive psychosis treated with amitripty-
line. J Neurol. 1981;226:131–135.
52. Rumpl E, Lorenzi E, Hackl J, et al. The EEG at different stages of
acute secondary traumatic midbrain and bulbar brain syndromes.
Electroencephalogr Clin Neurophysiol. 1981;46:487–497.
53. Chatrian GE, Shaw CM, Leffman H. The significance of periodic
lateralized epileptiform discharges in EEG: an electrographic, clin-
ical and pathological study. Electroencephalogr Clin Neurophysiol.
1964;17:177.
54. Markland ON, Daly DO. Pseudoperiodic lateralized paroxysmal
discharges in electroencephalogram. Neurology. 1971;21:975.
55. Austin EI, Wilkus RJ, Longstreth WT. Etiology and prognosis of
alpha coma. Neurology. 1988;38:773.
56. Carroll WM, Mastaglia FL. Alpha and beta coma in drug intoxica-
tion uncomplicated by cerebral hypoxia. Electroencephalogr Clin
Neurophysiol. 1979;46:95.
57. Chokroverty S. “Alpha like” rhythms in electroencephalograms in
coma after cardiac arrest. Neurology. 1975;25:655.
58. Grindal AB, Suter C, Martinez AJ. Alpha-pattern coma: 24 cases
with 9 survivors. Ann Neurol. 1977;1:371.
59. Iragui VJ, McCutchen CB. Physiologic and prognostic significance
of “alpha coma.” J Neurol Neurosurg Psychiatry. 1983;46:632.
60. Sorensen K, Thomassen A, Wernberg M. Prognostic significance of
alpha frequency EEG rhythm in coma after cardiac arrest. J Neurol
Neurosurg Psychiatry. 1978;41:840.
61. Westmoreland B, Klass OW, Sharbrough FW, et al. Alpha-coma:
electroencephalographic, clinical, pathologic, and etiologic corre-
lations. Arch Neurol. 1975;32:713.
62. Jordan KG. Neurophysiologic monitoring in the neuroscience
intensive care unit. Neurol Clin. 1995;13:579–626.
63. Archibald JE, Drazkowski JF. Clinical applications of compressed
spectral analysis (CSA) in OR/ICU settings. Am J EEG Technol.
1985;25:13–36.
64. Vespa PM, Nuwer MR, Nenov V, et al. Increased incidence and
impact of nonconvulsive and convulsive seizures after traumatic
brain injury as detected by continuous EEG in the intensive care
unit. J Neurosurg. 1999;91:750–760.
65. Vespa PM, O’Phelan K, Mirabelli J, et al. Acute seizures after
intracerebral hemorrhage: a factor in progressive midline shift and
outcome. Neurology. 2003;60:1441–1446.
66. Claassen J, Hirsch LJ, Kreiter KT, et al. Quantitative continuous EEG
for detecting delayed cerebral ischemia in patients with poor-grade
subarachnoid hemorrhage. Clin Neurophysiol. 2004;115:2699–2710.
67. Pandian JD, Cascino GD, So EL, et al. Digital video-electroen-
cephalographic monitoring in the neurological–neurosurgical
intensive care unit: clinical features and outcome. Arch Neurol.
2004;61(7):1090–1094.
68. Maynard D, Prior PF, Scott DF. Device for continuous monitoring
of cerebral activity in resuscitated patients. Br Med J. 1969;4:
545–546.
69. Prior P. The EEG in Acute Cerebral Anoxia. Amsterdam: Excerpta
Medica; 1973.
70. Stidham GL, Nugent SK, Rogers MC. Monitoring cerebral electri-
cal function in the ICU. Crit Care Med. 1980;8:519–523.
71. Young GB, Jordan KG, Doig GS. An assessment of nonconvulsive
seizures in the intensive care unit using continuous EEG monitoring:
an investigation of variables associated with mortality. Neurology.
1996;47(1):83–89.
72. Grand’Maison F, Reiher J, Laduke CP. Retrospective inventory of
EEG abnormalities in partial status epilepticus. Electroencephalogr
Clin Neurophysiol. 1991;79:264–270.
73. Scholtes FB, Renier WO, Meinardi H. Nonconvulsive status epilep-
ticus: causes, treatment, and outcome in 65 patients. J Neurol
Neurosurg Psychiatry. 1996;61:93–95.
74. Litt B, Dizon L, Ryan D. Fatal nonconvulsive status epilepticus in
the elderly. Epilepsia. 1994;35:10–17.
75. Shinton RA, Gill JS, Melnick AK. The frequency, characteristics,
and prognosis of epileptic seizures at the onset of stroke. J Neurol
Neurosurg Psychiatry. 1988;51:273–276.
76. Olsen TS, Hogenhaven H, Thage O. Epilepsy after stroke.
Neurology. 1987;37:1209–1211.
77. Lancman ME, Crolinstock A, Morseini J, et al. Risk factors for
developing seizures after a stroke. Epilepsia. 1993;34:141–143.
78. Vespa PM, Ronne-Engström E, Smith M, et al. Increase in extracel-
lular glutamate caused by reduced cerebral perfusion pressure and
seizures after human traumatic brain injury: a microdialysis study.
J Neurosurg. 1998;89:971–982.
79. Vespa PM, Martin NA, Nenov V, et al. Delayed increase in extracel-
lular glycerol with post-traumatic electrographic epileptic activity:
support for the theory that seizures induce secondary injury. Acta
Neurochir Suppl. 2002;81:355–357.
80. Bergamasco B, Bergamini L, Doriguzzi T. Clinical value of the
sleep electroencephalographic patterns in post-traumatic coma.
Acta Neurol Scand. 1968;44:495–511.
 Chapter 35 ■ Principles and Techniques for Long-Term EEG Recording (EMU, ICU, Ambulatory)
81. Bergamasco B, Bergamini L, Doriguzzi T, et al. EEG sleep pat-
terns as a prognostic criterion in post-traumatic coma. Electro-
encephalogr Clin Neurophysiol. 1968;24:374–377.
82. Bricolo A, Turella G, Ore GD, et al. A proposal for the EEG eval-
uation of acute traumatic coma in neurosurgical practice.
Electroencephalogr Clin Neurophysiol. 1973;34:789.
83. Bricolo A, Turazzi S, Faccioli F, et al. Clinical application of
compressed spectral array in long-term EEG monitoring of
comatose patients. Electroencephalogr Clin Neurophysiol. 1978;
45:211–225.
84. Bricolo A, Turazzi S, Faccioli F. Combined clinical and EEG exam-
inations for assessment of severity of acute head injuries: electro-
physiological methods. Acta Neurochir Suppl. 1979;28:35–39.
85. Astrup J, Siesjö BK, Symon L. Thresholds in cerebral ischemia—
the ischemic penumbra. Stroke. 1981;12:723–725.
86. Sundt TM Jr, Sharbrough FW, Piepgras DG, et al. Correlation of
cerebral blood flow and electroencephalographic changes during
carotid endarterectomy: with results of surgery and hemody-
namics of cerebral ischemia. Mayo Clin Proc. 1981;56:533–543.
87. Baron JC. Perfusion thresholds in human cerebral ischemia: his-
torical perspective and therapeutic implications. Cerebrovasc Dis.
2001;11:2–8.
88. Nuwer MR, Jordan SE, Ahn SS. Evaluation of stroke using EEG
frequency analysis and topographic mapping. Neurology. 1987;
37:1153–1159.
89. Jordan KG. Continuous EEG monitoring in the neuroscience
intensive care unit and emergency department. J Clin Neuro-
physiol. 1999;16:14–39.
90. Finnigan SP, Rose SE, Walsh M, et al. Correlation of quantita-
tive EEG in acute ischemic stroke with 30-day NIHSS score:
comparison with diffusion and perfusion MRI. Stroke.
2004;35:899–903.
91. Suzuki A, Yoshioka K, Yasui N. Clinical application of EEG
topography in cerebral ischemia: detection of functional
reversibility and hemodynamics. Brain Topogr. 1990;3:167–174.
92. Van Putten MJ, Tavy DL. Continuous quantitative EEG monitor-
ing in hemispheric stroke patients using the brain symmetry
index. Stroke. 2004;35:2489–2492.
93. Schneider AL, Jordan KG. Regional attenuation without delta
(RAWOD): a distinctive EEG pattern that can aid in the diagno-
sis and management of severe acute ischemic stroke. Am J
Electroneurodiagn Technol. 2005;45:102–117.
94. Daly DD, Markand ON. Focal brain lesions. In: Daly DD, Pedley
TA, eds. Current Practice of Clinical Electroencephalography. 2nd
ed. New York, NY: Raven Press; 1990:335–370.
95. Niedermeyer E. Cerebrovascular disorders and EEG. In:
Niedermeyer E, Lopes da Silva F, eds. Electroencephalography. 5th
ed. Baltimore, MD: Urban & Schwarzenberg Inc.; 2005:339–362.
96. Claassen J, Hirsch LJ, Frontera JA, et al. Prognostic significance of
continuous EEG monitoring in patients with poor-grade sub-
arachnoid hemorrhage. Neurocrit Care. 2006;4:103–112.
97. Hirsch LJ, Claassen J, Mayer SA, et al. Stimulus-induced rhyth-
mic, periodic, or ictal discharges (SIRPIDs): a common EEG
phenomenon in the critically ill. Epilepsia. 2004;45:109–123.
98. Weir B, Grace M, Hansen J, et al. Time course of vasospasm in
man. J Neurosurg. 1978;48:173–178.
99. Hijdra A, Van Gijn J, Nagelkerke NJ, et al. Prediction of delayed
cerebral ischemia, rebleeding, and outcome after aneurysmal
subarachnoid hemorrhage. Stroke. 1988;19:1250–1256.
100. Murayama Y, Malisch T, Guglielmi G, et al. Incidence of cerebral
vasospasm after endovascular treatment of acutely ruptured
aneurysms: report on 69 cases. J Neurosurg. 1997;87:830–835.
739
101. Charpentier C, Audibert G, Guillemin F, et al. Multivariate analy-
sis of predictors of cerebral vasospasm occurrence after aneurys-
mal subarachnoid hemorrhage. Stroke. 1999;30:1402–1408.
102. Hop JW, Rinkel GJ, Algra A, et al. Initial loss of consciousness and
risk of delayed cerebral ischemia after aneurysmal subarachnoid
hemorrhage. Stroke. 1999;30:2268–2271.
103. Qureshi AI, Sung GY, Razumovsky AY, et al. Early identification
of patients at risk for symptomatic vasospasm after aneurismal
subarachnoid hemorrhage. Crit Care Med. 2000;28:984–990.
104. Claassen J, Bernardini GL, Kreiter K, et al. Effect of cisternal and
ventricular blood on risk of delayed cerebral ischemia after sub-
arachnoid hemorrhage: the Fisher scale revisited. Stroke. 2001;
32:2012–2020.
105. Kirmani JF, Qureshi AI, Hanel RA, et al. Silent cerebral infarction
in poor-grade patients with subarachnoid hemorrhage. Neurology.
2002;58(7):159.
106. Labar DR, Fisch BJ, Pedley TA, et al. Quantitative EEG monitor-
ing for patients with subarachnoid hemorrhage. Electro-
encephalogr Clin Neurophysiol. 1991;78:325–332.
107. Vespa PM, Nuwer MR, Juhasz C, et al. Early detection of
vasospasm after acute subarachnoid hemorrhage using continu-
ous EEG ICU monitoring. Electroencephalogr Clin Neurophysiol.
1997;103:607–615.
108. Hellström-Westas L, De Vries LS, Rosén I. Neonatal ICU moni-
toring. In: Marc R, Nuwer MR, eds. Intraoperative Monitoring of
Neural Function. Handbook of Clinical Neurophysiology.
Amsterdam: Elsevier; 2008.
109. Greisen G, Hellström-Westas L, Lou H, et al. EEG depression and
germinal layer haemorrhage in the newborn. Acta Paediatr Scand.
1987;76:519–525.
110. Hellström-Westas L, Rosen I, Svenningsen NW. Cerebral function
monitoring during the first week of life in extremely small low
birthweight (ESLBW) infants. Neuropediatrics. 1991;22:27–32.
111. Hellström-Westas L, Klette H, Thorngren-Jerneck K, et al. Early
prediction of outcome with aEEG in preterm infants with large
intraventricular hemorrhages. Neuropediatrics. 2001;32:319–324.
112. Olischar M, Klebermass K, Kuhle S, et al. Reference values for
amplitude-integrated electroencephalographic activity in preterm
infants younger than 30 weeks’ gestational age. Pediatrics. 2004;
113:e61–e66.
113. Sisman J, Campbell D, Brion LP. Amplitude integrated EEG in
preterm infants: maturation of background pattern and amplitude
voltage with postmenstrual age and gestational age. J Perinatol.
2005;25:391–396.
114. Al Naqeeb N, Edwards AD, Cowan F, et al. Assessment of neona-
tal encephalopathy by amplitude integrated electroencephalogra-
phy. Pediatrics. 1999;103:1263–1271.
115. Hellström-Westas L, Rosen I, Svenningsen NW. Predictive value
of early continuous amplitude integrated EEG recordings on
outcome after severe birth asphyxia in full term infants. Arch Dis
Child. 1995;72:F34–F38.
116. Toet MC, Hellström-Westas L, Groenendaal F, et al. Amplitude
integrated EEG 3 and 6 hours after birth in full term neonates
with hypoxic-ischaemic encephalopathy. Arch Dis Child Fetal
Neonatal Ed. 1999;81:F19–F23.
117. Bickford RG, Fleming N, Billinger T. Compression of EEG data.
Trans Am Neurol Assoc. 1971;96:118–122.
118. Karnaze DS, Marshall LF, Bickford RG. EEG monitoring of
clinical coma: the compressed spectral array. Neurology. 1982;32:
289–292.
119. Nuwer MR. EEG and evoked potentials: monitoring cerebral func-
tion in the neurosurgical ICU. In: Martin NA, ed. Neurosurgery
740 Part V ■ Complementary and Special Techniques
Clinics of North America, Vol. 5, No. 4, Neurosurgical Intensive Care.
Philadelphia, PA: W.B. Saunders; 1994:647–659.
120. Scheuer ML, Wilson SB. Data analysis for continuous EEG mon-
itoring in the ICU: seeing the forest and the trees. J Clin
Neurophysiol. 2004;21:353–378.
121. Bridgers SL, Ebersole J. Ambulatory cassette EEG in clinical practice:
experience with 500 patients. Neurology. 1985;35(12): 1167–1168.
122. Liporace J, Tatum W 4th, Morris GL 3rd, et al. Clinical utility of
sleep-deprived versus computer-assisted ambulatory 16-channel
EEG in epilepsy patients: a multi-center study. Epilepsy Res.
1998;32:357–362.
123. Blum DE, Eskola J, Bortz JJ, et al. Patient awareness of seizures.
Neurology. 1966;47:260–264.
124. Chang BS, Ives JR, Schomer DL, et al. Outpatient EEG monitor-
ing in the presurgical evaluation of patients with refractory tem-
poral lobe epilepsy. J Clin Neurophysiol. 2002;19(2):152–156.
125. Nuwer MR, Engel J Jr, Sutherling WW, et al. Monitoring at the
University of California, Los Angeles. Electroencephalogr Clin
Neurophysiol. 1985;37:385–402.
126. Riley T, Porter RJ, White BG, et al. The hospital experience and
seizure control. Neurology. 1981;31:912–915.
127. Schomer DL. Ambulatory EEG telemetry: how good is it? J Clin
Neurophysiol. 2006;23(4):294–305.
128. Ebersole J. Ambulatory cassette EEG. J Clin Neurophysiol. 1985;
2(4):397–418.
129. Ives JR, Woods J. 4-Channel 24 hour cassette recorded for long-
term EEG monitoring of ambulatory patients. Electroencephalogr
Clin Neurophysiol. 1975;39:88–92.
130. Ebersole JS, Bridgers SL. Direct comparison of 3- and 8-channel
ambulatory cassette EEG with intensive inpatient monitoring.
Neurology. 1985;35:846–854.
131. Schomer DL, Ives JR, Schachter SC. The role of ambulatory EEG in
the evaluation of patients for epilepsy surgery. J Clin Neurophysiol.
1999;16(2):116–129.
132. Gotman J. Automatic recognition of epileptic seizures in the
EEG. Electroencephalogr Clin Neurophysiol. 1982;54:530–540.
133. Morris GL III, Galezowska J, Leroy R, et al. The results of
computer-assisted ambulatory 16-channel EEG. Electro-
encephalogr Clin Neurophysiol 1994;91:229–231.
134. Tatum WO, Winters L, Gieron M, et al. Results of 502 patients
using computer-assisted ambulatory EEG. J Clin Neurophysiol.
2001;18(1):14–19.
135. Tomson T, Nashef L, Ryvlin P. Sudden unexpected death in
epilepsy: current knowledge and future directions. Lancet Neurol.
2008;7(11):1021–1031.
136. Spanaki M, Smith B, Burdette D, et al. Chronic measurement of
increased epileptiform activity during menses using the responsive
neurostimulator system (RNS) in a patient with catamenial
seizures; 2005. Available at: http://www.neuropace.com/resources/
publications/0512.html#8. Accessed May 19, 2010.
137. Vossler DG, Tcheng T. Ambulatory intracranial ictal electroen-
cephalogram patterns recorded chronically using the first
implanted, self-contained ECOG recording and analysis instru-
ments; 2005. Available at: http://www.neuropace.com/resources/
publications/0512.html#9. Accessed May 19, 2010.
138. Fischell RE, Rischell DR. Integrated system for EEG monitoring
and electrical stimulation with a multiplicity of electrodes.
United States patent no. 6,320,049 B1. 2001.
139. Sauer C, Stanacevic M, Cauwenberghs G, et al. Power harvesting
and telemetry in CMOS for implanted devices. IEEE Trans
Circuits Syst I Regular Pap. 2005;52(10):2605–2613.
140. Holleman J, Yeager D, Prasad R, et al. NeuralWISP: an energy-
harvesting wireless neural interface with 1-m range. In: IEEE
Biological Circuits and Systems (BioCAS) Conference, November
2008. Available at: http://web.media.mit.edu/~jrs/ neuralwisp.pdf.
Accessed May 19, 2010.
141. Powercast I; 2008. Available at: http://powercastco.com/life-
time_power_app.html. Accessed May 19, 2010.
142. Bhutani A, McGarvey Z, Frye J, et al. Remote monitoring of EEG
signals through wireless sensor networks. In: Multi-Disciplinary
Engineering Design Conference; 2008. Rochester Institute of
Technology. Available at: http://edge.rit.edu/content/P08050/
pu blic/ PDF%20Docu m en t at ion / P08050_Tech Paper.pdf.
Accessed May 19, 2010.
143. Yuce MR, Ng SW, Myo NL, et al. Wireless body sensor network
using medical implant band. J Med Syst. 2007;31:467–474.
144. Vespa PM, Nenov V, Nuwer MR. Continuous EEG monitoring in
the intensive care unit: early findings and clinical efficacy. J Clin
Neurophysiol. 1999b;16:1–13.
145. Young GB, Doig GS. Continuous EEG monitoring in comatose
intensive care patients: epileptiform activity in etiologically dis-
tinct groups. Neurocrit Care. 2005;2(2):5–10.

Infraslow EEG Activity
SAMPSA VANHATALO, JUHA VOIPIO, AND KAI KAILA
INTRODUCTION
The conventional way of recording EEG is more than half a cen-
tury old. The early pioneers were faced with the challenge to
design routine EEG techniques that enabled recordings of the
most salient features in the human EEG known at that time. A
major obstacle was encountered in attempts to faithfully record
slow events. Electrode drifts that produced artifactual slow sig-
nals tended to contaminate the recordings and to saturate the
amplifier’s dynamic range, and they also pushed the polygraph
recorder pens out of scale. In order to circumvent these prob-
lems, the EEG amplifiers became furnished with an inbuilt
high-pass filter. Hence, all kinds of slow signals, whether physi-
ologic or artifactual, were eliminated.
In practice, the technical compromise described above
means that the conventional EEG has a poor low-frequency
response, which results in attenuation and distortion of signals
at <0.5 Hz. This is in sharp contrast with a wealth of data show-
ing that the physiologic frequency range of human EEG signals
ranges from infraslow (0.01 Hz or even less; this chapter) to
ultrafast (up to several hundreds of hertz; Chapter 37) frequen-
cies. In this chapter, several lines of evidence will be presented
to demonstrate that, at the present stage of EEG technology, sta-
bility is not an issue that would require any trade-off at the
expense of low-frequency response. Notably, filtering off the
lower end of the EEG spectrum by the standard high pass at
around 0.5 Hz can lead to a situation where the most salient
features of the EEG are effectively deleted. This kind of a mas-
sive loss of relevant data is clearly evident, for instance, in stan-
dard EEG recordings of the immature human brain and during
epileptic events.
Because of historical reasons, the terminology related to
recording slow EEG events is somewhat confusing. The term
direct current EEG (DC-EEG) has been widely used in older lit-
erature to imply a (ideal) frequency response of the EEG with a
minimum at 0 Hz. However, the term “DC” has at least two
connotations, which are misleading. First, it is impossible to
have a per definition 0 Hz response, since at 0 Hz the signal level
is constant and never changing. Second, the use of the term
DC-EEG puts an a priori emphasis on the low-frequency part
of the EEG spectrum but, as we will show below, looking at low
frequencies does not compromise any simultaneous analyses
carried out at fast or even ultrafast frequencies. The results
described here show that EEG responses, with a time course of
several minutes to achieve their peak levels, can be easily
recorded if attention is paid to trivial aspects of EEG electrode
and amplifier design and characteristics (1–5).
CHAPTER
36
Hence, instead of the old “DC-EEG” nomenclature, we pre-
fer using the term Full-band EEG (FbEEG). This conveys the
idea that the full, physiologically and clinically relevant EEG
bandwidth can be readily examined without favoring any fre-
quency band at the expense of another. The term DC shift is
used here to indicate slow shifts in FbEEG baseline that would
go unnoticed in conventional EEG. DC-coupled amplifiers are
devices that record electrode signals without elimination of the
FbEEG baseline by inbuilt high-pass filters.
EVENTS VERSUS OSCILLATIONS
In the context of infraslow activity, it is especially important to
acknowledge the distinction between events and oscillations (6).
A variety of EEG phenomena with a duration of seconds are,
indeed, not genuine oscillations but discrete events. As will be
shown below, spontaneous activity transients (SATs) in
preterm babies or arousal shifts in older subjects are very slow
transient events nesting complex patterns of high-frequency
EEG activity, whereas ictal DC shifts during epileptic seizures
are seen as slow deflections in the recording baseline. Although
the infraslow activity seen during sleep or cognitive tasks
clearly reflects cyclic changes, a well-defined oscillation fre-
quency is not discernible in the spectrograms. Hence it has
been called either infraslow oscillation (ISO) (7) or infraslow
fluctuation (ISF) (8).
TECHNICAL ASPECTS
Recording of a genuine FbEEG signal sets requirements both to
the amplifier and to the electrode–skin interface. It is important
to note that these requirements are not only related to the DC
stability already mentioned above, but addition of infraslow
and ultrafast frequency ranges introduces new sources of noise
that are of technical or of biologic origin (e.g., baseline drifts
and electrical interference from other devices or skin potential,
respectively). Practical solutions to overcome these problems
are, however, well established, often trivial, and widely available
in several modern higher end installations.
Regarding the amplifier, detection of the infraslow events
requires a genuine DC-coupled amplifier with a sufficiently
high input impedance and DC stability as well as a wide-
enough dynamic range (preferably ± 100 mV or higher),
whereas detection of the ultrafast frequencies requires a high-
enough sampling rate and a large data storage capacity. Several
clinically suitable, DC-coupled amplifiers of this kind have
recently become commercially available.
741
742 Part V ■ Complementary and Special Techniques
Regarding the electrode–skin interface, detection of the
infraslow events requires a DC-stable electrode–skin interface
(Fig. 36.1). This is achieved by using electrode materials that are
nonpolarizable (in practice Ag/AgCl) together with a gel con-
taining chloride, which is an absolute requirement for obtain-
ing a stable electrode potential (3). In addition, electrodes need
to be firmly fixed on the skin to minimize artifacts caused by gel
leakage or drying. Drying of the gel gives rise to an increase in
its electrolyte content, and, for example, a gradual 10% increase
in the Cl concentration seen by the Ag/AgCl electrode will
generate a negative drift of 2.5 mV. Finally, skin-borne signals
(9) caused by electrode movements or galvanic skin reactions
(GSR/SSR) can be strongly suppressed or eliminated by short-
circuiting the epithelium at the recording site (various tech-
niques are available for this purpose, see also Refs. 10, 11).
Sufficient short-circuiting of the skin is painless enough to be
performed in even sleeping neonates (12).
Compromises in recording setup are sometimes necessary,
especially under clinical conditions (13,14). For instance, if
the electrode–skin interface is not optimized as described
above (Fig 36.1), most slow events are hard if not impossible
to be distinguished from noncerebral artifacts. The situation
is, of course, different if the subjects are calm and cooperative
(as in research labs) as opposed to recording restless patients
(as in intensive care units). Our own experience has shown
that a suboptimal electrode–skin interface may still give rea-
sonably good clinical recordings in calm and/or sleeping
patients (4,5,15) if the EEG events of interest are phasic with
relatively short duration (such as SATs in preterm babies, see
section "Slow Activity in the Preterm Human EEG").
However, recordings of longer responses (such as ISOs dur-
ing sleep) or phasic events accompanied with changes in the
autonomic nervous system (such as arousal shifts, which may
include GSRs) will no longer be reliable. Movement artifacts
and slow baseline drifts will also be very much larger if the
Figure 36.1. A schematic drawing illustrating the requirements for
a stable electrode–skin interface. A: A nonpolarizing electrode
element (Ag| AgCl) is sealed in an insulating plastic housing that is
filled with chloride-containing electrode gel and tightly attached
to the skin in order to avoid gel leakage and drying. The epithelium
under the electrode is scratched or punctured in order to short circuit
skin-borne potentials. The housing of the electrode may have a
small opening for applying gel and scratching/ puncturing the skin.
B: A simplified equivalent circuit of the skin where the resistance and
potential difference across intact skin epithelium are represented
by Re and Ee , and the subepithelial resistance by Rd . Short-
circuiting the epithelium introduces a low-resistance shunt pathway
Rsh , which causes the transepithelial potential difference to fall
to Ee (Rsh )>(Re Rsh ) .
transepithelial potential is left intact under the electrode.
Another compromise that is sometimes encountered is the
use of polarizable (DC unstable) electrodes in combination
with high-pass filtering at an extremely low frequency. While
such recordings may be used to detect genuine DC shifts of a
foreseeable duration, they may equally give spurious results if
the evident technical pitfalls are ignored. Unlike reversible
electrodes that maintain their electrode potential and provide
a primarily resistive coupling of slow signals across the elec-
trode interface, capacitive signal coupling plays a major role
in polarizable electrodes. Therefore, it is possible that the
low-frequency cutoff is not set by the selected high-pass filter
but rather by the circuit formed by the electrode and the
amplifier input, resulting in unpredictable distortion of slow
signals. Furthermore, in this kind of recordings slow fluctua-
tions of the drifting electrode potential (3) will appear as
infraslow noise on an apparently stable baseline.
The challenges of interpretation of FbEEG data are mostly
similar to those of conventional EEG data. Lack of high-pass fil-
tering in FbEEG changes the visual appearance of slow artifacts
(e.g., eye and tongue movements) (16–18), but their visual
recognition in FbEEG recordings is easy with little experience.
Median filters and/or spatial filters, for near-complete remov-
ing of many artifacts, are now readily available in pertinent
analysis software packages. Skin-borne potentials (“sweat arti-
facts”), in turn, do not cause problems in FbEEG recordings if
they are excluded by short-circuiting (see above).
SLOW ACTIVITY IN THE
PRETERM HUMAN EEG
In FbEEG recordings from preterm neonates the most salient
frequency range that dominates the total spectral power is
much below the conventional EEG frequency band (Fig. 36.2B)
(12,15,19). This is consistent with previous well-known obser-
vations that an abundance of slow activity is a major feature of
EEG activity in the immature brain (20,21) (see also Chapters 9
and 25). In babies born 2 or more months before term age, the
spectral power peaks at frequencies as low as 0.01–0.1 Hz. It is
obvious that when recorded using conventional EEG, this kind
of activity must be highly attenuated and distorted by filtering.
Hence, only faint echoes of the slow neonatal EEG activity can
be detected at around the lower edge of the conventional
recording bandwidth.
Most notably, FbEEG showed that the preterm human EEG
exhibits very slow (up to 5 seconds) SATs. Furthermore, most
of the faster EEG activity is nested within these slow events,
and hence, practically all of the EEG activity of the most
immature cortex is associated with SATs (Fig. 36.2A) (12,15).
This is an extremely interesting finding in itself and in light of
animal experiments that have demonstrated that endoge-
nously driven, spontaneous activity is crucial in shaping neu-
ronal connectivity in the developing brain wiring already at an
immature stage where sensory input has little or no role at all
(reviewed in Refs. 19, 22, 23). Moreover, it has been shown that
sensory input to the as yet poorly wired brain is able to trigger

Figure 36.2. Slow activity in the perinatal human cortex. Specimen
recording of FbEEG activity (at O1 vs. Cz) of a premature neonate at 33
weeks of conceptional age shows prominent spontaneous activity tran-
sients. FbEEG data (A, lower trace) was high-pass filtered off-line at
0.5 Hz (upper trace) in order to demonstrate the pronounced loss
(attenuation, distortion) of slow activity. This is also clearly seen in the
power spectrum of FbEEG data from the same experiment (B). Note the
small fraction of the total EEG power above the conventional cutoff fre-
quency of 0.5 Hz (dashed line). FFT was obtained from three 3-minute
segments, using 60-second Hanning window. (C) Tactile stimulation
(asterisk) to palm area in a preterm baby (30 weeks conceptional age)
results in very prominent EEG events in the contralateral electrode (C3)
above somatosensory cortex. These events are characterized by a slow-
wave nesting faster activity, and they have striking similarities to some
SAT events generated spontaneously (i.e., without apparent stimuli) in
the same cortical area (19).
self-organized network events of this kind (Fig. 36.2C)
(reviewed in Ref. 24; see also Ref. 25).
FbEEG recordings do not only unravel a fully novel type of
activity in the immature human cortex in the low-frequency
bandwidth. In addition, it is obvious that these data call for a
revised nomenclature for the perinatal EEG (19,22). Future
studies will show whether work of this kind might set the stage
Chapter 36 ■ Infraslow EEG Activity 743
for the identification and diagnostics of activity-dependent dis-
eases and malfunctions of the immature brain as proposed by
Shatz et al. (26).
INFRASLOW EEG OSCILLATIONS
AND VOLTAGE SHIFTS DURING SLEEP
Using FbEEG, ISOs that take place at a wide range of
frequencies (0.02–0.2 Hz) with an amplitude of up to several
tens of microvolts are readily observed during non-REM
sleep (7). Interestingly, the phase of ISO shows a robust corre-
lation with higher frequency EEG activities (Fig. 36.3) as well as
with phasic brain events, such as the K-complex or interictal
epileptiform activity. ISOs recorded by FbEEG in humans may
thus reflect slow cyclic modulation of cortical excitability under
both physiologic and pathophysiologic conditions. This idea
fits well with results obtained in in vivo experiments on
rats (27), rabbits (28), and monkeys (29).
Transient arousals during sleep are associated with a vertex-
positive DC shift with an amplitude that may exceed hundred
microvolts and a duration of several tens of seconds (Fig. 36.4).
A comparable, very slow positive DC shift is also seen during
awakening or change in sleep stage to a more superficial one,
whereas a vertex-negative DC shift takes place when a person
falls asleep or shifts to deeper sleep stages (30).
INFRASLOW EEG OSCILLATIONS
AND PSYCHOPHYSICAL PERFORMANCE
Recently, we addressed the question whether infraslow EEG
oscillations in awake subjects (31) are linked to behavioral
performance. Prominent infraslow EEG fluctuations were
observed during a somatosensory detection task experiment,
with a log–log linear increase in the power spectrum down to
below 0.01 Hz (Fig. 36.5B) (8). The subject’s ability to detect
the stimulus was strongly correlated with the phase but not
with the amplitude of the infraslow EEG fluctuations in the
0.01 to 0.1 Hz band (Fig. 36.5A). In addition, the phase of the
infraslow EEG was strongly correlated with higher frequency EEG
activity. These results and the observation that EEG activity at
Figure 36.3. Infraslow oscillation during sleep. A 2-minute epoch of
FbEEG activity recorded during slow-wave sleep (midfrontal derivation
against a calculated linked mastoid reference, adult subject). Data are
shown as FbEEG and following band-pass filtering at 0.02–0.2 Hz to visu-
alize the prominent infraslow oscillation and its phase locking to activity
at higher frequencies. (Modified from Vanhatalo S, Palva JM, Holmes MD,
et al. Infraslow oscillations modulate interictal epileptic activity in the
human cortex during sleep. Proc Natl Acad Sci USA. 2004;191:5053–5057.)
744 Part V ■ Complementary and Special Techniques
Figure 36.4. Direct current (DC) shifts associated with arousals during
sleep. Large positive DC shifts are seen at vertex during transient
arousals. One event is shown on an expanded time scale in the middle
and lower traces to illustrate how only the faster activity that is associ-
ated with the ascending phase of the positive shift can be seen when
using a conventional EEG bandwidth. Unpublished observations of
S. Vanhatalo, J.M. Palva, J. Voipio, and K. Kaila.
conventional frequencies correlates with <0.1 Hz blood oxy-
genation level-dependent (BOLD) signal fluctuations in func-
tional magnetic resonance imaging (32,33) strongly support the
idea that FbEEG provides a powerful means for the monitoring
of slow fluctuations in the functional state of cerebral networks.
SLOW EEG SIGNALS GENERATED
BY SEIZURE ACTIVITY
Invasive recordings in experimental animals (34–37) and
humans (38–40) have established that seizures are associated
with very slow EEG responses. Early pioneering work in the
1960s (41–43) demonstrated that ictal DC shifts may be
recorded also from the scalp during generalized seizures, but
technical difficulties have limited noninvasive, ictal recordings of
focal seizures until recently (11,16,44). In addition, some more
recent invasive studies using DC-incompatible electrodes (3)
and long-time constant AC amplifiers have demonstrated vari-
able low-frequency fluctuations at the seizure focus (44–46).
Our recent studies (11,16) using FbEEG recordings on
epilepsy patients undergoing presurgical evaluation demon-
strated that focal seizures are associated with DC shifts that
confine to the area with seizure activity (Fig. 36.6). The ictal DC
Figure 36.6. Lateralized direct current (DC)-shift during a
seizure. FbEEG recording from scalp during a subclinical par-
tial seizure from an adult patient with temporal lobe epilepsy
reveals an unilateral negative DC shift with a clear spatiotem-
poral correlation with spiking activity. All traces are shown
against a linked Cz + Oz reference. (Modified from Vanhatalo
S, Holmes MD, Tallgren P, et al. Very slow EEG responses lat-
eralize temporal lobe seizures: an evaluation of noninvasive
DC-EEG. Neurology. 2003; 60:1098–1102.)
Figure 36.5. Infraslow fluctuation in EEG and cognitive performance
in an awake subject. A specimen FbEEG recording at 0 to 200 Hz dur-
ing a somatosensory detection task experiment in which threshold
level stimuli were delivered at random 1.5- to 4.5-second intervals
(A). Successful detections of stimuli are indicated with vertical lines,
and they show pronounced clustering and phase locking to the rising
phase of the infraslow EEG fluctuation. The mean power spectrum
of FbEEG data from 10 subjects shows a log–log linear increase in
power from 5 to 0.005 Hz (B). (Modified from Monto S, Palva S,
Voipio J, et al. Very slow EEG fluctuations predict the dynamics of
stimulus detection and oscillation amplitudes in humans. J Neurosci.
2008;28: 8268–8272.)
shifts usually begin within seconds after the initiation of elec-
trographic seizure activity, continuing throughout the seizure
with slow fluctuation of the amplitude that can reach a level of
over 100 V (Fig. 36.6). While the polarity of the ictal DC shifts
has been invariably negative when recording directly on the
cortical surface above the ictal focus, it may vary when record-
ing seizures from the scalp.
Using simultaneous intracranial and FbEEG recordings, we
found that the onset of the scalp-recorded DC shift can often
disclose the side (lateralization) of seizure initiation (Fig. 36.6),
even in cases where other noninvasive studies (such as video-
EEG and neuroimaging) are equivocal (16, 47). These findings
strongly suggest that in noninvasive recordings of ictal EEG
activity (48), the bandwidth should be extended to the low
frequencies that are readily seen in FbEEG.
NONNEURONAL EEG SIGNALS ASSOCIATED
WITH CHANGES IN RESPIRATION
AND HEMODYNAMICS
While it is evident that fast EEG activity has a neuronal origin,
slow EEG signals may arise from a variety of sources, including
both neuronal and nonneuronal generators. Below, we use the

Figure 36.7. A large nonneuronal direct current (DC) shift caused by
hyperventilation. Voluntary hyperventilation gives rise to a progressively
increasing negative DC shift that has a millivolt-scale amplitude at ver-
tex (mastoid reference, healthy adult subject). The DC shift is closely
paralleled by a fall in end-tidal partial pressure of CO2 (PCO2; continu-
ous capnograph signal). (Modified from Voipio J, Tallgren P, Heinonen
E, et al. Millivolt-scale DC shifts in the human scalp EEG: evidence for
a nonneuronal generator. J Neurophysiol. 2003;89: 2208–2214.)
term nonneuronal when referring to generators of EEG signals
that do not arise from the brain tissue parenchyma, that is,
neurons and glial cells (2).
In order to study the contribution of nonneuronal potential
shifts to human FbEEG signals, we have used manipulations
known to induce pH shifts at the level of the blood–brain bar-
rier (BBB) as well as large hemodynamic changes in healthy sub-
jects (2,17) (Fig. 36.7). Voluntary hyperventilation for a few
minutes induced a progressive negative shift with an astonish-
ingly high peak value of up to 2 mV at vertex, whereas hypoven-
tilation caused positive shifts (2). As discussed in detail
elsewhere (2), the high amplitude and duration (several min-
utes) of these scalp-recorded signals cannot be explained on the
basis of any neuronal generator mechanism, such as a change in
the state of tonic excitation of apical dendrites in cortical pyram-
idal neurons (36,49). Furthermore, bilateral jugular vein com-
pression, head-up and head-down tilt, and Valsalva and Mueller
maneuvers (i.e., exhaling and inhaling against a load, respec-
tively) all resulted in pronounced DC shifts that were highest at
midline derivations and were paralleled by hemodynamic
changes monitored using near-infrared spectroscopy (17).
The results described above are in agreement with early works
(50–52, see also Ref. 53) that have demonstrated large shifts in the
potential of brain tissue that correlate strikingly well with
changes in cerebral blood flow and/or CO2/pH and are generated
across the BBB. Such nonneuronal changes in the potential of
brain parenchyma are readily conducted to scalp, which fully
accounts for the large DC shifts seen in FbEEG recordings (2).
The fact that slow EEG signals can arise at the level of the
BBB has a number of implications. For instance, it is easy to
envisage that epileptic activity, known to cause a large change in
local blood flow (54), may produce a DC shift (see above) that
is at least to some extent generated by the BBB in the region
associated with neuronal hyperactivity. Also, slow signals
related to cognitive paradigms may be prone to “contamina-
tion” by BBB-generated potentials that arise in response to
uncontrolled changes in respiration and systemic CO2 during
the cognitive tasks (see below) (2).
Chapter 36 ■ Infraslow EEG Activity 745
In light of the available data, it is clear that much more work
is needed to elucidate the relative contributions and interactions
of nonneuronal and neuronal/glial generators to DC shifts seen
in both invasive and noninvasive recordings of brain activity.
SLOW EEG ACTIVITY, COGNITIVE STATES,
AND EVENT-RELATED POTENTIALS
During the last decades, much of the research on low-frequency
EEG has focused on slow potentials that are associated with var-
ious kinds of cognitive tasks and states (49,55), such as contin-
gent stimulation (contingent negative variation, CNV), motor
movements (Bereitschaftspotential), and the orienting para-
digm. Subjects may also learn to deliberately induce slow scalp-
recorded potentials, which has been used in attempts to
construct functional brain–computer interfaces (56). All these
potentials have a duration of up to several seconds and often an
amplitude in the order of only a few microvolts, hence requir-
ing an FbEEG as well as electrodes with an optimal DC per-
formance (3) for their accurate recording. In addition, the fact
that nonneuronal EEG signals are readily elicited by changes in
respiration patterns (see section above) or tongue movements
(18) call for caution in the interpretation of slow EEG signals
related to cognitive tasks and preparatory states (2,18).
In addition to slow EEG responses seen in cognitive studies
of the kind referred above, careful analysis of long-latency com-
ponents (up to ~2500 milliseconds) of sensory-evoked poten-
tials may possess significant potential for clinical monitoring of
brain functions. For instance, it has been shown that long-
latency components of auditory-evoked potential correlate
nicely with deepening of anesthesia as well as with emergence
from anesthesia (57), thus raising the possibility that they could
be effectively utilized in anesthesia monitoring. In addition,
large slow potential shifts, which may share a number of char-
acteristics with the ISOs described above, have also been
reported during anesthesia (58).
OTHER INFRASLOW MECHANISMS
The evidence for the presence of infraslow brain activity does
not only come from direct demonstration of infraslow EEG sig-
nals. Several lines of indirect evidence have recently emerged to
support the idea that an underlying infraslow modulation is a
widespread mechanism in brain functions. Many different phys-
iological and pathological EEG activities exhibit infraslow peri-
odicity in humans (7,8,28,59–61) as well as in animals (29,62).
In addition, fundamental studies on temporal dynamics of
human cognition have shown that cognitive performance (such
as reaction time) fluctuates at infraslow frequencies (63), and,
strikingly, this fluctuation in performance is tightly linked to the
temporal phase of an underlying, scalp-recorded infraslow EEG
activity (8) (Fig. 36.5). These observations are not only theoret-
ically interesting, but they might provide completely novel clin-
ical paradigms in diagnostics of patients, such as attention
deficit disorders, where infraslow mechanisms underlying cog-
nitive processes are disturbed (64). Finally, widespread mecha-
nisms of infraslow brain activity are implicated by the presence
746 Part V ■ Complementary and Special Techniques
of the so-called resting state (or default mode) networks that are
shown to produce distinct, infraslow activity in magnetic reso-
nance imaging (32,65,66).
CONCLUSIONS
It is obvious that FbEEG with its capability to record slow activ-
ity of the human brain has a wide range of potential applica-
tions in both basic science and in the clinic. Above, we have
shown that FbEEG is mandatory for faithful, nondistorted, and
nonattenuated recording of the most salient aspects of EEG
activity seen under a wide variety of circumstances. These
include recordings of immature human cortical activity and
epileptic seizures as well as EEG studies of sleep and of various
kinds of cognitive tasks and states.
It is likely that the scope of FbEEG measurements will
strongly expand in the near future. For instance, animal studies
have shown that large waves of spreading depression take place
in ischemic brain tissue, giving rise to high-amplitude, infra-
slow electrical responses (67). It is likely that such events can be
readily measured from the human scalp during early phases of
ischemia. It will also be of interest to use FbEEG in a reexami-
nation of the hitherto unresolved question whether slow DC
shifts related to cortical spreading depression take place during
migraine attacks in humans (68–71).
The technical problems related to FbEEG recording of slow
activity can be readily solved, as described above. Hence, it is
likely that FbEEG will become a standard technique in various
fields of research including neurophysiology, neurology as well
as experimental psychology.
ACKNOWLEDGMENTS
Our research is supported by the Academy of Finland, by the
Sigrid Jusélius Foundation, by Arvo and Lea Ylppö Foundation,
and by European IP “EPICURE” EFP6-037315 (KK).
REFERENCES
1. Bauer H, Korunka C, Leodolter M. Technical requirements for
high-quality scalp DC recordings. Electroencephalogr Clin
Neurophysiol. 1989;72:545–547.
2. Voipio J, Tallgren P, Heinonen E, et al. Millivolt-scale DC shifts in
the human scalp EEG: evidence for a nonneuronal generator.
J Neurophysiol. 2003;89:2208–2214.
3. Tallgren P, Vanhatalo S, Kaila K, et al. Evaluation of commercially
available electrodes and gels for recording of slow EEG potentials.
Clin Neurophysiol. 2005;116:799–806.
4. Vanhatalo S, Voipio J, Kaila K. Full-Band EEG (FbEEG): an emerg-
ing standard in electroencephalography. Clin Neurophys.
2005;116:1–8.
5. Vanhatalo S, Voipio J, Kaila K. Full-Band EEG (FbEEG): a new
standard in clinical electroencephalography. Clin EEG and
Neurosci. 2005;36:311–317.
6. Bullock TH, Mcclune MC, Enright JT. Are the electroencephalo-
grams mainly rhythmic? Assessment of periodicity in wide-band
time series. Neuroscience. 2003;121:233–252.
7. Vanhatalo S, Palva JM, Holmes MD, et al. Infraslow oscillations
modulate interictal epileptic activity in the human cortex during
sleep. Proc Natl Acad Sci USA. 2004;191:5053–5057.
8. Monto S, Palva S, Voipio J, et al. Very slow EEG fluctuations pre-
dict the dynamics of stimulus detection and oscillation amplitudes
in humans. J Neurosci. 2008;28:8268–8272.
9. Picton TW, Hillyard SA. Cephalic skin potentials in electroence-
phalography. Electroencephalogr. Clin Neurophysiol. 1972;33:
419–424.
10. Julkunen P, Pääkkönen A, Hukkanen T, et al. Efficient reduction of
stimulus artifact in TMS-EEG by epithelial short circuiting by
mini-punctures Clin Neurophys. 2008;119:475–481.
11. Miller JW, Kim W, Holmes MD, et al. Ictal localization by source
analysis of infraslow activity in DC-coupled scalp EEG recordings.
NeuroImage. 2007;35:583–597.
12. Vanhatalo S, Tallgren P, Andersson S, et al. DC-EEG unmasks very
slow activity patterns during quiet sleep in preterm infants. Clin
Neurophysiol. 2002;113:1822–1825.
13. Vanhatalo S, Metsäranta M, Andersson S. High fidelity recording
of brain activity in the extremely preterm babies: feasibility study
in the incubator. Clin Neurophysiol. 2008;119:439–445.
14. Vanhatalo S, Kaila K. Generation of ‘positive slow waves’ in the
preterm EEG: by the brain or by the EEG setup? Clin Neurophysiol.
2008;119:1453–1454.
15. Vanhatalo S, Palva M, Andersson S, et al. Slow endogenous activity
transients and developmental expression of K-Cl cotransporter 2 in
the immature human cortex. Eur J Neurosci. 2005c;22:2799–2804.
16. Vanhatalo S, Holmes MD, Tallgren P, et al. Very slow EEG
responses lateralize temporal lobe seizures: an evaluation of non-
invasive DC-EEG. Neurology. 2003; 60:1098–1102.
17. Vanhatalo S, Tallgren P, Becker C, et al. Scalp-recorded slow EEG
responses generated in response to hemodynamic changes in the
human brain. Clin Neurophysiol. 2003;114:1744–1754.
18. Vanhatalo S, Voipio J, Dewaraja A, et al. Topography and elimina-
tion of slow EEG responses related to tongue movements.
NeuroImage. 2003;20:1419–1423.
19. Vanhatalo S, Kaila K. Spontaneous and evoked activity in the early
human brain. In: Lagercrantz H, Hanson M, Evrard P, et al., ed.
The Newborn Brain: Neuroscience & Clinical Applications. 2nd ed.
New York: Cambridge University Press; 2009: 229–243.
20. Watanabe K, Hayakawa F, Okumura A. Neonatal EEG: a powerful
tool in the assessment of brain damage in preterm infants. Brain
Dev. 1999;21:361–372.
21. Lamblin MD, Andre M, Challamel MJ, et al. Electroencephalography
of the premature and term newborn. Developmental features and
glossary. Neurophysiol Clin. 1999;29:123–219.
22. Vanhatalo S, Kaila K. Ontogenesis of EEG activity: from phenom-
enology to physiology. Semin Fetal Neonatal Med.2006;11:
471–478.
23. Sipilä ST, Kaila K. GABAergic Control of CA3-driven network
events in the developing hippocampus. In: Darlison M, ed.
Inhibitory Regulation of Excitatory Neurotransmission. Results
Probl Cell Differ. Heidelberg: Springer-Verlag; 2008;44:99–121.
24. Vanhatalo S, Lauronen L. Neonatal SEP: back to bedside with basic
science. Semin Fetal Neonatal Med. 2006; 11:464–470.
25. Milh M, Kaminska A, Huon C, et al. Rapid cortical oscillations and
early motor activity in premature human neonate. Cereb Cortex.
2007;17:1582–1594.
26. Penn AA, Shatz CJ. Brain waves and brain wiring: the role of
endogenous and sensory-driven neural activity in development.
Pediatr Res. 1999;45:447–458.

27. Penttonen M, Nurminen N, Miettinen R, et al. Ultra-slow oscilla-
tion (0.025 Hz) triggers hippocampal afterdischarges in Wistar
rats. Neuroscience. 1999;94:735–743.
28. Prince DA. Cyclical spike driving in chronically isolated cortex.
Epilepsia. 1965;6:226–242.
29. Leopold DA, Murayama Y, Logothetis NK. Very slow activity fluc-
tuations in monkey visual cortex: implications for functional brain
imaging. Cereb. Cortex. 2003;13:422–433.
30. Marshall L, Molle M, Fehm HL, et al. Scalp recorded direct current
brain potentials during human sleep. Eur J Neurosci.
1998;10:1167–1178.
31. Girton DG, Benson KL, Kamiya J. Observation of very slow poten-
tial oscillations in human scalp recordings. Electroencephalogr Clin
Neurophysiol. 1973;35:561–568.
32. Mantini D, Perrucci MG, Del Gratta C, et al. Electrophysiological
signatures of resting state networks in the human brain. Proc Natl
Acad Sci USA. 2007;104:13170–13175.
33. Fox MD, Raichle ME. Spontaneous fluctuations in brain activity
observed with functional magnetic resonance imaging. Nat Rev
Neurosci. 2007;8:700–711.
34. Gumnit RJ, Takahashi T. Changes in direct current activity during
experimental focal seizures. Electroencephalogr Clin Neurophysiol.
1965;19:63–74.
35. Mayanagi Y, Walker AE. DC potentials of temporal lobe seizures in
the monkey. J Neurol. 1975;209:199–215.
36. Caspers H, Speckmann E-J, Lehmenkuhler A. DC potentials of the
cerebral cortex: seizure activity and changes in gas pressures. Rev
Physiol Biochem Pharmacol. 1987;106:127–178.
37. Nita DA, Vanhatalo S, Lafortune FD, et al. Nonneuronal origin of
CO 2-related DC EEG shifts: an in vivo study in the cat.
J Neurophysiol. 2004;92:1011–1022.
38. Goldring S. Negative steady potential shifts which lead to seizure
discharge. UCLA Forum Sci. 1963;1:215–236.
39. Wieser HG, Elger CE, Stodieck SR. The ‘foramen ovale electrode’:
a new recording method for the preoperative evaluation of
patients suffering from mesio-basal temporal lobe epilepsy.
Electroencephalogr Clin Neurophysiol. 1985;61:314–322.
40. Bragin A, Claeys P, Vonck K, et al. Analysis of initial slow waves
(ISWs) at the seizure onset in patients with drug resistant tempo-
ral lobe epilepsy. Epilepsia. 2007;48:1883–1894.
41. Bates JAV. The unidirectional potential changes in petit mal
epilepsy. UCLA Forum Sci. 1963;1:237–279.
42. Cohn R. DC recordings of paroxysmal disorders in man.
Electroencephalogr Clin Neurophysiol. 1964;17:17–24.
43. Chatrian GE, Somasundaram M, Tassinari CA. DC changes
recorded transcranially during “typical” three per second spike
and wave discharges in man. Epilepsia. 1968;9:185–209.
44. Ikeda A, Taki W, Kunieda T, et al. Focal ictal direct current shifts in
human epilepsy as studied by subdural and scalp recording. Brain.
1999;122:827–838.
45. Gross DW, Gotman J, Quesney LF, et al. Intracranial EEG with very
low frequency activity fails to demonstrate an advantage over con-
ventional recordings. Epilepsia. 1999;40:891–898.
46. Rodin E, Modur P. Ictal intracranial infraslow EEG activity. Clin
Neurophysiol. 2008;119:2188–2200.
47. Vanhatalo S, Voipio J, Kaila K, et al. Use of Full-band EEG for non-
invasive ictal localization. In: Miller JW, Silbergeld DL, eds.
Epilepsy Surgery, Principles and Controversies. New York: Informa
Healthcare, Taylor & Francis Group; 2006:745–751.
48. Lagerlund TD, Gross RA. DC-EEG recording: a paradigm shift in
seizure localization? Neurology. 2003;60:1062–1063.
Chapter 36 ■ Infraslow EEG Activity 747
49. Birbaumer N, Elbert T, Canavan AG, et al. Slow potentials of the
cerebral cortex and behavior. Physiol Rev. 1990;70:1–41.
50. Besson JM, Woody CD, Aleonard P, et al. Correlations of brain d-
c shifts with changes in cerebral blood flow. Am J Physiol.
1970;218:284–291.
51. Woody CD, Marshall WH, Besson JM, et al. Brain potential shift
with respiratory acidosis in the cat and monkey. Am J Physiol.
1970;218:275–283.
52. Sorensen E, Olesen J, Rask-Madsen J, et al. The electrical potential
difference and impedance between CSF and blood in unanes-
thetized man. Scand J Clin Lab Invest. 1978;38:203–207.
53. Tétrault S, Chever O, Sik A, et al. Opening of the blood-brain bar-
rier during isoflurane anaesthesia. Eur J Neurosci. 2008;28:
1330–1341.
54. Duncan JS. Imaging and epilepsy. Brain. 1997;120:339–377.
55. Cui RQ, Huter D, Egkher A, et al. High resolution DC-EEG map-
ping of the Bereitschaftspotential preceding simple or complex
bimanual sequential finger movement. Exp Brain Res. 2000;
134:49–57.
56. Hinterberger T, Kubler A, Kaiser J, et al. A brain-computer inter-
face (BCI) for the locked-in: comparison of different EEG classifi-
cations for the thought translation device. Clin Neurophysiol.
2003;114:416–425.
57. Fitzgerald RD, Lamm C, Oczenski W, et al. Direct current auditory
evoked potentials during wakefulness, anesthesia, and emergence
from anesthesia. Anesth Analg. 2001;92:154–160.
58. Roughan JV, Laming PR. Large slow potential shifts occur during
halothane anaesthesia in gerbils. J Comp Physiol. 1998;182:
839–848.
59. Bullock TH, McClune MC, Achimowicz JZ, et al. Temporal fluctu-
ations in coherence of brain waves. Proc Natl Acad Sci USA.
1995;92:11568–11572.
60. Parrino L, Halasz P, Tassinari CA, et al. CAP, epilepsy and motor
events during sleep: the unifying role of arousal. Sleep Med Rev.
2006;10:267–285.
61. Nir Y, Mukamel R, Dinstein I, et al. Interhemispheric correlations
of slow spontaneous neuronal fluctuations revealed in human sen-
sory cortex. Nature Neurosci. 2008;11:1100–1108.
62. Wichmann T, Kliem MA, Soares J. Slow oscillatory discharge in the
primate basal ganglia. J Neurophysiol. 2002;87:1145–1148.
63. Gilden DL. Cognitive emissions of 1/f noise. Psychol Rev.
2001;108:33–56.
64. Castellanos FX, Sonuga-Barke EJ, Scheres A, et al. Varieties of
attention-deficit/hyperactivity disorder-related intra-individual
variability. Biol Psychiatry. 2005;57:1416–1423.
65. Fransson P, Skiöld B, Horsch S, et al. Resting-state networks in the
infant brain. Proc Natl Acad Sci USA. 2007;104:15531–15536.
66. Raichle ME, Snyder AZ. A default mode of brain function: a brief
history of an evolving idea. NeuroImage. 2007;37:1083–1090.
67. Nedergaard M. Mechanisms of brain damage in focal cerebral
ischemia. Acta Neurol Scand. 1988;77:81–101.
68. Lauritzen M. Cortical spreading depression in migraine.
Cephalalgia. 2001;21:757–760.
69. Eikermann-Haerter K, Ayata C. Cortical spreading depression and
migraine. Curr Neurol Neurosci Rep. 2010;10:167-173.
70. James MF, Smith JM, Boniface SJ, et al. Cortical spreading depres-
sion and migraine: new insights from imaging? Trends Neurosci.
2001;24:266–271.
71. Rogawski MA. Common pathophysiologic mechanisms in
migraine and epilepsy. Arch Neurol. 2008;65:709–714.

High-Frequency EEG Activity
INTRODUCTION
The electroencephalogram (EEG) is traditionally recorded and
clinically interpreted with a low-pass filter set at 60 or 70 Hz. The
most common EEG activities are divided into the delta, theta,
alpha, and beta bands, the latter term usually used for activity
extending up to ~30 Hz. There has been much interest recently in
the gamma band, a term commonly used for activity faster than
beta ( 30 Hz). Because of considerable variability in the fre-
quency bounds of gamma band activity reported across different
studies, it may be premature, or even altogether inappropriate to
select a specific frequency range to define it. Nevertheless, there is
a practical need for terminological consistency in order to facili-
tate communication in both scientific discourse and translation
to clinical practice. Based on accumulating evidence from studies
of both physiologic and pathologic responses in gamma frequen-
cies (see below), it may be useful to arbitrarily, but provisionally
define the gamma band as 30 to 70 Hz. It has recently been dis-
covered that frequencies higher than 70 Hz can also be recorded
and can have both physiologic and pathologic significance. A
variety of terms have been used to describe physiologic EEG
activity above 70 Hz, for example, “high-gamma” (1,2) and the
chi-band (3). In the context of epilepsy, pathologic patterns at
high frequencies have also been described, for example: ripples
between 100 and 200 Hz (sometimes between 80 and 160 Hz,
sometimes between 100 and 250 Hz), fast ripples between 160,
200, or 250 Hz and 500 or 600 Hz. These events have other prop-
erties besides their frequency, and in general, we think that fre-
quency bands should be defined independently from specific
EEG phenomena at a time when much exploration is taking
place. Rather than using a name, we will therefore refer to bands
by their boundaries (e.g., the 100- to 250-Hz band).
The Skull, the Generator Volume,
and the Electrodes
It is often said and sometimes written that “the skull filters out
high frequencies.” This would imply that high frequencies, for
instance above 60 Hz, would be more attenuated by the skull
than lower frequencies, making it essentially impossible to
record high frequencies except with intracranial electrodes or
with magnetoencephalography (MEG) that is not sensitive to
skull conductivity. In fact the skull does not attenuate any fre-
quency until approximately 10,000 Hz. This has been clearly
demonstrated in the excellent study of Oostendorp et al. (4).
There seems to be, however, an inverse relationship between the
frequency of an EEG rhythm and the volume of brain that gen-
erates it. For instance, very large regions can generate synchro-
CHAPTER
JEAN GOTMAN, NATHAN E. CRONE 37
nous delta activity, and smaller regions can generate synchro-
nous higher frequencies. As very high frequencies are generated
in a very small volume, they are unlikely to be “seen” from the
scalp, simply because of the small angle under which they would
be seen (5). For the same reason (small generator volume), high
frequencies usually have low amplitude and the skull attenua-
tion reduces this amplitude further. So it is indeed unlikely that
high frequencies can be seen from the scalp, but it is not because
they are filtered by the skull; it is because of the characteristics of
their generator. For these reasons (distance from generator to
sensor and small amplitude of signal), it may be just as difficult
to record high frequencies with MEG as it is with scalp EEG.
In the context of a small generator volume, the size of
recording electrodes becomes a potentially important factor,
although the question of electrode size has rarely been
addressed in the field of intracranial EEG (iEEG). If a generator
is on the order of a few hundred microns, two issues can be
raised. The first is that of spatial sampling: if such small gener-
ators are scattered and present in a small number of locations,
the likelihood that an electrode would be in their vicinity is low.
This is one extreme of the general problem of poor spatial sam-
pling of intracranial investigations. The second one is that an
electrode larger than the field generator and in contact with it
could significantly alter that field, and an electrode much larger
than the field generator could attenuate it, possibly to the point
of making it undetectable. It is very difficult to know the size of
a generator, unless one performs a systematic exploration with
microelectrodes (6). It is therefore not easy to determine the
optimal electrode size. Electrodes have usually been designed to
allow electrical stimulation without an excessive current density
(7,8); for this purpose, a relatively large contact surface is
important but this requirement may be in conflict with that of
optimizing the recording of high-frequency activity.
It may also be useful to discuss some terminology issues. In
the animal experimentation domain, the notion of “local field
potential” is often used. For instance filtering an extracellular
microelectrode recording such that only activity below 100 Hz,
or below 500 Hz, remains (i.e., removing action potentials), is
said to produce the local field potential (LFP). This extracellular
potential fluctuation is not easily differentiated from the EEG,
as recorded locally with a small electrode. Although the term
“EEG” is most often associated with relatively large electrodes,
there is nothing in the definition of the EEG that implies elec-
trode size. It is rather defined as potential fluctuations within a
certain frequency range. From our point of view, there is no for-
mal difference between LFP and EEG, but the terms tend to be
used in different contexts.
749
750 Part V ■ Complementary and Special Techniques
We will first discuss the high frequencies that are recorded in
the context of normal function. These have been documented
in experimental animals as well as in humans in response to
specific sensory, motor, or cognitive activities. In humans, some
of these high frequencies can be recorded from the scalp but
most were investigated in patients with intracerebral electrodes
in the context of the evaluation of the surgical treatment of
their medically refractory epilepsy. We will then discuss the
high-frequency activities that have been related to focal
epilepsy, here again investigated in experimental animals and in
patients with intracerebral electrodes.
HIGH-FREQUENCY ACTIVITY
AND NORMAL CORTICAL FUNCTION
Experimental Background and Significance
The functional relevance of high-frequency EEG activity in nor-
mal brain physiology has been the subject of many investigations
over the past several decades in both animals and humans and is
still hotly debated. Some of the earliest studies using EEG recog-
nized that cortical activation was associated with the disappear-
ance of low-frequency activity (alpha) and the appearance of fast
activity, though the extent of this fast activity was difficult to
ascertain given the technical limitations of existing equipment
(9–11). Subsequent microelectrode recordings in rabbit olfactory
cortex (12) showed that different odors elicited different spatial
patterns in the amplitude of sniff-triggered gamma oscillations
(38 to 80 Hz), suggesting a mechanistic role in stimulus discrim-
ination. Similar studies in cat and rat revealed oscillations with a
similar frequency range (13). A decade later, local field recordings
in cat visual cortex found that oscillatory neuronal firing in a fre-
quency range of 40 to 60 Hz can become synchronized during
visual stimulation between spatially separate columns (14),
between areas 17 and 18 (15), and even between areas 17 of the
two hemispheres (16), and that this synchronization depends on
global stimulus properties. Similar activity, ranging up to 35 and
50 Hz was recorded in monkey sensorimotor cortex during vol-
untary hand movements (17,18). In these studies bursts of single
unit activity were commonly synchronized with LFP oscillations,
suggesting that these oscillations facilitated and/or were facili-
tated by the synchronization of neuronal firing between spatially
segregated but functionally related neurons.
Synchronization of neural firing has been hypothesized to
form the basis for temporal coding by which temporary assem-
blies of neurons represent higher-order, or global, stimulus
properties (19). This mechanism of temporal coding (20) was
recognized as an exciting potential solution to the binding prob-
lem associated with psychophysiologic phenomena such as sen-
sory segmentation and invariant object recognition. More recent
theoretical refinements have further proposed that subthreshold
gamma oscillations, driven by synchronous inhibitory neuron
firing, produce temporal windows for cortical–cortical commu-
nication, whereby coherently oscillating ensembles of neurons
interact more effectively because their communication windows
are open at the same time (21). Another theory proposes that
rhythmic gamma oscillations from inhibitory interneuronal fir-
ing interact with excitatory inputs to pyramidal cells such that
more excited cells fire earlier in the gamma cycle. This recording
of the amplitude of excitatory drive into phase values relative to
the gamma cycle would provide a more efficient coding mecha-
nism by enabling the readout of amplitude information within
a single gamma cycle without requiring the temporal integration
of firing rates (22).
Band-Limited Gamma Oscillations
A common feature of the aforementioned theories relating high-
frequency EEG (gamma) activity to neural coding and cortical
function is that they all more or less require rhythmic oscilla-
tions of membrane potentials and/or action potentials at rela-
tively stable, that is, band-limited, gamma frequencies. Although
synchronous neural activity is not necessarily oscillatory and
temporal coding could conceivably exist without oscillations,
the explanatory potential of band-limited gamma oscillations
has been compelling (23). Perhaps inspired by this, many inves-
tigators have attempted to correlate band-limited gamma oscil-
lations in human EEG with cognitive and perceptual operations
in cortex.
Early attempts to relate band-limited gamma oscillations to
human cortical function were designed to demonstrate task-
specific lateralization of 40-Hz activity in scalp-recorded EEG
during long blocks of task performance. Although compensa-
tions had to be made for 40-Hz EMG artifact, these studies did
show lateralization of 40 Hz to the left hemisphere during ver-
bal tasks (24), and to contralateral central scalp region during
both simple and complex choice reaction time tasks (25,26).
However, like block designs in fMRI experiments, estimates of
gamma activity in long temporal segments during both resting
and activated conditions can be contaminated with extraneous
behavioral fluctuations. Furthermore, this block-design
approach does not exploit the fine temporal scale at which
gamma activity can change during task performance. For these
reasons, event-related designs have been used much more
extensively when studying gamma responses.
Phase-Locked Gamma Responses
Using event-related analyses of MEG, Pantev et al. (27)
observed band-limited gamma responses to tone bursts at 35 to
40 Hz. This was consistent with previous reports of 40-Hz clicks
driving large steady-state evoked responses in auditory cortex
(28). Notably, the gamma-band responses (GBRs) observed by
Pantev et al. (27) were obtained by filtering event-related poten-
tials (ERPs) produced by averaging raw signal from many indi-
vidual trials in the time domain. This procedure reveals
phase-locked evoked GBR, consisting of gamma components
in the ERP. It is important to distinguish these evoked GBRs
from induced gamma band activity that is time-locked, but not
phase-locked (29) (Fig. 37.1; see the section “Non–Phase-Locked
Gamma Responses”).
Phase-locked gamma responses have been investigated in a
variety of experimental contexts. In many cases narrow band-
pass filters have been used to focus on ERP components in and
around 40 Hz. This practice is potentially susceptible to arti-
facts from filtering ERP impulses containing a wide range of
 Chapter 37 ■ High-Frequency EEG Activity 751

Figure 37.1 Signal analysis of phase-locked and non–phase-locked responses using band-pass filtering. Schematic of sig-
nal analyses for one channel of intracranial EEG recorded over dominant (left) superior temporal gyrus (filled circle on
brain) during an auditory speech discrimination task. To minimize the contribution of phase-locked responses to subse-
quent analyses of non–phase-locked responses, the ERP may be subtracted from the raw EEG signal in each individual trial
(depicted in steps A, B, and C; for alternative approaches, see Kalcher et al. (42) and Trautner et al. (69)). In this approach,
the raw signal (A) is averaged across N trials to obtain the event-related potential (ERP, B), and the ERP is then subtracted
from each individual trial (C) prior to band-pass filtering (D, 80 to 100 Hz in this illustration). Band-pass filtered signals
from each individual trial are then squared to obtain power values (E). These band-specific power estimates, which are all
positive, may then be averaged across trials to obtain the power average (F), or submitted to statistical analyses to calcu-
late the % change in poststimulus power from prestimulus baseline power (G). Event-related power increases (F,G) are
also known as induced responses and are dominated by non–phase-locked response components. Note the variability in
the latency and magnitude of these responses across individual trials (E). (From Crone NE, Boatman D, Gordon B, et al.
Induced electrocorticographic gamma activity during auditory perception. Clin Neurophysiol. 2001;112(4):565–582.)

frequencies. Nevertheless, evoked GBRs have been observed in
both animals and humans during cortical activation in a vari-
ety of functional–anatomic domains (30–32), and in humans
this class of responses has been used to explore mind–brain
relationships (33,34).
In addition to ERP components in the gamma band, ERP
components have also been observed in even higher frequen-
cies, for example, ultrafast frequencies (400 to 1000 Hz), in ani-
mals and humans. For example, both EEG and MEG recordings
(Fig. 37.2) have shown that somatosensory evoked potentials
from electrical stimulation of the median nerve contain a brief
(10 to 15 msec) burst of ~600-Hz spikelike wavelets, called
sigma bursts (35–37). Sigma bursts temporally overlap the thal-
amic P15 component and the N20 primary cortical response
and, based on converging evidence from animals and humans,
are thought to represent both far-field and near-field potentials
generated by highly synchronized population spikes in
cuneothalamic and thalamocortical relay cells and a variety of
cortical neurons. These sigma bursts are diminished during
sleep, particularly slow-wave sleep, though the N20 is not (38).
EEG wavelet bursts at similar frequencies have been observed in
animals (39). In awake monkeys (40), for example, a subset of
SI units exhibits bursts of spikes and single spikes that are
phase-locked to surface EEG wavelets elicited by electrical
median nerve stimuli (Fig. 37.3). Because many central
synapses transmit bursts but filter out single spikes, coincident
spike bursts from thalamocortical relay neurons may drive
postsynaptic cortical cells more efficiently than single spikes.
Furthermore, they may also be capable of richer coding because
intraburst spike frequency can convey graded information
about sensory stimuli while bursts of different durations can
code different stimulus features.
Non–Phase-Locked Gamma Responses
Averaging electrophysiologic (EEG, MEG, LFP, etc.) responses
in the time domain extracts phase-locked signal components as
ERPs and discards non–phase-locked components as noise. In
contrast, averaging these responses in the frequency domain
focuses on event-related changes in the spectral energy of
responses that may have both phase-locked and non–phase-
locked components (41). The latter approach is still temporally
linked to an event across multiple trials, but the result is not
limited to phase-locked components, as in ERPs. Different
methods have been used to recognize and minimize the contri-
bution of phase-locked components (ERPs) to analyses of
non–phase-locked electrophysiologic responses (42,43).
752 Part V ■ Complementary and Special Techniques
Figure 37.2 Simultaneous magnetoencephalogram/ electroencephalo-
gram (MEG/ EEG) recordings of somatosensory responses (8000 aver-
ages) after electrical left median nerve stimulation. Somatosensory
evoked potentials (SEPs) (traces A/ B) recorded with a C4 –C3 derivation
(negativity at C4 upward). MEG somatically evoked fields (SEF) (traces
C/ D) were recorded at the N20 field maximum over the lower part of the
right central sulcus (upward deflections: magnetic field lines entering the
head). A/ D: Original records (0.5 to 1500 Hz). B/ C: High-pass-filtered
versions ( 430 Hz). The thin line at 21 msec demonstrates the peak
latency preservation before versus after digital filtering. The solid line to
the left indicates median nerve electrical stimulation (artifacts clipped).
The MEG sigma burst (C) is shorter than the simultaneously recorded
EEG burst (B), which starts earlier and lasts longer. While MEG detects
activity mainly from tangential generators (e.g., area 3b), early EEG burst
wavelets correspond to radially oriented thalamocortical afferences, and
later EEG peaks reflect radially oriented sources in area 1 at the crown of
the postcentral gyrus. (Modified from Curio G. Ultrafast EEG activities. In:
Niedermeyer E, Lopes da Silva F, eds. Electroencephalography: Basic
Principles, Clinical Applications, and Related Fields. 5th ed. Philadelphia,
PA: Lippincott, Williams & Wilkins; 2005:495–504.)
However, the distinction between phase-locked responses
(ERPs) and non–phase-locked increases in signal energy (often
termed “induced” responses) may depend more on the meth-
ods by which they are extracted, than on fundamental differ-
ences between their generators.
Averaging in the time domain necessarily yields phase-locked
responses. However, significant variability (jitter) in the latency
(or phase) of ERPs can distort their appearance in time-averaged
responses. High-frequency components of electrophysiologic
responses are more susceptible to latency jitter, and there is
Figure 37.3 Recordings from single units and epidural EEG in S-I of an
awake monkey. Peristimulus time histograms (PSTHs) summed over 15
single units (thick lines, upper right panel) coincide with the middle
part of epidural EEG burst averages (thin lines); middle right panel
depicts the results of high-pass filtering ( 428 Hz) the upper panel
traces. (Modified from Baker SN, Curio G, Lemon RN. EEG oscillations
at 600 Hz are macroscopic markers for cortical spike bursts. J Physiol.
2003;550(pt 2):529–534.)
usually more jitter of these responses (and their corresponding
cognitive processes) at longer latencies. This probably explains
why ERPs with high-frequency components (e.g., the sigma
bursts described above) are usually confined to early latencies.
On the other hand, ERPs at longer latencies (e.g., P300) usually
consist of low-frequency components that are more resistant
to jitter (44,45). Because averaging in the frequency domain
does not require phase-locking, it may be better suited to inves-
tigate cortical processing at longer or more variable latencies and
to investigate high-frequency electrophysiologic responses at
longer latencies. It may therefore be useful to conceive of electro-
physiologic responses as having different combinations of
frequencies, latencies, and phase-locking. Nevertheless, the dis-
tinction between phase-locked and non–phase-locked responses
is often still a practical one. Furthermore, many studies suggest
that these different classes of EEG responses may have distinct
functional response properties (29,46,47).
Although early studies of gamma activity using block
designs, strictly speaking, did relate non–phase-locked gamma
activity to human cortical function, one of the first event-
related analyses of this activity, that is, on a temporal scale com-
mensurate with task performance, was made by Pfurtscheller et
al. (48), who demonstrated a lateralized increase in 40-Hz activ-
ity during self-paced finger movements in three subjects. One
of these subjects was later shown to have a somatotopic pattern
of this activity during movements of the tongue, fingers, and
toes (49). Similar non–phase-locked gamma activity has also
been observed in a variety of experimental contexts, for exam-
ple, during the perception of illusory triangles (50).

Broadband “High Gamma” Activity
As mentioned above, most noninvasive studies of gamma activity
vis-à-vis normal cortical function, until recently, have focused on
relatively narrow band-limited gamma frequencies in and around
40 Hz. These studies were motivated in part by previous animal
studies of band-limited gamma oscillations and by the aforemen-
tioned theories of cortical coding that necessitate rhythmic oscil-
latory activity at consistent, narrowband gamma frequencies.
iEEG recordings in patients undergoing epilepsy surgery, however,
have allowed the investigation of physiologic gamma responses
across a broader range of frequencies than those studied in
humans by noninvasive methods. These studies have revealed
event-related, non–phase-locked responses in gamma frequencies
higher than those previously studied noninvasively (2). In addi-
tion, these “high gamma” responses have had a characteristically
broadband spectral profile that spans a large range of frequencies.
The lower and upper frequency boundaries of these responses
have been quite variable, but they have most commonly ranged
from ~60 to ~200 Hz, with the majority of event-related energy
changes occurring between 80 and 150 Hz (51,52).
Across a variety of iEEG studies broadband high gamma
responses have exhibited functional response properties that dis-
tinguish them from both phase-locked responses (ERPs) and
non–phase-locked responses previously observed with scalp EEG
in other frequency bands (2). For example, the temporal and spa-
tial distributions of these responses are typically more discrete
and/or functionally specific for task-related cortical activation
than other electrophysiologic responses. Furthermore, broad-
band high gamma responses have been demonstrated in a great
variety of functional–anatomic domains, including motor cortex
(1,53–61), frontal eye fields (62), somatosensory cortex (63–66),
auditory cortex (43,67–69), visual cortex (70–74), olfactory
cortex (75), and language cortex (74,76–79). The functional–
anatomic ubiquity of these responses suggests that they may
serve as a general electrophysiologic index of cortical processing.
Motor Cortex
Investigations of broadband high gamma responses in motor
cortex have provided many insights into their functional
response properties. In a study using a visually cued motor task
(1,80) power augmentation was observed from 75 to 100 Hz
within somatotopically defined regions of sensorimotor cortex
contralateral to the cued movement (1) and corresponded well to
electrocortical stimulation maps of motor function. The tempo-
ral patterns of these responses were also quite specific, occurring
in brief bursts limited to the onset and offset (81) of movement,
with latencies that covaried with the latencies of movement
onset/offset. In another study using self-paced finger and wrist
movements, Ohara et al. (53) observed non–phase-locked
gamma activity in S1 and M1 extending up to 90 Hz. High
gamma activity (60 to 90 Hz) in particular was time-locked to
movement onset and was short-lived after it. Pfurtscheller et al.
(54) also observed broadband high gamma activity (60 to 90 Hz)
over sensorimotor cortex while subjects performed self-paced
tongue and finger movements. As in the previous studies, the
topographic pattern of high gamma activity was more discrete
Chapter 37 ■ High-Frequency EEG Activity 753
and somatotopically specific than the more widespread power
changes in mu (alpha) and beta bands, and its temporal pattern
was also briefer, corresponding to movement onset.
In the largest series to date (22 subjects) of iEEG recordings
in sensorimotor cortex, Miller et al. (56) compared the topo-
graphic patterns of power suppression in low frequencies (8 to
32 Hz) versus power augmentation in high gamma frequencies
(76 to 100 Hz) during a variety of motor tasks. As in the previ-
ously mentioned studies, high gamma responses had a more
focused spatial distribution than did power suppression in low
frequencies. In addition, these responses had a somatotopic
organization corresponding to the movement of different body
parts and corresponded well to the results of electrocortical
stimulation mapping. This and other reports (57,82,83) suggest
that high-frequency responses may be a useful tool for mapping
motor function in patients undergoing surgical resections near
or within motor cortex.
Language and Auditory Cortices
Broadband high gamma responses have also been used to map
language cortex (74,76–79). In one of these studies (Fig. 37.4),
the spatiotemporal patterns of high gamma responses were com-
pared during tasks with different modalities of input (visual vs.
auditory stimuli) and output (spoken vs. signed responses). The
location and timing of high gamma responses during these tasks
were consistent with the functional neuroanatomy of human lan-
guage and with the latencies of verbal and signed responses. In
addition, high gamma responses often, but not always, corre-
sponded to the results of electrocortical stimulation mapping.
The potential clinical utility of high gamma responses for map-
ping language cortex was subsequently studied in 13 patients.
Maps of language cortex based on high gamma responses during
picture naming were compared with maps of naming and mouth
movements (responsible for verbal output) from electrocortical
stimulation (77). With respect to electrocortical stimulation (the
“gold standard”), the calculated specificity of maps based on high
gamma responses was 84% but their sensitivity was only 43%.
The relatively good specificity suggested that high gamma
responses were potentially useful for constructing a preliminary
functional map in order to identify cortical sites of lower priority
for electrocortical stimulation mapping. However, its low sensi-
tivity indicated that it could not yet replace electrocortical stim-
ulation even though stimulation carries the added risk of eliciting
seizures. These results also raise the question as to whether some
high gamma responses are falling between the cracks of the stan-
dard 1-cm spaced electrode arrays. Would smaller, more closely
spaced electrodes capture more of these responses? A previous
study of intracranial gamma responses suggested an optimal
electrode spacing of less than 5 mm (84).
Intracranial EEG recordings of auditory cortex have also
demonstrated high gamma responses during tone and speech
discrimination (43; Fig. 37.5), during an auditory oddball par-
adigm (67), and during an auditory sensory gating (P50) para-
digm (69). Broadband high gamma responses induced by
acoustic stimuli are concentrated over posterior superior tem-
poral gyrus in a spatial distribution similar to but not identical
with the N100 of the auditory evoked response associated with
754 Part V ■ Complementary and Special Techniques

Figure 37.4 Plots of induced high gamma

(80 to 100 Hz) activity in one electrode over
posterior superior temporal gyrus, during
three language tasks using either spoken or
signed responses. In each plot the x-axis is
time in milliseconds after stimulus onset,
and data points occur every 100 msec.
Event-related high gamma responses
(HGRs, geometric mean percentage change
in poststimulus 80 to 100 Hz power from
baseline) are plotted as thick lines; thin
lines above and below represent 95% con-
fidence intervals. HGRs begin 100 to 200
msec after onset of the auditory stimulus
during word repetition (peaks marked “S”).
In all three language tasks, HGRs occurred
later during spoken responses (peaks
marked “R”), but not during signed
responses. The onset of the late HGRs
covaried with the latencies of the patient’s
verbal responses. (From Crone NE, Hao L,
Hart J Jr, et al. Electrocorticographic gamma
activity during word production in spoken
and sign language. Neurology. 2001;57(11):
2045–2053.)

the onset of acoustic stimuli. In addition, the magnitudes of
these responses appear to be correlated with the degree of func-
tional activation, for example, greater magnitude in dominant
superior temporal gyrus during discrimination of speech stim-
uli than during discrimination of tone stimuli. Responses in a
similar frequency range have been observed in microelectrode
recordings of auditory cortex in monkeys (85,86).
Band-Limited versus Broadband
Gamma Responses
Studies in humans and animals have shown that non–phase-
locked responses in the traditional 40-Hz gamma band are
more variable and less sensitive to functional activation of cor-
tex than responses in high gamma frequencies (43,86). This
may be due to variability in the frequency range of event-
related power suppression. If this power suppression extends
into low gamma frequencies, it may mask power augmentation
in this frequency range. This is a potential pitfall for analyses of
responses with narrow band-pass filters at low gamma frequen-
cies, for example, at 40 Hz. Although the lower boundary of
broadband gamma power increases may extend down into
40-Hz frequencies, the most consistent responses occur
above 60 Hz. The greater reliability of power changes in high
gamma frequencies was recently demonstrated in a study
using iEEG to classify movements in different body parts (3).
The limited classification accuracy of lower gamma frequencies
(30 to 70 Hz) was interpreted to result from an overlap in the
power spectrum of band-limited power suppression at lower
frequencies (ranging up to ~50 Hz) and broadband power
increases that included gamma frequencies greater than 70 Hz.
Figure 37.5 Power spectral analyses (PSA) of ECoG recorded during an auditory speech discrimination task. For selected
electrodes (stars) average power spectra during a baseline, or reference, interval ( 750 to 250 msec before stimulus onset)
are compared with average power spectra during auditory stimulation (0 to 500 msec after stimulus onset). For each elec-
trode, the upper plot is produced by subtracting the lower plot of the prestimulus (reference) power spectrum (gray line) from
the lower plot of the poststimulus (active) power spectrum (black line). Negative values in the upper plots indicate power
suppression; positive values indicate power augmentation. At the second site in Subject #1 (lower plots) there is a relative
power increase at 40 Hz during active versus reference intervals. However, at the first site in Subject #1 (upper plots) and the
second site in Subject #2 (lower plots), there is no net increase in ECoG power at 40 Hz. Furthermore, at the first site in Subject
#2 (upper plots) ECoG power is suppressed at 40 Hz. In all four sites, ECoG power is augmented in high gamma frequencies
greater than 80 Hz. This power augmentation is observed over a broad range of frequencies although it is most prominent
between ~ 80 and ~ 150 Hz. Dips and peaks in power spectrum at 60 Hz are from a 60-Hz notch filter (Subject #1) or from
electronic noise and its harmonics (Subject #2), respectively. (From Crone NE, Boatman D, Gordon B, et al. Induced electro-
corticographic gamma activity during auditory perception. Clin Neurophysiol. 2001;112(4):565–582.)

755
756 Part V ■ Complementary and Special Techniques
Broadband high gamma responses recorded with iEEG in
humans are difficult to reconcile with earlier conceptualiza-
tions of band-limited gamma oscillations and their putative
role in neural computation. Although more recent observa-
tions, of higher frequency and broader band oscillations, in
animals are extending the frequency range that is quoted for
“gamma,” gamma responses are still largely conceptualized as
band-limited network oscillations. However, for the broad-
band responses observed with iEEG to arise from band-limited
network oscillations, power changes would have to arise from
summation over multiple spatially overlapping neural popula-
tions or assemblies, each oscillating at different, perhaps over-
lapping or broadly tuned frequencies (1,87).
Generators of Broadband Gamma Responses
An alternative explanation for broadband high gamma
responses is that they are the time–frequency representations of
transient responses with a broad range of frequency compo-
nents. Intertrial jitter in the latency of these transients presum-
ably renders them invisible when averaged across trials in the
time domain. Recent investigations of microelectrode record-
ings from macaque SII cortex during tactile stimulation have
observed broadband high gamma responses with spectral pro-
files identical to those recorded in human iEEG (88,89).
Detailed time–frequency analyses of these responses using
matching pursuits revealed that they are temporally tightly
linked to neuronal spikes though their precise generating mech-
anisms could not be determined (88). In addition, LFP power
in the high gamma range was strongly correlated, both in its
temporal profile and in its trial-by-trial variation, with the fir-
ing rate of the recorded neural population (89).
Whether the underlying signals giving rise to broadband
gamma responses are oscillations or transients, activity in such a
high frequency range is much more likely to be recorded at the
mesoscale of subdural electrocorticography if there is some
degree of synchronization across a population of neural genera-
tors. In a recent simulation of iEEG responses using different fir-
ing patterns in the underlying cortical population, both an
increase in firing rate and an increase in neuronal synchrony
resulted in broadband high gamma power increases. However,
these responses were much more sensitive to increases in neu-
ronal synchrony than to increases in firing rate (89). Thus, broad-
band gamma response may index neuronal synchronization to
some extent even if the underlying firing pattern is not a band-
limited oscillation. Synchronization across subpopulations of
neurons, as mentioned above, has been hypothesized to consti-
tute a temporal coding strategy for cortical computation that
complements rate coding and plays a role in higher cortical func-
tions such as attention (90,91). Several human iEEG studies have
found an augmentation of broadband gamma activity in associ-
ation with attention (52,61,71,73,79,92). Whether these gamma
responses reflect an increase in firing rate or an increase in neu-
ral synchronization, there is ample evidence from the aforemen-
tioned studies that they likely reflect patterns of neural activity
that are relevant to cortical computation and may serve as useful
markers for cortical function mapping.
PATHOLOGIC HIGH-FREQUENCY ACTIVITY
Ripple activity, in the range of 100 to 200 Hz, had been recorded
in the LFPs of hippocampus in normal experimental animals
and, like the gamma responses discussed above, were considered
to be normal physiologic activity. It was then found that the hip-
pocampal formation of experimental animal models of epilepsy
presented a ripple type of activity, but at an even higher fre-
quency (93,94). These were termed fast ripples (FRs). Ripples
and FRs were also found in the mesial temporal structures of
epileptic patients in whom microelectrodes were inserted (93).
This discovery started an extensive series of investigations in
experimental animals and in humans using microelectrodes. It
was then found that low-amplitude activity in the same fre-
quency range (100 to 500 Hz) could also be recorded from EEG
macroelectrodes (95). This was followed by a series of studies
with clinical EEG recordings. We will review separately the
microelectrode and the macroelectrode studies and then discuss
the issues in interpreting their similarities and differences. A first
review of this topic was written by Rampp and Stefan (96).
High-Frequency Activity from Microelectrodes
Microelectrodes are normally used to record the activity of one
or a small number of cells (we only discuss here extracellular
recordings) in the form of action potentials. This requires a
small electrode tip (usually 40 µ in diameter or smaller) and a
sampling rate sufficient to identify action potentials (typically
20 kHz). Equipment required for such recordings is common in
the experimental laboratory but is not normally used when
investigating epileptic patients. Standard EEG equipment can-
not record properly from microelectrodes because of their high
impedance compared to EEG electrodes and because it cannot
deal with such high sampling rates. Microelectrode investiga-
tions in humans therefore require a complex setup where the
EEG is recorded with a standard clinical EEG system at the
same time that microelectrode activity is recorded with a paral-
lel system dedicated to this task.
In addition to providing action potentials, microelectrodes
also record the field potential, or the local EEG. This can be
revealed by filtering out action potentials (removing frequencies
above approximately 600 Hz). Figure 37.6 shows how the differ-
ent components of a microelectrode recording are obtained. The
knowledge of when action potentials fire is of course important
in interpreting the high frequencies of the LFP.
As indicated above, Bragin et al. (93) recorded ripples in the
frequency range of 80 to 200 Hz and FRs between 200 and
500 Hz in the hippocampal region of rats in which chronic
epilepsy was induced with intrahippocampal injection of kianic
acid, and in human hippocampal regions of epileptic patients
investigated with intracerebral electrodes, through which micro-
electrodes were inserted (Fig. 37.7). They noted that FRs were
not present in the recordings from kindled animals. The same
group then demonstrated that ripples were the more frequent
pattern in the least abnormal temporal lobe (in patients with
electrodes implanted on both sides because of possible bitempo-
ral epilepsy), whereas FRs were more frequent on the more

Figure 37.6 Example of band-pass filter settings used to detect specific
activities recorded from microelectrodes: (A) unfiltered EEG, (B) ripple
frequency (80 to 160 Hz), (C) fast ripple frequency (250 to 500 Hz), and
(D) unit activity (600 Hz to 5 kHz). Marker pulses shown in B and C
indicate trigger points used for waveform averaging. (From Bragin A,
Mody I, Wilson CL, et al. Local generation of fast ripples in epileptic
brain. J Neurosci. 2002;22(5):2012–2021.)
abnormal side (97). They also showed that FRs are generated
over a smaller volume than ripples in patients; in animals the FR
generator appeared to be less than 1 mm 3 (6). In rats, it was
found that the regions generating FRs were stable over time
(days and weeks); it was also found that the broader the region
in which FRs were found, the more frequent were the seizures,
implying that the FRs represented the epileptogenic region and
seizures were more frequent if the epileptogenic region was
larger. FRs appeared to be generated in multiple discrete clusters
of hyperexcitable neurons. Engel et al. (98) concluded that FRs
may be a more efficient marker of epileptogenic tissue than syn-
chronized cellular activity because the field potential of FRs can
be recorded more easily than cellular activity, given the small
fraction of neurons that participate in the synchronous activity
of epileptic discharges. In a different experimental context
(ketamine-induced anesthesia in the cat), Grenier et al. (99)
argued that fast oscillations result from a vicious cycle in which
bursts of population spikes generate a field potential which itself
facilitates neuronal firing.
In humans, it was observed that ripples and FRs were more fre-
quent in non-REM sleep than during wakefulness and REM sleep.
In addition, REM sleep suppressed ripples but not FRs, compared
to wakefulness (100). In patients with temporal lobe epilepsy,
Staba et al. (101) found that the ratio of FRs to ripples increased
with increased hippocampal atrophy, thus establishing a direct
link between the generation of FRs and the pathology underlying
mesial temporal lobe epilepsy. In an attempt to try to link directly
the generation of high-frequency oscillations (HFOs, including
ripples and FRs) and the generation of seizures, Bragin et al. (102)
studied the evolution of HFOs after kianic acid injection, a model
Chapter 37 ■ High-Frequency EEG Activity 757
Figure 37.7 A: MRI scan of an electrode track remaining after removal
of a clinical electrode implanted into the right posterior entorhinal cor-
tex (EC), with a proportionally sized Photoshop image of the electrode
superimposed over the track to illustrate the position of a recording
microelectrode within the EC. B: Enlarged Nissl-stained autopsy section
of a entorhinal cortex with an orientation similar to the MRI in A, and
with the same proportionally superimposed electrode image. Arrows
indicate locations of the recording sizes of the clinical electrode. (From
Bragin A, Mody I, Wilson CL, et al. Local generation of fast ripples in
epileptic brain. J Neurosci. 2002;22(5):2012–2021.)
in which status epilepticus is generated, followed by a latent
period and then spontaneous recurrent seizures. They found that
if no HFO developed, no seizure developed. They also found that
the earlier HFOs appeared following the status, the earlier seizures
developed, indicating that HFOs appear to reflect epileptogenicity
in this model. In a subsequent study of the same model, they
attempted to link seizure onset and HFOs, but found that seizures
started with an EEG wave before the first ictal HFO was observed,
concluding that HFOs do not seem to be the immediate trigger of
seizure discharges (103).
High-Frequency Activity from Macroelectrodes
EEG Activity between 40 and 120 Hz in Epilepsy
Activity in the range of 80 to 120 Hz was reported in subdural
recordings at the time of electrodecremental seizure onset, and
this activity appeared limited to the region of seizure onset (104).
The same year, Allen et al. (105) reported that frequencies around
100 Hz might be useful in localizing the seizure onset in frontal
758 Part V ■ Complementary and Special Techniques

Figure 37.8 Left: Ictal EEG of spasms in series in patient with symptomatic West syndrome. Top: A conventional trace.
Bottom: A temporally expanded trace. The time points of A and B (arrowheads) in the conventional trace correspond to
those of a and b (arrowheads) in the expanded trace. Rhythmic gamma activity at around 60 Hz is dominant over the pos-
terior head area and is associated with each spasm in the expanded trace. The power spectra in the right part of the fig-
ure correspond to the 2-second ictal EEG segments, with each segment beginning from each arrow in the conventional
trace. Delt., deltoid muscle. Right: Time evolution of spasm-associated gamma activity in averaged power. Spectral peaks
of the ictal gamma activity are indicated by arrows at around 60 or 70 Hz in the EEG. Activity from the deltoid muscles
has much broader and noisier spectra than does gamma activity. The peaks of gamma activity on the scalp precede those
of muscle activity by about 200 msec. (From Kobayashi K, Oka M, Akiyama T, et al. Very fast rhythmic activity on scalp
EEG associated with epileptic spasms. Epilepsia. 2004;45(5):488–496.) (See color insert)

lobe epilepsy. It took quite a few years until significant studies
evaluated further the importance of activity faster than the beta
band in epileptic patients.
There have been several studies relating activity in the 40- to
150-Hz band and different types of epileptic syndromes. Worrell
et al. (106) found that bursts of interictal activity in the 60 to 100
Hz range were often found in the intracerebral EEG of patients
with neocortical epilepsy, were indicative of the seizure onset
region, and became more frequent as seizures were approaching.
Studying the scalp EEG of a large group of children, Wu et al.
(107) found that interictal paroxysmal beta and gamma activity
was an accurate indicator of epilepsy and of the seizure onset
region, particularly in children younger than 3 years. In young
children with hemimegalencephaly, Yamazaki et al. (108) fre-
quently saw activity around 40 to 50 Hz. Kobayashi et al. (109)
showed that gamma activity (50 to 100 Hz) was present at the
time of infantile spasms in the vast majority of spasms. The
power spectrum of this activity showed a clear peak, thus
differentiating it from muscle activity with energy in the same
frequency range (Fig. 37.8). Akiyama et al. (110) also showed
activity between 60 and 150 Hz at the onset of spasms, and sug-
gested that high-frequency activity triggers the spasm.
Moving into frequencies above 120 Hz, but in relation to
infantile spasms, Ramachandranair et al. (111) demonstrated
activity between 150 and 250 Hz at spasm onset, localized to
frontal or rolandic areas in intracranial recordings. Meanwhile,
Akiyama et al. (112) presented a method to facilitate the visual-
ization of activity recorded from subdural grid electrodes
between 60 and 120 Hz at seizure onset. Gardner et al. (113)
described a method of automatic detection of short bursts of
activity between 40 and 100 Hz.
EEG Activity Higher Than 100 Hz in Epilepsy
Given the size of the presumed generators of HFOs, it would
seem unlikely that such activity can be recorded from the stan-
dard macroelectrodes used in the presurgical evaluation of
 Chapter 37 ■ High-Frequency EEG Activity 759

Figure 37.9 Discrete high-frequency oscillations identified in the EEG through visual inspection of the digitally filtered
signal. A: Unfiltered ictal EEG at standard timescale. Highlighted sample for further visual analysis in B and C. Ictal EEG
is shown at an expanded time-scale and increased vertical gain with high-pass filtering at 50 and 100 Hz in B and C. A
brief 160- to 210-Hz segmental oscillation is well visualized in the RAH1-2 channel at both filter settings. D: Unfiltered ictal
EEG at standard timescale. A highlighted sample is shown at increased vertical and horizontal gain in parts E and F with
filtering. A discrete 285- to 375-Hz segmental oscillation is clearly visualized in RMH2-3 and to a lesser extent RMH3-4 at
both filter settings. (From Jirsch JD, Urrestarazu E, LeVan P, et al. High-frequency oscillations during human focal seizures.
Brain. 2006;129(pt 6):1593–1608.)

some patients with intractable epilepsy. Electrodes used for
subdural recordings are usually disks of 4 to 5 mm 2 with an
exposed diameter of 2.3 mm, and depth electrodes used for
intracerebral implantation are made of cylinders of approxi-
mately the same surface area. This size is approximately 4000
times larger than a microelectrode. A group at the Montreal
Neurological Institute uses electrodes made in house with a sur-
face contact of 0.8 mm 2, somewhat smaller than commercial
electrodes. Using a recording system that allowed sampling at
2000 Hz after 500 Hz low-pass filtering, they embarked on a
series of studies that demonstrated that high-frequency activity
in the range of 100 to 500 Hz could be recorded with these
macroelectrodes and appeared to have a strong relationship
with the seizure-generating region.
The first study demonstrated HFOs during epileptic seizures
(95). High-frequency activity, either in the form of brief oscil-
lations or as prolonged discharges, was found most often in the
electrodes in which the seizure had started and rarely in regions
to which the seizure had propagated (Fig. 37.9). It was also
found that patients for whom it was felt that the seizure onset
had been missed (for instance if clinical signs preceded the first
EEG change), did not have high-frequency activity at seizure
onset. It was then demonstrated that interictal HFOs were also
relatively frequent in the ripple frequency band, that is, not just
in the FR frequency band (114). These HFOs occurred in three
situations: (i) on top of traditional EEG spikes (as was most
often the case for FRs recorded with microelectrodes); (ii)
totally independently of spikes; (iii) as a result of filtering a
760 Part V ■ Complementary and Special Techniques

A FR-Vis-S B FR-NVis-S C FR-Vis

(Fast Ripples visible in spike) (Fast Ripples not visible in spike) (Fast Ripples visible)

Unfiltered

HP 80 Hz

HP 250 Hz

300 µV 300 µV 50 µV

100 msec 100 msec 100 msec

Figure 37.10 Different types of HFOs, with and without spikes. A: Fast ripple visible in spike. B: Fast ripple not visible
in spike. C: Fast ripple visible, independently of a spike. Top: Nonfiltered EEG. Middle: EEG filtered with high-pass filter
of 80 Hz. Bottom: EEG filtered with high-pass filter of 250 Hz. (From Urrestarazu E, Chander R, Dubeau F, et al. Interictal
high-frequency oscillations (100–500 Hz) in the intracerebral EEG of epileptic patients. Brain. 2007;130(pt 9):2354–2366.)

sharp spike (Fig. 37.10). In both the above studies, HFOs were
found in neocortical regions as well as in mesial temporal lobe
structures. In one recent study, Kobayashi et al. (115), recording
with a 3000-Hz filter, demonstrated activity up to 900 Hz from
subdural electrodes at seizure onset, in the form of short oscil-
lations superimposed on spikes.
It is known that the zone of interictal spiking and the seizure
onset zone are related to each other, but the relationship is not
always tight. By examining the spatial relationship between the
seizure onset zone, the region in which interictal spiking took
place and the region in which HFOs were found, it could be deter-
mined that HFOs had a tighter correspondence with the seizure
onset zone than did interictal spikes (116; Fig. 37.11). HFOs there-
fore appeared to be a better candidate for a biomarker of ictogen-
esis than spikes. These results were largely confirmed for mesial
and polar structures of the temporal lobe in the study of Crepon
et al. (117). In a study that was not directly related to HFOs as dis-
crete entities but rather to activity in the ripple and FR frequency
bands as calculated by spectral analysis, Urrestarazu et al. (118)
showed that high-frequency activity was suppressed during the
slow wave following spikes, compared to the background. The
suppression was most pronounced in the region closest to the
maximum of the spike and was most prominent in the hippocam-
pus compared to the amygdala or neocortical region. The sup-
pression of high-frequency activity was interpreted as reflecting
the strength of the inhibition that follows a spike. A statistical
approach to the demonstration of this phenomenon and of other
high-frequency changes was presented by Kobayashi et al. (119).
The importance of HFOs was further underlined when it
was found that, in patients with lesions, HFOs were more
closely coupled with the region of seizure onset than with the
lesion, which is sometimes but not always the source of seizures
(120). The above studies were performed by measuring HFO
rates during non-REM sleep, since it was shown that, as in
microelectrode recordings, HFOs are most frequent during
non-REM sleep but their relative distribution across brain
regions is not state-dependent (121).
If there is a good spatial correspondence between the region of
seizure onset and HFOs, one can naturally wonder if there is also
a temporal coupling. The most natural question is whether HFOs
become more frequent as a seizure approaches. Khosravani et al.
(122) found that HFOs often increased in the few seconds imme-
diately preceding a seizure. This study also demonstrated that
HFOs could be recorded from commercially available subdural
electrodes, with a surface contact of 4 mm 2. In a study that exam-
ined fluctuations in HFO rate and energy in the 15-, 5-, and
1-minute intervals preceding seizure occurrence, Jacobs et al.
(123) did not find any systematic change. These two studies
appear to indicate that if there is a change in HFOs prior to
seizures, it is only immediately prior to their occurrence and thus
it may be difficult to distinguish them from the seizure onset itself.
Looking at HFO fluctuations at the time scale of days in rela-
tion to seizure occurrence and to changing antiepileptic medica-
tion, Zijlmans et al. (124) demonstrated that medication
reduction results directly in an increase in the rate of HFO occur-
rence. Seizures, on the other hand, when they occurred in the
 Chapter 37 ■ High-Frequency EEG Activity 761

Figure 37.11 HFOs are most prominent in the seizure onset region. This patient presented with interictal and
ictal signs of bitemporal epilepsy on scalp EEG. He was implanted bitemporally aiming at the amygdala (LA, RA)
and hippocampus (LH, RH). The MRI revealed a malrotation of the right hippocampus. All seizures originated
from the right mesial temporal structures, where the highest rates of HFOs were found. The figure shows three
sections selected from the same few seconds of EEG. In each panel the original EEG is on the left and the thin sec-
tion marked in grey is expanded at the right, after filtering and changing the gain as indicated. The unfiltered EEG
shows frequent spikes in both mesial TL structures. Top panel: Co-occurrence of ripples and fast ripples outside
spikes in channel RH1. Additionally a ripple is seen in channel RH2. Middle panel: The selection includes two
spikes over LA1 and LH1. Both spikes lied outside the seizure onset zone and in the healthier mesial TL. The
expanded sections show no high-frequency oscillation during this time period. Bottom panel: The gray selection
shows two spikes simultaneously at RH1 and RH2 in the seizure onset zone. No oscillation is visible in the spike
at RH1 in the extended unfiltered EEG, while the spike in RH2 shows a very short and small oscillation. The fil-
tered EEG segments clearly reveal a ripple and fast ripple oscillation during these spikes. Spikes looking similar in
the unfiltered EEG might differ in whether they carry high-frequency oscillations. (From Jacobs J, LeVan P,
Chander R, et al. Interictal high-frequency oscillations (80–500 Hz) are an indicator of seizure onset areas inde-
pendent of spikes in the human epileptic brain. Epilepsia. 2008;49(11):1893–1907.)
762 Part V ■ Complementary and Special Techniques
context of stable medications, did not cause any change in HFO
occurrence rates. This is in contrast to interictal spikes (125–127),
for which medication reduction does not result directly in an
increased occurrence rate; seizures, however, are followed by
increased spiking (this explains the often seen increase in spiking
after medication withdrawal, as seizures occur more often in this
context and mediate the increase in spiking). Comparing HFOs
and spikes, one therefore concludes that HFOs behave more like
seizures than spikes do in the context of changing medication
levels. They may therefore be a better marker of disease activity
than spikes. In another study relating HFOs to seizure occur-
rence, Zijlmans et al. (128) were not able to replicate in humans
the results of Bragin et al. (129) obtained in rats, to the effect that
the frequency of seizure occurrence was related to the size of the
region generating HFOs.
Given these results, the question naturally arises regarding the
possible use of HFOs in the surgical evaluation of epileptic
patients. In a small study, Ochi et al. (130) noted some correla-
tion between the removal of regions generating HFOs and suc-
cessful surgery. In a larger patient group and studying
systematically the correlation between regions generating HFOs,
spikes and the seizure onset, Jacobs et al. (131) showed that the
outcome of surgery was more closely related to whether the
HFO-generating region was removed than to whether spike- or
seizure-generating regions were removed. This suggests that
regions generating HFOs are potential seizure-generating regions.
The question of what is the optimum electrode size to record
HFOs remains open. We have seen above that they were well
recorded with “small” EEG electrodes (0.8 mm 2) but that they
could also be recorded with standard electrodes (4 mm 2; see Ref.
122). Worrell et al. (132) performed combined recordings with
standard depth electrodes (contact size 9.4 mm 2) and with
microelectrodes placed a few millimeters away. They concluded
that both electrode types could record HFOs but that high fre-
quencies were better recorded with the microelectrodes. Clemens
et al. (133) recorded activity up to 150 Hz with foramen ovale
electrodes. In preliminary findings by Chatillon et al. (134), it was
found, surprisingly, that electrodes having a surface of 0.2 mm 2
did not record HFOs better than electrodes having a 0.8-mm 2
surface. The optimum electrode size therefore remains to be
determined.
When looking for HFOs in intracerebral EEG, recorded by
depth or subdural recordings, an important word of caution is
necessary: artifacts from contracting scalp muscles can appear
surprisingly like HFOs (135). Just as EEG activity is transmitted
through the skull and can be recorded on the scalp, activity gen-
erated outside the skull is measurable in intracerebral record-
ings. This has been shown for eye blinks (136). Similarly, EMG
activity from scalp muscles is transmitted through the skull and
can be recorded on subdural electrodes and on the most super-
ficial contacts of depth electrodes. Because of skull attenuation
and because of the small surface of synchronized activity in
muscles, EMG recorded with intracerebral EEG electrodes is of
small amplitude, just like HFOs. This is a particular concern
during seizures, when scalp muscles can have strong contrac-
tions. One way to become aware of this possibility and to
observe the intracerebral distribution of EMG activity is to ask
patients to contract their jaws and observe the changes directly
related to the contractions.
Discussion
Activity between 40 and 100 Hz is often seen in relation to the
seizure onset zone in intracranial recordings. Activity in this fre-
quency range can also be seen in the scalp EEG, particularly in
young children and particularly in the context of infantile
spasms. Frequencies above 100 Hz have not been reported in the
scalp EEG, most likely as a result of a small generator region. In a
recent review, Engel et al. (137) discussed the origin and possible
meaning of normal and pathologic microelectrode-recorded
HFOs in the context of mesial temporal lobe epilepsy. They con-
cluded that ripples appear to reflect synchronized inhibitory
postsynaptic potentials and that FRs probably reflect bursts of
population spikes. It was possible to demonstrate in human
recordings that interneurons and pyramidal cells fire in sequence,
thus supporting the generation of ripples by inhibitory activity
(138). Engel et al. (137) also concluded that not all ripples repre-
sent normal phenomena since, for instance, ripples are not nor-
mally seen in the dentate gyrus but appear there prior to seizures
in kianic acid–treated rats (102). Abnormal ripples could be gen-
erated by slow bursts of population spikes. It is also possible that
not all FRs represent abnormal phenomena.
From the studies of HFOs recorded with intracerebral EEG
macroelectrodes, one can conclude that HFOs are an important
marker of the epileptogenic region not only in mesial temporal
structures, where they are particularly abundant, but also in
neocortical regions. Although activity in the FR band was
sometimes more specific than activity in the ripple band, the
distinction that was possible with microelectrodes has not fully
carried over with macroelectrodes. There is clearly a relation-
ship between the events recorded with microelectrodes and
those recorded with macroelectrodes, but the nature of this
relationship is uncertain. From microelectrode recordings, it
would seem that the field of HFOs should be too small to be
clearly visible with macroelectrodes. Microelectrode recordings
are made with a local reference however, and they may therefore
be blind to broader field changes that take place at the same
time as the very local event that they are best suited to detect.
The properties of these broader events may not be identical to
those of the very local events. There is a clear need for a system-
atic study of the distribution of these HFO fields and of the
electrode sizes most appropriate for recording the most dis-
criminating events. Whereas microelectrodes record activity
that is easiest to interpret scientifically because it relates directly
to the basic building block of the nervous system, that is, the
neuron, macroelectrodes record potentials that are more diffi-
cult to interpret in terms of the activity of individual cells.
However, activity recorded with macroelectrodes may be more
significant from the clinical point of view because it represents
what happens to neuronal populations on a larger scale.
There is a frequent overlap between the spike-generating
region and the seizure-generating region, justifying the interest in
spikes. The relationship is, however, uncertain, and it is not clear
how information about spikes should be integrated in the defini-
tion of the epileptogenic region. The investigations of HFOs

recorded with macroelectrodes suggest that they are more tightly
coupled to seizures than are spikes, in both time and in space.
They may therefore be of greater clinical utility. One study (131)
even suggests that HFOs may be a better indicator of epilepto-
genicity than the seizure onset itself. More studies are required
before these results can become clinically applicable.
It is becoming more common now that clinical EEG record-
ing systems allow recordings of activity up to 500 or 600 Hz, but
these systems are not equipped for microelectrode recordings.
If it is shown conclusively that microelectrode recordings pro-
vide useful clinical information that cannot be obtained from
EEG electrodes, it is possible that future clinical equipment will
allow microelectrode recordings in addition to the EEG.
An important issue that has not yet been addressed in
intracerebral EEG studies is the differentiation between normal
high-frequency responses related to cognitive activity and
pathologic HFOs related to epilepsy, which may occur in over-
lapping frequency ranges. It is unlikely that the epileptogenic
HFOs recorded during slow wave sleep, as done in most of the
studies reported above, are influenced by cognitive activity. On
the other hand, even though epileptogenic HFOs are more
common during sleep, they can also occur during wakefulness,
and thus can potentially contaminate recordings of normal
high-frequency responses. A few differences between these elec-
trophysiologic phenomena, however, may mitigate this con-
cern. First, event-related high gamma responses are usually
time-locked to a functional task, and although these responses
are robust enough to be differentiated in single trials, their sta-
tistical significance is usually established over a series of trials.
It is thus unlikely that epileptogenic HFOs will consistently
occur at the same times. Second, the epileptogenic HFOs
observed to date have typically consisted of brief oscillatory,
that is, sinusoidal, wave packets, whereas the most consistent
high-frequency responses during functional activation are
broadband and thus not really oscillations per se. The utility of
this distinction, however, will require further testing as investi-
gations of both phenomena continue.
Another major issue remains, of course, the limitation of HFO
recordings to intracerebral electrodes. Is there a possibility that
they might be visible noninvasively, from scalp EEG or from
MEG? Given what we know on the one hand about the size of the
region over which HFOs are synchronous (possibly up to 1 cm)
and the amplitude of this activity (10 to 20 V) and, on the other
hand, about the region of cortex that needs to be involved for a
spike of several hundred microvolts to be visible on the scalp (a
patch of 2.5 to 3 cm 2), it appears unlikely that either EEG or MEG
can record HFOs. If such recordings were to be possible, they
would probably be limited to the most superficial cortical
regions. An indication of this difficulty is given by a recent com-
bined MEG and intracerebral EEG study (139) evaluating
gamma activity evoked by cognitive tasks. This activity was
clearly visible in the intracerebral EEG but was barely measurable
in the MEG. Epileptogenic HFOs may have smaller generators
that the gamma activity generated during functional activation
and may thus have even smaller amplitudes at the scalp surface.
Although high gamma responses during functional activation
have been successfully captured with MEG and even with scalp
Chapter 37 ■ High-Frequency EEG Activity 763
EEG (140), this requires averaging in the frequency domain
across many trials of the same time-aligned activations. The tim-
ing of epileptogenic HFOs however, like epileptic seizures in gen-
eral, is inherently unpredictable, and cannot be exploited as
easily. Nevertheless, future investigations will undoubtedly con-
tinue to test the limits of both invasive and noninvasive record-
ings of high-frequency EEG activities and to explore their
relationships to both normal and pathologic neurophysiologic
mechanisms.
REFERENCES
1. Crone NE, Miglioretti DL, Gordon B, et al. Functional mapping of
human sensorimotor cortex with electrocorticographic spectral
analysis. II. Event-related synchronization in the gamma band.
Brain. 1998;121(pt 12):2301–2315.
2. Crone NE, Sinai AS, Korzeniewska A. High-frequency gamma
oscillations and human brain mapping with electrocorticography.
Prog Brain Res. 2006;159:279–302.
3. Miller KJ, Shenoy P, den Nijs M, et al. Beyond the gamma band: the
role of high-frequency features in movement classification. IEEE
Trans Biomed Eng. 2008;55(5):1634–1637.
4. Oostendorp TF, Delbeke J, Stegeman DF. The conductivity of the
human skull: results of in vivo and in vitro measurements. IEEE
Trans Biomed Eng. 2000;47(11):1487–1492.
5. Gloor P. Neuronal generators and the problem of localization in
electroencephalography: application of volume conductor theory
to electroencephalography. J Clin Neurophysiol. 1985;2(4):327–354.
6. Bragin A, Mody I, Wilson CL, et al. Local generation of fast ripples
in epileptic brain. J Neurosci. 2002;22(5):2012–2021.
7. McCreery DB, Agnew WF, Yuen TG, et al. Charge density and
charge per phase as cofactors in neural injury induced by electri-
cal stimulation. IEEE Trans Biomed Eng. 1990;37(10):996–1001.
8. Wei XF, Grill WM. Current density distributions, field distribu-
tions and impedance analysis of segmented deep brain stimulation
electrodes. J Neural Eng. 2005;2(4):139–147.
9. Adrian ED, Matthews BHC. The Berger rhythm: potential changes
from the occipital lobes in man. Brain. 1934;57:355–385.
10. Jasper HH, Andrews HL. Electro-encephalography. III. Normal
differentiation of occipital and precentral regions in man. Arch
Neurol Psychiatry. 1938;39:96–115.
11. Ectors L. Étude oscillographique des activités sensitives et motrice
du cortex cérébral chez l’animal éveillé. Comptes rendus des séances
de la Société de biologie et de ses filiales. 1935;120:1339–1343.
12. Freeman WJ. Spatial properties of an EEG event in the olfactory bulb
and cortex. Electroencephalogr Clin Neurophysiol. 1978;44:586–605.
13. Bressler SL, Freeman WJ. Frequency analysis of olfactory system
EEG in cat, rabbit, and rat. Electroencephalogr Clin Neurophysiol.
1980;50:19–24.
14. Gray CM, König P, Engel AK, et al. Oscillatory responses in cat
visual cortex exhibit inter-columnar synchronization which
reflects global stimulus properties. Nature. 1989;338:334–337.
15. Eckhorn R, Bauer R, Jordan W, et al. Coherent oscillations: a mech-
anism of feature linking in the visual cortex? Multiple electrode and
correlation analyses in the cat. Biol Cybern. 1988;60:121–130.
16. Engel AK, Konig P, Kreiter AK, et al. Interhemispheric synchro-
nization of oscillatory neuronal responses in cat visual cortex.
Science. 1991;252:1177–1179.
17. Murthy VN, Fetz EE. Coherent 25- to 35-Hz oscillations in the
sensorimotor cortex of awake behaving monkeys. Proc Natl Acad
Sci U S A. 1992;89:5670–5674.
764 Part V ■ Complementary and Special Techniques
18. Sanes JN, Donoghue JP. Oscillations in local field potentials of the
primate motor cortex during voluntary movement. Proc Natl Acad
Sci U S A. 1993;90:4470–4474.
19. Singer W, Gray CM. Visual feature integration and the temporal
correlation hypothesis. Annu Rev Neurosci. 1995;18:555–586.
20. Von der Malsburg C. Binding in models of perception and brain
function. Curr Opin Neurobiol. 1995;5:520–526.
21. Fries P. A mechanism for cognitive dynamics: neuronal communi-
cation through neuronal coherence. Trends Cogn Sci. 2005;9(10):
474–480.
22. Fries P, Nikolic D, Singer W. The gamma cycle. Trends Neurosci.
2007;30(7):309–316.
23. Bressler SL. The gamma wave: a cortical information carrier?
[news]. Trends Neurosci. 1990;13:161–162.
24. Spydell JD, Ford MR, Sheer DE. Task dependent cerebral lateraliza-
tion of the 40 Hertz EEG rhythm. Psychophysiology. 1979;16:
347–350.
25. Sheer DE, Grandstaff NW, Benignus VA. Behavior and 40-c-sec.
electrical activity in the brain. Psychol Rep. 1966;19:1333–1334.
26. DeFrance J, Sheer DE. Focused arousal, 40-Hz EEG, and motor
programming. In: Giannitrapani D, Murri L, eds. The EEG of
Mental Activities. New York, NY: Karger; 1988:153–168.
27. Pantev C, Makeig S, Hoke M, et al. Human auditory evoked
gamma-band magnetic fields. Proc Natl Acad Sci U S A. 1991;88:
8996–9000.
28. Galambos R, Makeig S, Talmachoff PJ. A 40-Hz auditory potential
recorded from the human scalp. Proc Natl Acad Sci U S A. 1981;78:
2643–2647.
29. Pantev C. Evoked and induced gamma-band activity of the human
cortex. Brain Topogr. 1995;7(4):321–330.
30. Franowicz MN, Barth DS. Comparison of evoked potentials and
high-frequency (gamma-band) oscillating potentials in rat audi-
tory cortex. J Neurophysiol. 1995;74(1):96–112.
31. Barth DS, MacDonald KD. Thalamic modulation of high-fre-
quency oscillating potentials in auditory cortex [see comments].
Nature. 1996;383(6595):78–81.
32. Basar-Eroglu C, Struber D, Schurmann M, et al. Gamma-band
responses in the brain: a short review of psychophysiological
correlates and functional significance. Int J Psychophysiol. 1996;
24(1–2):101–112.
33. Sheer DE. Focused arousal and the cognitive 40-Hz event-related
potentials: differential diagnosis of Alzheimer’s disease. Prog Clin
Biol Res. 1989;317:79–94.
34. Tiitinen H, Sinkkonen J, Reinikainen K, et al. Selective attention
enhances the auditory 40-Hz transient response in humans.
Nature. 1993;364(6432):59–60.
35. Curio G, Mackert B, Burghoff M, et al. Localization of evoked neu-
romagnetic 600 Hz activity in the cerebral somatosensory system.
Electroencephalogr Clin Neurophysiol. 1994;91:483–487.
36. Curio G. Linking 600-Hz “spikelike” EEG/MEG wavelets (“sigma-
bursts”) to cellular substrates: concepts and caveats. J Clin
Neurophysiol. 2000;17(4):377–396.
37. Curio G. Ultrafast EEG activities. In: Niedermeyer E, Lopes da
Silva F, eds. Electroencephalography: Basic Principles, Clinical
Applications, and Related Fields. 5th ed. Philadelphia, PA: Lippincott,
Williams & Wilkins; 2005:495–504.
38. Emerson RG, Sgro JA, Pedley TA, et al. State-dependent changes in
the N20 component of the median nerve somatosensory evoked
potential. Neurology. 1988;38(1):64–68.
39. Jones MS, Barth DS. Spatiotemporal organization of fast ( 200 Hz)
electrical oscillations in rat Vibrissa/Barrel cortex. J Neurophysiol.
1999;82(3):1599–1609.
40. Baker SN, Curio G, Lemon RN. EEG oscillations at 600 Hz are
macroscopic markers for cortical spike bursts. J Physiol. 2003;
550(pt 2):529–534.
41. Pfurtscheller G. Quantification of ERD and ERS in the time
domain. In: Pfurtscheller G, Lopes da Silva FH, eds. Event-
Related Desynchronization. 1st ed. Amsterdam: Elsevier; 1999:
89–105.
42. Kalcher J, Pfurtscheller G. Discrimination between phase-locked and
non-phase-locked event-related EEG activity. Electroencephalogr Clin
Neurophysiol. 1995;94:381–384.
43. Crone NE, Boatman D, Gordon B, et al. Induced electrocortico-
graphic gamma activity during auditory perception. Clin
Neurophysiol. 2001;112(4):565–582.
44. Spencer KM, Polich J. Poststimulus EEG spectral analysis and
P300: attention, task, and probability. Psychophysiology. 1999;36(2):
220–232.
45. Demiralp T, Ademoglu A, Comerchero M, et al. Wavelet analysis of
P3a and P3b. Brain Topogr. 2001;13(4):251–267.
46. Yordanova J, Kolev V, Polich J. P300 and alpha event-related desyn-
chronization (ERD). Psychophysiology. 2001;38(1):143–152.
47. Herrmann CS, Knight RT. Mechanisms of human attention:
event-related potentials and oscillations. Neurosci Biobehav Rev.
2001;25(6):465–476.
48. Pfurtscheller G, Neuper C, Kalcher J. 40-Hz oscillations during
motor behavior in man. Neurosci Lett. 1993;164:179–182.
49. Pfurtscheller G, Flotzinger D, Neuper C. Differentiation between
finger, toe and tongue movement in man based on 40 Hz EEG.
Electroencephalogr Clin Neurophysiol. 1994;90:456–460.
50. Tallon-Baudry C, Bertrand O, Delpuech C, et al. Stimulus specificity
of phase-locked and non-phase-locked 40 Hz visual responses in
human. J Neurosci. 1996;16(13):4240–4249.
51. Ray S, Jouny CC, Crone NE, et al. Human ECoG analysis during
speech perception using matching pursuit: a comparison between
stochastic and dyadic dictionaries. IEEE Trans Biomed Eng. 2003;
50(12):1371–1373.
52. Ray S, Niebur E, Hsiao SS, et al. High-frequency gamma activity
(80–150 Hz) is increased in human cortex during selective atten-
tion. Clin Neurophysiol. 2008;119(1):116–133.
53. Ohara S, Ikeda A, Kunieda T, et al. Movement-related change of
electrocorticographic activity in human supplementary motor
area proper. Brain. 2000;123(pt 6):1203–1215.
54. Pfurtscheller G, Graimann B, Huggins JE, et al. Spatiotemporal pat-
terns of beta desynchronization and gamma synchronization in cor-
ticographic data during self-paced movement. Clin Neurophysiol.
2003;114(7):1226–1236.
55. Miller KJ, denNijs M, Shenoy P, et al. Real-time functional brain
mapping using electrocorticography. Neuroimage. 2007;37(2):
504–507.
56. Miller KJ, Leuthardt EC, Schalk G, et al. Spectral changes in corti-
cal surface potentials during motor movement. J Neurosci. 2007;
27(9):2424–2432.
57. Szurhaj W, Bourriez JL, Kahane P, et al. Intracerebral study of
gamma rhythm reactivity in the sensorimotor cortex. Eur J Neurosci.
2005;21(5):1223–1235.
58. Leuthardt EC, Schalk G, Wolpow JR, et al. A brain-computer inter-
face using electrocorticographic signals in humans. J Neural Eng.
2004;1:63–71.
59. Gonzalez SL, de Peralta RG, Thut G, et al. Very high frequency oscil-
lations (VHFO) as a predictor of movement intentions. Neuroimage.
2006;32(1):170–179.
60. Gunji A, Ishii R, Chau W, et al. Rhythmic brain activities related to
singing in humans. Neuroimage. 2007;34(1):426–434.

61. Brovelli A, Lachaux JP, Kahane P, et al. High gamma frequency
oscillatory activity dissociates attention from intention in the
human premotor cortex. Neuroimage. 2005;28(1):154–164.
62. Lachaux JP, Hoffinann D, Minotti L, et al. Intracerebral dynamics
of saccade generation in the human frontal eye field and supple-
mentary eye field. Neuroimage. 2006;30(4):1302–1312.
63. Gross J, Schnitzler A, Timmermann L, et al. Gamma oscillations in
human primary somatosensory cortex reflect pain perception.
PLoS Biol. 2007;5(5):1168–1173.
64. Ihara A, Hirata M, Yanagihara K, et al. Neuromagnetic gamma-
band activity in the primary and secondary somatosensory areas.
Neuroreport. 2003;14(2):273–277.
65. Hauck M, Lorenz J, Engel AK. Attention to painful stimulation
enhances gamma-band activity and synchronization in human
sensorimotor cortex. J Neurosci. 2007;27(35):9270–9277.
66. Bauer M, Oostenveld R, Peeters M, et al. Tactile spatial attention
enhances gamma-band activity in somatosensory cortex and
reduces low-frequency activity in parieto-occipital areas. J Neurosci.
2006;26(2):490–501.
67. Edwards E, Soltani M, Deouell LY, et al. High gamma activity in
response to deviant auditory stimuli recorded directly from
human cortex. J Neurophysiol. 2005;94(6):4269–4280.
68. Kaiser J, Lutzenberger W. Human gamma-band activity: a window
to cognitive processing. Neuroreport. 2005;16(3):207–211.
69. Trautner P, Rosburg T, Dietl T, et al. Sensory gating of auditory
evoked and induced gamma band activity in intracranial record-
ings. Neuroimage. 2006;32(2):790–798.
70. Hoogenboom N, Schoffelen JM, Oostenveld R, et al. Localizing
human visual gamma-band activity in frequency, time and space.
Neuroimage. 2006;29(3):764–773.
71. Vidal JR, Chaumon M, O’Regan JK, et al. Visual grouping and the
focusing of attention induce gamma-band oscillations at different
frequencies in human magnetoencephalogram signals. J Cogn
Neurosci. 2006;18(11):1850–1862.
72. Lachaux JP, George N, Tallon-Baudry C, et al. The many faces of
the gamma band response to complex visual stimuli. Neuroimage.
2005;25(2):491–501.
73. Tallon-Baudry C, Bertrand O, Henaff MA, et al. Attention modulates
gamma-band oscillations differently in the human lateral occipital
cortex and fusiform gyrus. Cereb Cortex. 2005;15(5):654–662.
74. Tanji K, Suzuki K, Delorme A, et al. High-frequency gamma-band
activity in the basal temporal cortex during picture-naming and
lexical-decision tasks. J Neurosci. 2005;25(13):3287–3293.
75. Miyanari A, Kaneoke Y, Ihara A, et al. Neuromagnetic changes of
brain rhythm evoked by intravenous olfactory stimulation in
humans. Brain Topogr. 2006;18(3):189–199.
76. Crone NE, Hao L, Hart J Jr, et al. Electrocorticographic gamma
activity during word production in spoken and sign language.
Neurology. 2001;57(11):2045–2053.
77. Sinai A, Bowers CW, Crainiceanu CM, et al. Electrocorticographic
high gamma activity versus electrical cortical stimulation mapping
of naming. Brain. 2005;128(pt 7):1556–1570.
78. Mainy N, Jung J, Baciu M, et al. Cortical dynamics of word recog-
nition. Hum Brain Mapp. 2008;29(11):1215–1230.
79. Jung J, Mainy N, Kahane P, et al. The neural bases of attentive read-
ing. Hum Brain Mapp. 2008;29(10):1193–1206.
80. Crone NE, Miglioretti DL, Gordon B, et al. Functional mapping of
human sensorimotor cortex with electrocorticographic spectral
analysis. I. Alpha and beta event-related desynchronization. Brain.
1998;121(pt 12):2271–2299.
81. Zygierewicz J, Durka PJ, Klekowicz H, et al. Computationally effi-
cient approaches to calculating significant ERD/ERS changes in
Chapter 37 ■ High-Frequency EEG Activity 765
the time–frequency plane. J Neurosci Methods. 2005;145(1–2):
267–276.
82. Aoki F, Fetz EE, Shupe L, et al. Increased gamma-range activity in
human sensorimotor cortex during performance of visuomotor
tasks. Clin Neurophysiol. 1999;110(3):524–537.
83. Brunner P, Ritaccio AL, Lynch TM, et al. A practical procedure for
real-time functional mapping of eloquent cortex using electrocor-
ticographic signals in humans. Epilepsy Behav. 2009;15(3):278–286.
84. Menon V, Freeman WJ, Cutillo BA, et al. Spatio-temporal correlations
in human gamma band electrocorticograms. Electroencephalogr Clin
Neurophysiol. 1996;98:89–102.
85. Brosch M, Budinger E, Scheich H. Stimulus-related gamma oscil-
lations in primate auditory cortex. J Neurophysiol. 2002;87(6):
2715–2725.
86. Steinschneider M, Fishman YI, Arezzo JC. Spectrotemporal
analysis of evoked and induced electroencephalographic
responses in primary auditory cortex (A1) of the awake monkey.
Cereb Cortex. 2008;18(3):610–625.
87. Crone NE, Hao L. The functional significance of event-related spec-
tral changes (ERD/ERS) from the perspective of electrocorticogra-
phy. In: Reisin RC, Nuwer MR, Hallett M, et al., eds. Advances in
Clinical Neurophysiology (Suppl to Clin Neurophysiol). Amsterdam:
Elsevier Science; 2002:435–442.
88. Ray S, Hsiao SS, Crone NE, et al. Effect of stimulus intensity on
the spike-local field potential relationship in the secondary
somatosensory cortex. J Neurosci. 2008;28(29):7334–7343.
89. Ray S, Crone NE, Niebur E, et al. Neural correlates of high-
gamma oscillations (60–200 Hz) in macaque local field poten-
tials and their potential implications in electrocorticography. J
Neurosci. 2008;28(45):11526–11536.
90. Steinmetz PN, Roy A, Fitzgerald PJ, et al. Attention modulates
synchronized neuronal firing in primate somatosensory cortex.
Nature. 2000;404(6774):187–190.
91. Fries P, Reynolds JH, Rorie AE, et al. Modulation of oscillatory
neuronal synchronization by selective visual attention. Science.
2001;291(5508):1560–1563.
92. Jensen O, Kaiser J, Lachaux JP. Human gamma-frequency oscilla-
tions associated with attention and memory. Trends Neurosci.
2007;30(7):317–324.
93. Bragin A, Engel J Jr, Wilson CL, et al. Hippocampal and entorhinal
cortex high-frequency oscillations (100–500 Hz) in human epilep-
tic brain and in kainic acid-treated rats with chronic seizures.
Epilepsia. 1999;40(2):127–137.
94. Bragin A, Engel J Jr, Wilson CL, et al. High-frequency oscillations
in human brain. Hippocampus. 1999;9(2):137–142.
95. Jirsch JD, Urrestarazu E, LeVan P, et al. High-frequency oscillations
during human focal seizures. Brain. 2006;129(pt 6):1593–1608.
96. Rampp S, Stefan H. Fast activity as a surrogate marker of epileptic
network function? Clin Neurophysiol. 2006;117(10):2111–2117.
97. Staba RJ, Wilson CL, Bragin A, et al. Quantitative analysis of
high-frequency oscillations (80–500 Hz) recorded in human
epileptic hippocampus and entorhinal cortex. J Neurophysiol.
2002;88(4):1743–1752.
98. Engel J Jr, Wilson C, Bragin A. Advances in understanding the
process of epileptogenesis based on patient material: what can
the patient tell us? Epilepsia. 2003;44(suppl 12):60–71.
99. Grenier F, Timofeev I, Steriade M. Neocortical very fast oscilla-
tions (ripples, 80–200 Hz) during seizures: intracellular corre-
lates. J Neurophysiol. 2003;89(2):841–852.
100. Staba RJ, Wilson CL, Bragin A, et al. High-frequency oscillations
recorded in human medial temporal lobe during sleep. Ann Neurol.
2004;56(1):108–115.
766 Part V ■ Complementary and Special Techniques
101. Staba RJ, Frighetto L, Behnke EJ, et al. Increased fast ripple to rip-
ple ratios correlate with reduced hippocampal volumes and neu-
ron loss in temporal lobe epilepsy patients. Epilepsia. 2007;48(11):
2130–2138.
102. Bragin A, Wilson CL, Almajano J, et al. High-frequency oscilla-
tions after status epilepticus: epileptogenesis and seizure genesis.
Epilepsia. 2004;45(9):1017–1023.
103. Bragin A, Wilson CL, Fields T, et al. Analysis of seizure onset on
the basis of wideband EEG recordings. Epilepsia. 2005;46(suppl
5):59–63.
104. Fisher RS, Webber WR, Lesser RP, et al. High-frequency EEG activ-
ity at the start of seizures. J Clin Neurophysiol. 1992;9(3):441–448.
105. Allen PJ, Fish DR, Smith SJ. Very high frequency rhythmic
activity during SEEG suppression in frontal lobe epilepsy.
Electroencephalogr Clin Neurophysiol. 1992;82(2):155–159.
106. Worrell GA, Parish L, Cranstoun SD, et al. High-frequency oscil-
lations and seizure generation in neocortical epilepsy. Brain.
2004;127(pt 7):1496–1506.
107. Wu JY, Koh S, Sankar R, et al. Paroxysmal fast activity: an inter-
ictal scalp EEG marker of epileptogenesis in children. Epilepsy
Res. 2008;82(1):99–106.
108. Yamazaki M, Chan D, Tovar-Spinoza Z, et al. Interictal epilepto-
genic fast oscillations on neonatal and infantile EEGs in
hemimegalencephaly. Epilepsy Res. 2009;83(2–3):198–206.
109. Kobayashi K, Oka M, Akiyama T, et al. Very fast rhythmic activity
on scalp EEG associated with epileptic spasms. Epilepsia. 2004;
45(5):488–496.
110. Akiyama T, Otsubo H, Ochi A, et al. Focal cortical high-frequency
oscillations trigger epileptic spasms: confirmation by digital video
subdural EEG. Clin Neurophysiol. 2005;116(12):2819–2825.
111. Ramachandrannair R, Ochi A, Imai K, et al. Epileptic spasms in
older pediatric patients: MEG and ictal high-frequency oscilla-
tions suggest focal-onset seizures in a subset of epileptic spasms.
Epilepsy Res. 2008;78(2–3):216–224.
112. Akiyama T, Otsubo H, Ochi A, et al. Topographic movie of ictal
high-frequency oscillations on the brain surface using subdural
EEG in neocortical epilepsy. Epilepsia. 2006;47(11):1953–1957.
113. Gardner AB, Worrell GA, Marsh E, et al. Human and automated
detection of high-frequency oscillations in clinical intracranial
EEG recordings. Clin Neurophysiol. 2007;118(5):1134–1143.
114. Urrestarazu E, Chander R, Dubeau F, et al. Interictal high-fre-
quency oscillations (100–500 Hz) in the intracerebral EEG of
epileptic patients. Brain. 2007;130(pt 9):2354–2366.
115. Kobayashi K, Inoue T, Watanabe Y, et al. Spectral analysis of EEG
gamma rhythms associated with tonic seizures in Lennox-Gastaut
syndrome. Epilepsy Res. 2009;86(1):15–22.
116. Jacobs J, LeVan P, Chander R, et al. Interictal high-frequency
oscillations (80–500 Hz) are an indicator of seizure onset areas
independent of spikes in the human epileptic brain. Epilepsia.
2008;49(11):1893–1907.
117. Crepon B, Navarro V, Hasboun D, et al. Mapping interictal oscil-
lations greater than 200 Hz recorded with intracranial macro-
electrodes in human epilepsy. Brain. 2010;133(pt 1):33–45.
118. Urrestarazu E, Jirsch JD, LeVan P, et al. High-frequency intracere-
bral EEG activity (100–500 Hz) following interictal spikes.
Epilepsia. 2006;47(9):1465–1476.
119. Kobayashi K, Jacobs J, Gotman J. Detection of changes of high-
frequency activity by statistical time-frequency analysis in epilep-
tic spikes. Clin Neurophysiol. 2009;120(6):1070–1077.
120. Jacobs J, Levan P, Chatillon CE, et al. High frequency oscillations
in intracranial EEGs mark epileptogenicity rather than lesion
type. Brain. 2009;132(pt 4):1022–1037.
121. Bagshaw AP, Jacobs J, LeVan P, et al. Effect of sleep stage on interic-
tal high-frequency oscillations recorded from depth macroelec-
trodes in patients with focal epilepsy. Epilepsia. 2009;50(4):617–628.
122. Khosravani H, Mehrotra N, Rigby M, et al. Spatial localization and
time-dependant changes of electrographic high frequency oscilla-
tions in human temporal lobe epilepsy. Epilepsia. 2009;50(4):
605–616.
123. Jacobs J, Zelmann R, Jirsch J, et al. High frequency oscillations
(80–500 Hz) in the preictal period in patients with focal seizures.
Epilepsia. 2009;50(7):1780–1792.
124. Zijlmans M, Jacobs J, Zelmann R, et al. High-frequency oscilla-
tions mirror disease activity in patients with epilepsy. Neurology.
2009;72(11):979–986.
125. Gotman J, Marciani MG. Electroencephalographic spiking activ-
ity, drug levels, and seizure occurrence in epileptic patients. Ann
Neurol. 1985;17(6):597–603.
126. Gotman J, Koffler DJ. Interictal spiking increases after seizures
but does not after decrease in medication. Electroencephalogr Clin
Neurophysiol. 1989;72(1):7–15.
127. Spencer SS, Goncharova, II, Duckrow RB, et al. Interictal spikes
on intracranial recording: behavior, physiology, and implica-
tions. Epilepsia. 2008;49(11):1881–1892.
128. Zijlmans M, Jacobs J, Zelmann R, et al. High frequency oscillations
and seizure frequency in patients with focal epilepsy. Epilepsy Res.
2009;85(2–3):287–292.
129. Bragin A, Wilson CL, Engel J. Spatial stability over time of brain
areas generating fast ripples in the epileptic rat. Epilepsia. 2003;
44(9):1233–1237.
130. Ochi A, Otsubo H, Donner EJ, et al. Dynamic changes of ictal
high-frequency oscillations in neocortical epilepsy: using multi-
ple band frequency analysis. Epilepsia. 2007;48(2):286–296.
131. Jacobs J, Zijlmans M, Zelmann R, et al. High-frequency elec-
troencephalographic oscillations correlate with outcome of
epilepsy surgery. Ann Neurol. 2010;67(2):209–220.
132. Worrell GA, Gardner AB, Stead SM, et al. High-frequency
oscillations in human temporal lobe: simultaneous microwire
and clinical macroelectrode recordings. Brain. 2008;131(pt 4):
928–937.
133. Clemens Z, Molle M, Eross L, et al. Temporal coupling of
parahippocampal ripples, sleep spindles and slow oscillations in
humans. Brain. 2007;130(pt 11):2868–2878.
134. Chatillon C-E. The Effect of Contact Size on High Frequency
Oscillation Detection in Human and Rat Intracerebral EEG
Recording. Montreal: McGill; 2009.
135. Otsubo H, Ochi A, Imai K, et al. High-frequency oscillations of
ictal muscle activity and epileptogenic discharges on intracranial
EEG in a temporal lobe epilepsy patient. Clin Neurophysiol. 2008;
119(4):862–868.
136. Ball T, Kern M, Mutschler I, et al. Signal quality of simultaneously
recorded invasive and non-invasive EEG. Neuroimage. 2009;
46(3):708–716.
137. Engel J Jr, Bragin A, Staba R, et al. High-frequency oscillations:
what is normal and what is not? Epilepsia. 2009;50(4):598–604.
138. Le Van Quyen M, Bragin A, Staba R, et al. Cell type-specific fir-
ing during ripple oscillations in the hippocampal formation of
humans. J Neurosci. 2008;28(24):6104–6110.
139. Dalal SS, Baillet S, Adam C, et al. Simultaneous MEG and intracra-
nial EEG recordings during attentive reading. Neuroimage.
2009;45(4):1289–1304.
140. Ball T, Demandt E, Mutschler I, et al. Movement related activity
in the high gamma range of the human EEG. Neuroimage. 2008;
41(2):302–310.

Intraoperative Evoked
Potential Monitoring
S
ensory evoked potentials (EPs) are the electrical signals
produced by the nervous system in response to a sensory
stimulus. Auditory, visual, and somatosensory EPs are
used as diagnostic tests to noninvasively assess sensory pathways
within the central and peripheral nervous system, as well as to
assess patients for diseases of the ear and eye. Somatosensory
evoked potentials (SEPs) and auditory evoked potentials (AEPs)
have also found wide utility in intraoperative monitoring (IOM)
to detect neural compromise during surgery, when the patient is
anesthetized and a conventional neurologic examination cannot
be performed. The anesthetic sensitivity of visual EPs limits their
utility for IOM.
The central motor tracts can also be monitored by stimulating
them rostrally and recording from either muscles or neural struc-
tures caudal to the part of the nervous system that is at risk. By
analogy to sensory EPs, the signals that are recorded are typically
called motor evoked potentials (MEPs). The brain can be stimu-
lated either electrically or magnetically in order to elicit MEPs. As
discussed in Chapter 52, MEPs to magnetic stimulation are diffi-
cult to record consistently in anesthetized patients during sur-
gery; intraoperative MEP monitoring is therefore typically
performed using transcranial electrical stimulation (TES) of the
brain.
The EP modalities commonly used for IOM are described in
this chapter. For the sensory EPs, the description includes the
anatomical generators of the components that are useful for
IOM. Recording techniques, assessment of the IOM data,
causes of adverse EP changes, and typical intraoperative appli-
cations are described.
FACTORS COMMON TO MULTIPLE
MODALITIES
Recording Techniques
Sensory EPs are in general too small to be seen in the raw data
following a single stimulus. Signal averaging is therefore
employed—multiple stimuli are delivered and the data epochs or
sweeps recorded following the individual stimuli are averaged
together, after rejecting those that contain large artifacts. Signal
averaging (Fig. 38.1) improves the signal-to-noise ratio of the
averaged EP waveforms by a factor equal to the square root of the
number of sweeps included in the average (1). If the EP data are
noisy and/or the EP amplitudes are small, the number of sweeps
per average may have to be increased. Conversely, if the signal-to-
noise ratio of the raw data is favorable, the number of sweeps per
average can be decreased, allowing averages to be acquired more
CHAPTER
ALAN D. LEGATT
38
frequently and thus facilitating more rapid notification to the
surgeons should adverse EP changes occur.
EPs should only be recorded to unilateral stimulation; if
simultaneous bilateral stimulation were employed, the presence
of a normal EP to stimulation of one side might prevent recog-
nition of an abnormal or absent EP to stimulation of the other
side. Modern EP recording systems permit interleaving of sen-
sory stimuli (e.g., auditory stimuli are delivered alternately to the
left and to the right ears). The data sweeps are then sorted (after
artifact rejection) into separate left- and right-sided averages.
Thus, even though left- and right-sided averages are acquired
simultaneously, simultaneous bilateral stimulation is never actu-
ally delivered.
Assessment of IOM Data
Sensory EP amplitudes vary substantially across different sub-
jects, whereas component latencies are much more consistent.
Therefore, interpretation of extraoperative EP studies per-
formed for diagnostic purposes is predominantly based on
latencies (2,3). However, EP component amplitudes on
repeated recordings in the same subject are usually quite consis-
tent if the recording techniques are not altered. Moreover, if
parts of the nervous system are compromised, amplitude
changes may occur earlier than, or in the absence of, latency
changes. Therefore, both the amplitudes and latencies should be
assessed during IOM of SEPs and AEPs.
For extraoperative sensory EP studies performed for diagnos-
tic purposes, the patient’s signals are compared to those
recorded in a control population of normal subjects. In order for
this comparison to be valid, the techniques used and the state of
the subjects must be identical in the patient’s study and in the
control studies. During IOM, the anesthetic regimen, the body
temperature, the delivered stimulus intensity, and other factors
will differ from patient to patient. Therefore, during IOM each
patient serves as his or her own control; EPs recorded during the
time when parts of the nervous system are at risk are compared
to those recorded earlier during the same operation (4). If the
anesthetic regimen or other factors that will affect the EP data
are changed, the baseline amplitude and latency values should be
reassessed.
Causes of Adverse EP Changes
Changes in the signals recorded during IOM can be classified
into three categories: “true positive” changes, which reflect
compromise of the structures that the monitoring is intended
to safeguard (e.g., Figs. 38.2 and 38.3); changes produced by
other physiologic mechanisms such as anesthetic effects (e.g.,
767
768 Part V ■ Complementary and Special Techniques

Figure 38.1 Averaged brainstem auditory evoked potential

(BAEP) waveforms in an awake human subject, showing the
improvement in the signal-to-noise ratio as the number of
epochs in the average is increased. The BAEPs were elicited
by right ear stimulation with rarefaction clicks delivered at a
rate of 11.3 per second; the number of epochs included in
the average is shown to the left of each waveform. Amplifier
bandpass 150 to 3000 Hz, vertex (Cz) to right mastoid
recording. In this and subsequent BAEP figures, Cz positivity
is plotted as an upward deflection.

Figs. 38.4 and 38.5) or hypothermia; and changes due to tech-
nical problems (e.g., Fig. 38.6) rather than to actual changes in
neuronal function (5). True positive changes can reflect tissue
damage due to direct mechanical or thermal injury. They can
also reflect ischemia caused by compromise of the blood supply
to the neural tissue (Figs. 38.2 and 38.3), by vasospasm, by gen-
eralized hypotension, or by a combination of these acting syn-
ergistically.
EPs can be lost due to technical problems such as equipment
malfunction, dislodged electrodes or stimulators (Fig. 38.6), dis-
connected or broken wires, and operator error (the use of incor-
rect protocols or settings). A variety of artifacts can obscure the
EPs (Fig. 38.7). The artifact from the cavitational ultrasonic suc-
tion aspirator (CUSA) device (Fig. 38.7C) deserves special men-
tion, as it may manifest as a low-voltage, high-frequency artifact
that obscures the EP but does not trigger automatic artifact
rejection (4). Monopolar cautery use can saturate the EP record-
ing system’s input amplifiers, preventing recording of valid EP
data for some time after the cautery is stopped.
Anesthesia (Figs. 38.4 and 38.5), systemic hypothermia, and
hypotension can produce EP changes that do in fact reflect
changes in neural function but are different from the neural
dysfunction due to surgical manipulations that the IOM was
specifically intended to detect. Irrigation of the surgical field
with cold fluids can produce localized tissue hypothermia,
causing EP changes that resemble those that would be caused by
surgical compromise of the neural tissue there (Fig. 38.8).
AUDITORY EVOKED POTENTIALS
The earliest electrical signals recorded following a transient
auditory stimulus are generated within the cochlea and are
recorded as part of the electrocochleogram. The subsequent
neurally generated signals have been divided into latency
classes—short- latency AEPs, with latencies of under 10 msec;
long- latency AEPs, with latencies over 50 msec; and middle-
latency AEPs, with intermediate latencies. The long- latency
AEPs are generated within cerebral cortex, including cortical
association areas, and are profoundly affected by the degree to
which the subject is attending to the auditory stimulus and
extracting information from it. The long- latency AEPs are sup-
pressed by surgical anesthesia, and thus are not useful for IOM.
 Chapter 38 ■ Intraoperative Evoked Potential Monitoring 769

Figure 38.2 Consecutive SEP runs recorded to left median nerve stimulation during a right carotid endarterectomy in
a 78-year-old man. There was a delayed attenuation of the cortical SEPs (left) following carotid cross-clamping. The
surgeons were notified and placed a shunt; the SEPs then returned to baseline. The patient was unchanged neurolog-
ically after the surgery. The simultaneously recorded cervicomedullary SEPs (right) did not change, demonstrating that
the left median nerve was still being stimulated and that afferent somatosensory activity was still reaching the level of
the medulla. Vertical calibration: 3 µV/ division. Horizontal calibration: 5 msec/ division. In this and subsequent SEP
figures, negativity at input 1 is plotted as an upward deflection.

Middle- latency AEPs are also generated within cerebral cortex,
including primary auditory cortex and surrounding areas (6).
They are also markedly affected by surgical anesthesia, which
limits their utility for IOM to detect surgery- related focal
neurologic dysfunction. However, this anesthetic sensitivity has
led to attempts to use middle- latency AEPs as an indicator of
the depth of anesthesia (7).
The short-latency AEPs are the most useful AEPs for neuro-
logic and otologic diagnosis because they are relatively easy to
record and their waveforms and latencies are highly consistent
across normal subjects (8). They are also useful for IOM because
surgical anesthesia produces only minor changes in them
(5,9,10); some of the intraoperative changes that are observed in
them may be due more to changes in body temperature than to
anesthetic effects (11–13). Although the earliest components are
generated, at least in part, in the eighth nerve, most of the com-
ponents are generated in the parts of the brainstem auditory
pathways, and the short-latency AEPs are most commonly
referred to as brainstem auditory evoked potentials (BAEPs).
Components, Generators, and Recording
Electrode Locations
BAEP components are typically recorded between the vertex
(position “Cz” of the International 10-20 System) and both ear-
lobes or both mastoids (labeled “Ai” or “Mi” for the ipsilateral
ear and mastoid, respectively, and “Ac” or “Mc” for those con-
tralateral to the stimulated ear); recording the two channels pro-
vides a measure of redundancy in case one of the ear/mastoid
electrodes becomes unusable during surgery and may also help in
the identification of wave V (see next paragraph). Positive BAEP
peaks are most often displayed as upward deflections (the oppo-
site polarity convention from SEPs and from electroencephalog-
raphy and electromyography, in which negativity is displayed as
an upward deflection), and the positive peaks in the Cz-Ai record-
ing channel are typically labeled with Roman numerals according
to the convention of Jewett and Williston (14) (Fig. 38.9). Most of
the BAEP components are recorded from the scalp as far-field
potentials, which means that small displacements of the record-
ing electrodes do not significantly alter the BAEP waveform.
770 Part V ■ Complementary and Special Techniques

Figure 38.3 BAEP and SEP changes due to brainstem ischemia and infarction during surgery for a basilar artery
aneurysm in a 50-year-old woman. Cortical SEPs to stimulation of both median nerves and BAEPs to left ear stimulation
are shown. The aneurysm ruptured during an attempt to place a clip on it and the cortical SEP to left median nerve
stimulation disappeared. A clip was placed on the basilar artery to control the bleeding and the cortical SEP to right
median nerve then disappeared as well. A BAEP run was not obtained between the aneurysm rupture and clipping of
the basilar artery, but following the clipping BAEP wave V (arrow) became delayed and attenuated, then only question-
ably present (arrow with question mark), and eventually disappeared, while BAEP wave IV (triangle) persisted. The
cortical SEP to right median nerve stimulation reappeared and returned to baseline; the cortical SEP to left median nerve
stimulation recovered only partially. The patient suffered multiple brainstem infarcts. In the SEP waveforms, cortical
negativity is shown as an upward deflection. The first 8 msec of each SEP waveform was cropped off to remove the
large stimulus artifacts. (Reprinted from Legatt AD. Mechanisms of intraoperative brainstem auditory evoked potential
changes. J Clin Neurophysiol. 2002;19:396–408.)

Thus, if a CPz electrode is being placed to record lower limb
SEPs, it can be used in place of the Cz electrode to record BAEPs,
with minimal effect on the waveforms obtained. Wave I, how-
ever, is recorded as a near-field potential on the surface of the
head in the vicinity of the stimulated ear.
Waves I, III, and V are the most consistent BAEP peaks, and
those upon which interpretation of both extraoperative BAEP
studies and intraoperative BAEP monitoring is based. Wave I is
recorded as a negativity by the ipsilateral ear/mastoid electrode
and is largely absent in the Cz-Ac recording (see Fig. 38.9).
Wave III is typically seen in both channels, though it is smaller
in Cz- Ac than in the Cz- Ai recording. Waves IV and V are seen
in both channels at similar amplitudes and typically overlap,
forming a IV–V complex with variable morphology (15)
(Fig. 38.10). In the Cz- Ac channel, the latency of wave IV is typ-
ically a little shorter than that in the Cz-Ai channel, and that of
wave V is typically a little longer (see Fig. 38.9). The increased
separation of waves IV and V in the Cz- Ac channel may make
wave V easier to identify in that channel.
Wave I originates in the first volley of action potentials in the
auditory nerve as it begins in the most distal portion of the nerve
(16). It represents the same electrical phenomenon as the N1
component of the eighth nerve compound action potential in
the electrocohleogram (17), and at the skin surface, it appears as
a negativity in a circumscribed area around the stimulated ear
(18,19). The negativity picked up by the Ai electrode gives rise to
the upgoing (positive) peak in the Cz-Ai recording. Since wave I
is recorded as a near-field potential around the stimulated ear,
repositioning of the Ai/Mi recording electrode can substantially
alter it, and alternate electrode positions can be used to obtain a
clearer wave I. An electrode within the external auditory canal
may pick up an even larger wave I. Because wave I arises from the
 Chapter 38 ■ Intraoperative Evoked Potential Monitoring 771

Figure 38.4 Cortical and cervicomedullary SEPs to left median nerve stimulation recorded (simultaneously) during a
left carotid endarterectomy in a 65-year-old man. There were no SEP changes with carotid clamping, but the cortical
SEPs became markedly attenuated after nitrous oxide was turned on during closing. Volume-conducted cortical SEPs
visible in the cervicomedullary SEP waveforms (asterisk) also became attenuated, but the cervicomedullary SEP com-
ponents themselves (arrow) were unchanged. Vertical calibration: 1.5 µV/ division for the cortical SEP waveforms and
0.7 µV/ division for the cervicomedullary SEP waveforms. Horizontal calibration: 5 msec/ division.

most distal portion of the auditory nerve, it may persist after the
nerve is sectioned at a more proximal location, such as during
surgery for eighth nerve tumors (20,21) (Fig. 38.11).
Subsequent BAEP components are the composites of contri-
butions of multiple generators. The complexity of the generators
of human BAEPs derives in part from the anatomy of the brain-
stem auditory pathways, with ascending fibers both synapsing in
and bypassing various relay nuclei (22–24). It also reflects the
presence of at least two bursts of activity in the auditory nerve
(corresponding to the N1 and N2 components of the eighth
nerve compound action potentials in the electrocochleogram),
which can drive the more rostral pathways. Because of both of
these factors, several different structures within the infratentor-
ial auditory pathways may be active simultaneously, and thus
contribute to the same BAEP component (16).
Wave II originates, in part, in the first auditory nerve volley
(the N1 component of the eighth nerve compound action poten-
tial, which gave rise to wave I when it began in the distal eighth
nerve) that has propagated to the proximal end of the auditory
nerve and to the cochlear nucleus. However, this occurs simulta-
neously with the onset of the second auditory nerve volley (the N2
component of the eighth nerve compound action potential) in the
distal nerve (17). The latter contributes to the scalp-recorded
BAEP in the same manner as the N1 component did when it was
at the same location. This can cause persistence of a wave II in
cases where the proximal eighth nerve has been destroyed (8).
Wave III predominantly originates in the caudal pontine
tegmentum, including the region of the superior olivary com-
plex (8). Since ascending projections from the cochlear nucleus
are bilateral, wave III may receive contributions from brainstem
auditory structures both ipsilateral and contralateral to the
stimulated ear. In patients with asymmetrical lesions of the
brainstem, wave III abnormalities are usually most pronounced
following stimulation of the ear ipsilateral to the lesion (25–27),
though occasionally they are more pronounced following con-
tralateral stimulation (28).
The anatomical generators of waves IV and V are most likely
in close anatomical proximity, since they are usually either both
772 Part V ■ Complementary and Special Techniques

Figure 38.5 Consecutive MEP runs recorded from right leg muscles to transcranial electrical stimulation with the
anode on the left in a 19-year-old man having spinal instrumentation and fusion surgery for Scheuermann’s kyphosis.
During closing, the upper two pairs of MEP waveforms were recorded under total intravenous anesthesia using propofol
and sufentanil. The anesthetic regimen was then changed to 2% sevoflurane. The MEPs became attenuated and then
disappeared.

affected or both unaffected by brainstem lesions (28–30). They
may, however, be differentially affected (16,28,31), including by
intraoperative brainstem damage (Fig. 38.12). Wave IV appears
to reflect activity predominantly in ascending auditory fibers
within the dorsal and rostral pons, just caudal to the inferior
colliculus, whereas wave V predominantly reflects activity at the
level of the inferior colliculus, perhaps including activity in the
rostral portion of the lateral lemniscus as it terminates in the
inferior colliculus (8). As is the case with wave III, wave V
abnormalities due to unilateral brainstem lesions are usually
most pronounced following stimulation of the ear ipsilateral to
the lesion (25–27,32,33), though there are exceptions (34,35).
Waves VI and VII are absent in some normal subjects. While
they may in part reflect activity in more rostral structures such
as the medial geniculate nucleus, they also receive contributions
from activity in the inferior colliculus (8); the latter generator
may cause persistence of these waves in patients with auditory
pathway damage rostral to the inferior colliculus. Therefore,
BAEPs cannot be used to assess or monitor the auditory path-
ways rostral to the mesencephalon.
Recording Techniques
Acoustic clicks, produced by delivering trains of 100- µsec-
duration electrical square pulses to the acoustic transducer, are
most often used to elicit BAEPs for IOM; brief tone pips can
also be used. Since the responses to rarefaction and compres-
sion clicks may differ (36,37), extraoperative diagnostic BAEP
studies typically employ a single click polarity. Stimulation with
alternating click polarities and averaging the responses to rar-
efaction and compression clicks together is useful to cancel a
large stimulus artifact and/or the cochlear microphonic and is
often employed during intraoperative BAEP monitoring.
Since headphones would be impractical for IOM, the
acoustic stimuli are typically delivered using ear inserts, incor-
porating foam cylinders that can be compressed and then grad-
ually expand to achieve a tight fit with the ear canal. The foam
can be covered with a thin layer of metal foil, to serve as a near-
field electrode for recording of wave I from the external audi-
tory canal. The ear insert can be held in place with bone wax or
cotton padding, and secured with nonporous tape or an adhe-
sive waterproof dressing pad (4,38). This is to prevent fluids
from entering the ear canal, where they might interfere with
transmission of sound to the inner ear. The ear insert is con-
nected to the acoustic transducer using a segment of plastic
tubing. Care should be taken to ensure that this tubing remains
in place during positioning of the patient and is not kinked. The
time required for the acoustic signal to propagate through the
tubing typically prolongs the latencies of all BAEP components
by approximately 0.9 to 1.0 msec as compared to those recorded
using headphones. This causes no problems in the evaluation of
the BAEPs, since each patient serves as his or her own control
and the acoustic propagation delay is constant. Moreover, the
delay helps to prevent obscuration of wave I by the electrical
stimulus artifact, because (i) it prolongs the latency of wave I,
 Chapter 38
Figure 38.6 Loss of BAEPs due to a technical problem in a patient
undergoing surgery for a left-sided vestibular schwannoma. Consecutive
BAEP recordings (Cz-Ai recording) to right ear stimulation are shown,
with the earliest waveform at the top. During positioning of the patient’s
head, these BAEPs disappeared. Investigation revealed that the tube car-
rying the acoustic stimulus had been pulled out of the patient’s ear. The
tube was replaced and the BAEPs reappeared. The arrow indicates a pos-
sible delayed BAEP peak when the tube was dislodged, reflecting
acoustic stimulus reaching the ear at a reduced intensity. (Reprinted from
Legatt AD. Mechanisms of intraoperative brainstem auditory evoked
potential changes. J Clin Neurophysiol. 2002;19:396–408.)
helping to separate it in time from the electrical stimulus arti-
fact (which remains simultaneous with the activation of the
acoustic transducer), and (ii) it permits increasing the distance
between the acoustic transducer and the recording electrodes,
reducing the amplitude of the electrical stimulus artifact.
The stimulus intensity delivered during intraoperative BAEP
monitoring cannot be precisely controlled, due to variability in
the positioning of the ear insert, but this does not cause a prob-
lem since each patient serves as his or her own control. The
intensity setting chosen should be loud enough to produce a
clear BAEP but not loud enough to cause ear damage. A stimu-
lus rate of approximately 10 per second is typical, but a rate of
exactly 10 Hz or another submultiple of the power line fre-
quency should be avoided; if such a submultiple were chosen,
signal averaging would not eliminate line- frequency artifact
that is present in the data. Contralateral white noise masking,
such as is typically used during extraoperative diagnostic BAEP
studies to prevent acoustic crosstalk, cannot be used during
IOM if left and right ear stimuli are interleaved. This is not a
■ Intraoperative Evoked Potential Monitoring 773
Figure 38.7 Single (unaveraged) data epochs recorded during intraop-
erative BAEP monitoring. A: An electrical stimulus artifact is present at
the beginning of the epoch. The BAEP, which is less than 1 µV in ampli-
tude, is not visible in this single epoch of raw data. B: Bipolar cautery
produces a large artifact that triggers automatic artifact rejection. C: The
cavitational ultrasonic surgical aspirator (CUSA) device produces a very
high-frequency but low-voltage artifact that completely obliterates the
neurophysiologic data but does not trigger automatic artifact rejection.
D: The light source of the operating microscope produces a repetitive,
sharply contoured artifact that recurs at a harmonic of the line frequency
but is composed of higher frequencies that would not be removed by a
line-frequency notch filter. In A, B, and C, the horizontal dotted lines
show the input window of the analog-to-digital converter and the thres -
hold for automatic artifact rejection. In D, the amplifier gain was
reduced to show the light source artifacts in their entirety. Voltage cali-
bration bar: 10 µV in A, B, and C, and 40 µV in D. (Reprinted from Legatt
AD. Mechanisms of intraoperative brainstem auditory evoked potential
changes. J Clin Neurophysiol. 2002;19:396–408.)
problem because acoustic crosstalk is much smaller with ear-
insert transducers than with headphones (39). Also, the major
reason for white noise masking is to prevent acoustic crosstalk
from producing a BAEP when a deaf ear is stimulated during a
diagnostic BAEP study; absence of white noise masking will not
prevent recognition of new auditory pathway compromise
when a functioning ear is stimulated.
Typical filter settings for BAEP recordings are 100 to 150 Hz
for the low-frequency (high- pass) filter and 3000 Hz for the
high- frequency (low-pass) filter. Line- frequency (e.g., 60 Hz)
774 Part V ■ Complementary and Special Techniques

Figure 38.8 BAEPs to right ear

stimulation recorded during surgery
for a right cerebellopontine angle
meningioma. The BAEPs were stable
during cerebellar retraction and
tumor resection. After the tumor had
been removed, copious irrigation of
the surgical field with cold fluids
produced a transient prolongation of
the I-to-III interpeak interval, reflect-
ing slowed neural conduction within
the eighth nerve resulting from
local cooling. The peak latencies of
waves I, III, and V are marked by
the small diamonds, and the clock
times, esophageal temperatures,
and surgical procedures correspon-
ding to each of the BAEP waveforms
are noted at the right. (Reprinted
from Legatt AD. Mechanisms of
intraoperative brainstem auditory
evoked potential changes. J Clin
Neurophysiol. 2002;19:396–408.)

“notch” filters can be used for BAEP recordings. The analog
gain depends on the input window of the analog- to- digital con-
verter; a value of approximately 100,000 is typical.
While an averaging epoch duration of 10 msec is often used
for extraoperative diagnostic BAEP recordings in adults, a longer
epoch duration, typically 15 msec, should be used during intra-
operative BAEP monitoring because component latencies can be
prolonged by preexisting pathology, hypothermia, or intraoper-
ative compromise of the auditory system. The choice of the
number of sweeps per average will depend on the signal-to-
noise ratio of the raw data. A value of 1000 to 2000 sweeps per
average is typical.
If the risk is to the eighth nerve (e.g., during surgery for an
intracanalicular eighth nerve tumor) and the surgical exposure
permits placement of a recording electrode on the proximal
eighth nerve, a near- field eighth nerve compound action poten-
tial that is typically much larger than the far- field BAEP can be
recorded (4,40) (Fig. 38.13). This permits averaging using fewer
sweeps, facilitating more frequent assessment and more rapid
notification to the surgeons should adverse changes occur.

Assessment of IOM Data

Figure 38.9 Brainstem auditory evoked potentials recorded simultane-
ously from three different recording electrode linkages following monaural
stimulation in a normal subject. The vertical dashed lines indicate the
peak latencies of waves IV and V in the Cz-Ai waveforms; these peaks are
more widely separated in the Cz-Ac waveforms. (Reprinted from Legatt
AD. Brainstem auditory evoked potentials: methodology, interpretation,
and clinical application. In: Aminoff MJ, ed. Electrodiagnosis in Clinical
Neurology. 5th ed. New York, NY: Churchill Livingstone; 2005:489–523.)
The peak latencies of waves I, III, and V are measured, and the
amplitude of each of these components is measured with
respect to the trough that follows it. The central transmission
time, the interval between the peaks of waves I and V, can also
be calculated. Typical alarm criteria are a 50% drop in compo-
nent amplitudes and a 1 msec increase in component latencies
or in the central transmission time.
Loss or delay of wave V with preservation of waves I and III
(Fig. 38.12) reflects compromise of the auditory pathways within
the brainstem, between the lower pons and the mesencephalon.
If the BAEP changes are unilateral or asymmetrical, the predom-
inant pathology is most often, though not always, ipsilateral to
the ear whose stimulation gives the most abnormal BAEPs.
 Chapter 38 ■ Intraoperative Evoked Potential Monitoring 775

Figure 38.11 Intraoperative BAEPs to left ear stimulation recorded dur-

ing surgery for left vestibular schwannomas in two different patients,
showing persistence of wave I (arrows) after transection of the intracra-
nial eighth nerve. The nerves were intentionally sacrificed to permit
total resection of the tumors. (Reprinted from Legatt AD. Mechanisms
of intraoperative brainstem auditory evoked potential changes. J Clin
Neurophysiol. 2002;19:396–408.)

Internal auditory artery compromise probably accounts for

most of the cases where all BAEP components, including wave I,
Figure 38.10 Various IV–V complex morphologies in Cz-Ai wave-
are suddenly lost (Fig. 38.14), during surgery for cerebellopon-
forms recorded in normal subjects. (Reprinted from Chiappa K,
tine angle tumors (42).
Gladstone KJ, Young RR. Brain stem auditory evoked responses. Studies
When the cerebellum is retracted to gain access to the cere-
of waveform variations in 50 normal human subjects. Arch Neurol.
bellopontine angle (e.g., for a cerebellopontine angle tumor or
1979;36:81–87. Copyright © 1979 American Medical Association. All
during microvascular decompression surgery), this also moves
rights reserved.)
the brainstem away from the internal auditory meatus and
Loss or delay of waves III and V with preservation of wave I
(Fig. 38.11) reflects compromise of the eighth nerve (though
not of its most distal part, within the modiolus) or of the lower
pons. If the BAEP changes are unilateral or asymmetrical, the
predominant pathology is again most often ipsilateral to the ear
whose stimulation gives the most abnormal BAEPs.
Loss or delay of wave I and subsequent waves (Fig. 38.14), if
not due to technical problems, reflects cochlear dysfunction.
This may include cochlear ischemia or infarction due to com-
promise of the blood supply to the cochlea.

Causes of Adverse EP Changes

As previously noted, true positive BAEP changes may reflect
direct mechanical or thermal tissue injury as well as
ischemia/infarction. Two situations specifically pertaining to
BAEPs deserve special mention: internal auditory artery com-
promise and eighth nerve stretch.
The cochlea receives its blood supply from the intracranial
circulation via the internal auditory artery. This vessel, which is
usually a branch of the anterior inferior cerebellar artery, passes
through the internal auditory canal alongside the eighth nerve
(41). Damage to this artery will cause cochlea ischemia or infarc-
tion, affecting wave I and all subsequent BAEP components.
Figure 38.12 Intraoperative BAEPs to left ear stimulation recorded
during the surgery of Figure 38.2. A: Clear waves I through VI were
present in these BAEPs, recorded just before the aneurysm ruptured.
B: BAEPs recorded after the aneurysm had been clipped and the basilar
artery had been reopened show a loss of waves V and VI; waves I
through IV were unaffected. (Modified from Legatt AD, Arezzo JC,
Vaughan HG Jr. The anatomic and physiologic bases of brain stem audi-
tory evoked potentials. Neurol Clin. 1988;6:681–704.)
776 Part V ■ Complementary and Special Techniques
Figure 38.13 BAEPs recorded from a platinum pad electrode placed
on the intracranial eighth nerve (top) compared to surface-recorded
BAEPs (bottom) during surgery in a 52-year-old woman with a left-
sided intracanalicular acoustic neuroma. Note the voltage calibrations;
the near-field response is considerably larger. The patient had transient
BAEP changes during the resection but BAEPs were at baseline at its
end, and postoperative hearing was normal. (Courtesy of Dr. Timothy
A. Pedley.)
stretches the eighth nerve, which can cause hearing loss. BAEP
monitoring can be used to detect excessive eighth nerve stretch
(Fig. 38.15) and to notify the surgeons that the retraction needs
to be reduced or readjusted. Uncorrected excessive traction on
the eighth nerve can lead to avulsion of the distal nerve from
the cochlea (43), causing hearing loss.
Drilling of bone will produce high noise levels in both ears
via bone conduction and can alter the BAEPs due to acoustic
masking (44), simulating cochlear dysfunction. This would rep-
resent a change due to physiologic factors; it is not a “true pos-
itive” change reflecting compromise of the auditory system. It is
best to pause BAEP recording during drilling.
Typical Intraoperative Applications
BAEPs can be used for IOM of the ears, the eighth nerves, and
the infratentorial auditory pathways up through the level of the
mesencephalon. They are not useful for monitoring the thala-
mus (medial geniculate) or the auditory pathways within the
cerebrum. They are most often used to monitor surgery for
eighth nerve tumors such as vestibular schwannomas (formerly
called acoustic neuromas) and for tumors or vascular abnor-
malities within the posterior fossa, both extra- axial and within
the brainstem. BAEPs are also useful to detect excessive eighth
nerve stretch, which can lead to hearing loss, from cerebellar
retraction during surgery in the cerebellopontine angle, such as
microvascular decompression procedures (45).
Figure 38.14 Consecutive BAEPs to left ear stimulation (earliest wave-
form at the top) recorded in a patient undergoing surgery for a left
acoustic neuroma. A clear wave I and a poorly formed wave V were ini-
tially present and were stable during the initial dissection, but all BAEP
components disappeared simultaneously during dissection within the
internal auditory canal and remained absent through the end of the
operation. This was most likely due to interruption of the blood supply
to the cochlea via the internal auditory artery. (Reprinted from Legatt
AD. Mechanisms of intraoperative brainstem auditory evoked potential
changes. J Clin Neurophysiol. 2002;19:396–408.)
SOMATOSENSORY EVOKED POTENTIALS
While SEPs can be elicited by mechanical stimulation, for extra-
operative diagnostic studies and for IOM SEPs are elicited by
electrical stimulation of mixed peripheral nerves, most often the
median and ulnar nerves in the upper extremities and the
peroneal and posterior tibial nerves in the lower extremities
(Fig. 38.16). Components generated in the peripheral nerves, the
spinal cord, the brainstem, and in cerebral cortex can be identi-
fied. SEP components are typically named by their scalp polarity
and their average latency in unanesthetized normal subjects; for
example, N20 is a negativity with a typical latency of 20 msec.
Upper extremity SEPs are most often elicited by stimulation of
the median and ulnar nerves just proximal to the wrist. Median
nerve SEPs are most often used for extraoperative diagnostic SEP
studies, because of the common occurrence of a degree of ulnar
neuropathy at the elbow; carpal tunnel syndrome will not affect
 Chapter 38 ■ Intraoperative Evoked Potential Monitoring 777

Figure 38.15 Intraoperative BAEPs to

right ear stimulation recorded during sur-
gery for a right acoustic neuroma, showing
two runs recorded prior to (top) and after
(bottom) retraction of the cerebellum. The
most prominent change in the BAEPs was
an increase in the I–III interpeak interval of
more than 1 msec, reflecting stretching of
the eighth nerve. The smaller change in the
III–V interpeak interval may reflect effects
of the retraction on the brainstem.
(Reprinted from Legatt AD. Mechanisms of
intraoperative brainstem auditory evoked
potential changes. J Clin Neurophysiol.
2002;19:396–408.)

Figure 38.16 Examples of the four most common SEPs used for IOM, recorded in four different patients. Data from three
consecutive runs are superimposed. The arrowheads indicate the primary cortical SEPs (upper waveforms in each pair)
and the cervicomedullary SEP components (lower waveforms in each pair). Vertical calibration bar = 1 µV for the cortical
SEPs to peroneal nerve stimulation, 2 µV for the cortical SEPs to median nerve and to posterior tibial nerve stimulation,
and 1.5 µV for all of the other waveforms.
778 Part V ■ Complementary and Special Techniques
them because the stimulation site is proximal to the wrist.
Sensory fibers of the median nerve gain access to the spinal cord
through the C6 and C7 spinal roots (46). If the spinal cord path-
ways were compromised at a level below C6, the afferent activity
traversing the C6 nerve root could generate SEPs and mask the
presence of more caudal spinal cord dysfunction. Thus, median
nerve SEPs by themselves are not adequate for monitoring the
spinal cord below the C6 level. Similarly, ulnar nerve SEPs are not
adequate for monitoring the spinal cord below the C8 level,
because sensory fiber of the ulnar nerve gains access to the spinal
cord through the C8 and T1 nerve roots (46).
In the lower extremity, the common peroneal nerve is typi-
cally stimulated at the level of the knee, behind the head of the
fibula. The posterior tibial nerve is stimulated at the ankle,
behind the medial malleolus; this permits recording of the
peripheral nerve compound action potential behind the knee,
which can help in the determination of the cause of a change in
the more rostrally recorded SEPs. Other advantages of posterior
tibial nerve SEPs as compared to peroneal nerve SEPs are that
posterior tibial nerve SEPs are typically larger (47), that stimu-
lating electrodes at the ankle are usually easier to reach and
replace than those at the knee should they become dislodged
during surgery, when the patient is draped, and that if the
patient is not pharmacologically paralyzed, posterior tibial
nerve stimulation causes less patient movement than does per-
oneal nerve stimulation. If posterior tibial nerve SEPs cannot be
used due to peripheral neuropathy, the anatomy at the ankle, or
a distal amputation, then peroneal nerve SEPs offer another way
to monitor the afferent somatosensory pathways from the legs.
Similar to auditory EPs, a complex series of somatosensory
EPs with latencies ranging from a few milliseconds to hundreds
of milliseconds can be recorded following a single peripheral
nerve shock. The shorter-latency SEP components, generated in
the peripheral nerves, the spinal cord, and at the level of the cer-
vicomedullary junction, are relatively insensitive to anesthetic
effects. The primary cortical SEP components—N20 to stimula-
tion of the median or ulnar nerve at the wrist, P37 to stimulation
of the posterior tibial nerve at the ankle, and P27 to stimulation
of the peroneal nerve at the knee—are affected more by anesthe-
sia (Fig. 38.4), but can be recorded and used for IOM if the anes-
thetic regimen is not excessive. Longer-latency SEP components
are even more attenuated by anesthesia and may be completely
suppressed, which limits their utility for IOM. Thus, SEPs can be
used to monitor the central somatosensory pathways only up
through the level of primary somatosensory cortex.
Components, Generators, and Recording
Electrode Locations
For extraoperative diagnostic studies of upper limb SEPs, the
signal generated by afferent peripheral nerve volley as it trav-
erses the brachial plexus is typically recorded by an electrode at
Erbs’ point as the N9 component. This can also be done during
IOM, but in some cases Erbs’ point is too close to the surgical
field, or the positioning of the patient may preclude secure
placement of an electrode there. The afferent peripheral volley
can also be recorded by an electrode near the nerve at, or just
proximal to, the elbow; use of a reference electrode on the lat-
eral or medial aspect of the limb, at the same level, maximizes
in- phase cancellation of the electrocardiogram (4). The periph-
eral nerve volley to posterior tibial nerve stimulation is
recorded from a midline electrode behind the knee, referred to
either the lateral or medial aspect of that knee or to the midline
electrode on the other knee.
Components originating in the gray matter of the spinal
cord at the level of root entry, N13 following median nerve
stimulation and N22 following posterior tibial nerve stimula-
tion, are used during clinical SEP studies (3) but are of limited
utility during IOM. They are often small and difficult to record,
their recording sites may be too close to the surgical field, and
they are generated caudal to the central nervous system struc-
tures that the IOM is intended to safeguard.
Activity in the afferent somatosensory pathways at the level
of the dorsal column nucleus/cervicomedullary junction is typ-
ically recorded as a far- field potential, P14 for upper limb stim-
ulation and P31 for posterior tibial nerve stimulation (Fig.
38.16), with the positivity of the equivalent dipole field directed
upwards and anteriorly and the negativity directed downwards
and posteriorly. These cervicomedullary far- field potentials are
best recorded between an electrode in the area where the dipole
field is positive and one in which the dipole field is negative. In
extraoperative diagnostic SEP studies, these components are
typically recorded between the dorsal scalp and a noncephalic
electrode (3). During IOM, recording electrode pairs with such
large interelectrode distances tend to pick up more noise. A
forehead- to- inion recording linkage can also pick up the SEP
components generated at the level of the cervicomedullary
junction; if the inion is within the surgical field, a forehead- to-
earlobe recording linkage can also be used (48).
The primary cortical SEP to upper limb stimulation, N20, is
recorded from the scalp over the parietal region contralateral to
the stimulated arm, for example, at positions CP3 and CP4 of the
International 10-10 System (49) for right arm and left arm stim-
ulation, respectively. The reference can be either the electrode
over the other parietal region (CPi, ipsilateral to the stimulus) or
an electrode on the forehead. Recordings to a forehead reference
may also contain the N18 component, which is generated in sub-
cortical white matter structures. If SEPs are being used to moni-
tor the cerebrum, a CPc-Cpi recording linkage can be used to
cancel the N18 component, which has a symmetrical distribution
over the top of the head (50), to ensure that the SEPs being mon-
itored are entirely of cerebral origin. However, if SEPs are being
used to monitor the spinal cord, a forehead reference may give a
somewhat larger SEP waveform, and is acceptable because N18
and N20 would both be affected by spinal cord dysfunction.
The scalp topographies of the primary cortical SEPs to lower
limb stimulation—P37 following posterior tibial nerve stimula-
tion at the ankle and P27 following peroneal nerve stimulation at
the knee—vary from one subject to another because of anatom-
ical variability in the location and orientation of the foot area of
primary somatosensory cortex (51). These components are typi-
cally largest in the midline or over the parietal area ipsilateral to
the stimulated limb—a so-called “paradoxical” localization
(Fig. 38.17). Preoperative SEPs studies can be performed to
determine the optimal location for recording electrodes for IOM
 Chapter 38 ■ Intraoperative Evoked Potential Monitoring 779

Figure 38.17 Extraoperative recordings of cortical SEPs to posterior tibial nerve stimulation during diagnostic SEP testing
in two different patients, and P37 generator locations and equivalent dipole orientations that would explain them. A P37
component that is largest in the midline (A) is explained by an almost vertically oriented P37 dipole (B). A P37 component
that is “paradoxical,” largest over the hemisphere ipsilateral to the stimulated nerve and inverted over the contralateral hemi-
sphere (C), is explained by a generator in the mesial wall of the hemisphere that produces a horizontally oriented P37 dipole
(D). Note that an electrode at CPz would not pick up a P37 component that is suitable for IOM in the example shown
in C. The larger peak that follows P37 has a wider distribution; suppression by anesthesia makes it not useful for IOM.

of lower limb SEPs in each individual patient, or electrodes can
be placed at multiple recording positions (e.g., CP3, CPz, and
CP4) to permit adequate recording of the cortical SEPs in all
patients no matter what the scalp topographies of their cortical
SEPs are.
The primary cortical SEP to median or ulnar nerve stimula-
tion (N20) is generated in the posterior bank of the central sul-
cus (Fig. 38.18) and thus inverts in polarity across the central
sulcus in recordings from the cortical surface (52–54). This
inversion can be used to identify the central sulcus during resec-
tive surgery (55–57). The SEPs are typically recorded from metal
disk electrodes embedded in a sheet of silastic plastic that is
placed on the exposed cortical surface after the dura is opened.
Electrode strips containing several electrodes arranged in a lin-
ear array can be used, but use of an electrode grid containing a
rectangular grid of electrodes may make identification of the
central sulcus easier and faster (58). SEPs recorded from the
cortical surface may also be used to monitor for compromise of
cortical somatosensory tracts during resection of cerebral
lesions or during surgery for aneurysms or vascular malfor-
mations when the craniotomy opening permits placement of
cortical surface electrodes.
Recording Techniques
SEPs are elicited using a pair of stimulating electrodes over each
tested peripheral nerve, with the cathode proximal to prevent
the possibility of anodal block of the afferent signal. Either sur-
face electrodes or subdermal needle stimulating electrodes can
be used. The spacing between the surface electrodes, or between
the skin entry points of needle electrodes, should be about 3
cm. If subdermal needles are used, they should be directed away
from each other; otherwise, the tips could be too close to each
other and the stimulating current could travel between them
without effectively stimulating the underlying nerve. If surface
cup electrodes with electrode gel are used, care should be taken
to ensure that the gel does not leak out and form a salt bridge
between the stimulating electrodes, since this would shunt the
stimulus current between them. Were this to happen, the nerve
780 Part V ■ Complementary and Special Techniques
Figure 38.18 Diagram showing generation of the N20/ P20 component
of the median nerve SEP in primary somatosensory cortex located in the
posterior bank of the central sulcus (shaded area). The arrow represents
the equivalent voltage dipole produced by activation of this area of cortex.
(Reprinted from Legatt AD, Kader A. Topography of the initial cortical
component of the median nerve somatosensory evoked potential: relation-
ship to central sulcus anatomy. J Clin Neurophysiol. 2000;17:321–325.)
would not be stimulated but the stimulating equipment would
not indicate an open- circuit condition (4).
During extraoperative SEP studies, the stimulus intensity is
typically adjusted to just above the threshold level necessary to
obtain a consistent muscle twitch, so as to not cause undue
patient discomfort. During IOM, higher stimulus intensities are
used, to give a safety margin should the efficacy of peripheral
nerve stimulation decrease during the surgery due to technical
factors, such as the development of edema that would increase
the distance between the nerve and the stimulating electrodes.
Peripheral nerve stimulation is typically accomplished using
trains of 200- µsec- duration constant-current stimulus pulses.
Stimulus rates of 2 to 8 per second have been used for SEP
monitoring, but rates that are exact submultiples of the line fre-
quency (e.g., 5 Hz or 6 Hz in North America) should be
avoided; if such a submultiple were chosen, signal averaging
would not eliminate line- frequency artifact that is present in
the data. Typical filter settings are 5 to 30 Hz for the low-
frequency (high- pass) filter and 1000 to 3000 Hz for the high-
frequency (low- pass) filter. Line- frequency (e.g., 60 Hz)
“notch” filters should not be used for SEP recordings because
they can cause a “ringing” oscillatory artifact (59).
The epoch or sweep duration used for averaging should be at
least twice the latency of the longest EP component that is of
interest. A typical epoch duration for recording of upper limb
SEPs is 50 msec; epoch durations in the 80- to 100-msec range
are appropriate for recording of lower limb SEPs. The choice of
the number of sweeps per average will depend on the signal- to-
noise ratio of the raw data and may range from 200 to 1000.
SEPs recorded from the cortical surface are typically much
larger than those recorded from the scalp, and the number of
sweeps per average can be smaller.
In some patients with spinal cord lesions, the volleys along
the afferent pathways may be sufficiently desynchronized that
the cortical and cervicomedullary SEP components are not
identifiable. In such patients, recording electrodes placed
directly on the spinal cord rostral to the lesion may pick up sig-
nals that are sufficiently reproducible to be used for IOM (4).
Assessment of IOM Data
The peak latencies of the peripheral nerve potentials (if appli-
cable), the cervicomedullary far- field potentials, and the pri-
mary cortical SEPs are measured, and the amplitude of each of
these components is measured with respect to the following
peak of the opposite polarity. Depending on the configuration
of the SEP and the degree to which the SEP waveform is con-
taminated by lower- frequency noise, it may also be useful to
measure the amplitude of the SEP peak with respect to the pre-
ceding peak of the opposite polarity or with respect to the base-
line level of the waveform preceding the SEP. Typical alarm
criteria are a 50% drop in component amplitudes and a 10%
increase in component latencies.
If the rostrally recorded SEPs are lost, recording of the
peripheral nerve compound action potential permits immedi-
ate categorization of the cause of the SEP change. If the periph-
eral nerve compound action potential also disappears, the cause
of the SEP change is most likely due to technical factors leading
to inadequate stimulation of the peripheral nerve, though the
possibility of mechanical or ischemic compromise of the
peripheral nerve in the limb must also be considered. If the
peripheral nerve compound action potential persists but the
rostral SEPs disappear, the cause is most likely compromise of
the afferent somatosensory pathways due to surgical manipula-
tions, and the surgeons should be notified accordingly.
If the region of the neuraxis that is at risk is caudal to the cer-
vicomedullary junction (e.g., the spinal cord), then both the
primary cortical SEPs and the cervicomedullary far- field SEP
components are generated rostral to the region that is at risk,
and either may serve to monitor the condition of the spinal
cord. If, however, the part of the nervous system that is at risk
during the surgery is above the cervicomedullary junction, then
the subcortical far- field potentials can be used to verify that the
peripheral nerves are being adequately stimulated and that the
afferent activity has reached the level of the cervicomedullary
junction, but the cortical potentials must be examined to assess
the region of the brain or brainstem that is at risk during the
operation (Fig. 38.2).
Cortical SEPs are affected by surgical anesthesia, and
increases in the anesthetic regimen can cause cortical SEP
changes resembling those due to surgical compromise of neural
structures (compare Figs. 38.2 and 38.4). If the anesthetic regi-
men is changed substantially, the baseline amplitude and latency
levels for the patient’s SEPs may have to be reassessed. High
doses of inhalational agents, especially high doses of multiple
agents, are undesirable. Even more important than avoiding
high anesthetic doses is avoiding large changes in the anesthetic
regimen (i.e., a sudden increase in the concentration of inhala-
tional anesthetic or a bolus dose of an intravenous anesthetic
agent), particularly around the time of critical maneuvers such
 Chapter 38 ■ Intraoperative Evoked Potential Monitoring 781

as positioning of the head and neck in a patient with cervical
spinal stenosis or instability, clamping of cerebral vessels, or dis-
traction during spinal deformity surgery (4).
Causes of Adverse EP Changes
As with all IOM, true positive SEP changes may reflect direct
mechanical or thermal tissue injury as well as ischemia/
infarction of the tissue due to compromise of its blood supply
(Figs. 38.2 and 38.3).
During aortic surgery with retrograde bypass via the femoral
artery, anterograde blood flow in the femoral artery is blocked,
and the peripheral nerve in that leg will become ischemic. This
may lead to disappearance of all SEP components, including the
peripheral nerve components, to stimulation of that leg follow-
ing the placement of the femoral artery catheter. The absence of
the peripheral nerve volley would preclude IOM of the central
somatosensory pathways from that leg.
Compression of the peripheral nerve in a limb or limb
ischemia due to positioning of the limb may also lead to atten-
uation or disappearance of the SEPs to stimulation of a nerve in
that limb. Recording of the peripheral nerve SEP in that limb
may help to recognize that situation. If peripheral nerve SEPs
were not recorded, investigations to determine the cause of an
adverse change in the rostral SEPs should include assessment of
the stimulated limb.
If the constant- current SEP stimulator is unable to deliver
the preset current level, most contemporary IOM equipment
will indicate a high- impedance or open- circuit stimulus condi-
tion, which can help to identify technical problems with stimu-
lation such as broken wires or dislodged electrodes. If, however,
the paired stimulating electrodes are short- circuited together, as
by a salt bridge, then the nerve will not be stimulated effectively
and the SEPs will deteriorate, but the equipment will not indi-
cate the error condition.
As previously mentioned, an increase in the anesthetic con-
centration can cause attenuation and latency prolongation of
the cortical SEPs that resembles SEP changes caused by
somatosensory pathway compromise. If the region of the neu-
raxis at risk is caudal to the cervicomedullary junction and cer-
vicomedullary far- field SEPs adequate for IOM are obtained,
they can be used to determine whether the changes in the cor-
tical SEPs are due to anesthesia or to surgical manipulations.
Typical Intraoperative Applications
SEPs can be used to monitor the condition of the central
somatosensory pathways during surgery on the brain, brainstem,
or spinal cord that places those structures at risk or during ortho-
pedic surgery that poses a risk to the spinal cord (Fig. 38.19).
Neural structures are at risk of ischemic injury during vascular
surgery on the aorta, the carotid arteries, and during surgery for

Figure 38.19 Consecutive SEP runs recorded during spinal instrumentation and fusion surgery for congenital kyphosco-
liosis with multiple hemivertebrae in a 15-year-old boy; only the cortical SEPs are shown. The lower limb SEPs and MEPs
disappeared bilaterally during placement of pedicle screws; upper limb SEPs and MEPs remained present. The surgeons
were notified and the most recently placed screws were removed. A wake-up test was performed and the patient had left
leg weakness. The patient was reanesthetized and the remaining hardware was removed. Postoperatively the patient had
motor and sensory deficits in the left leg, with subsequent improvement. Shown here are the cortical SEPs from the last
four runs recorded before the adverse changes (top) and the first four runs recorded after them (bottom). The arrows indi-
cate the primary cortical SEP components (P37 for posterior tibial nerve stimulation and N20 for ulnar nerve stimulation).
Calibrations: 10 msec/ division and 1.5 µV/ division for posterior tibial nerve SEPs, 5 msec/ division and 2 µV/ division for
ulnar nerve SEPs.
782 Part V ■ Complementary and Special Techniques

aneurysms or arteriovenous malformations of vessels that supply
the brain and spinal cord.
During surgery on the thoracic and lumbar spine, the
brachial plexus or nerves in the arms may be stretched due to
the positioning of the patient’s arms. Monitoring of upper limb
SEPs can be used to detect nerve stretch during surgery (60,61),
permitting intraoperative correction of the arm positioning to
reduce postoperative morbidity.
Afferent somatosensory activity is carried rostrally, within
the spinal cord, in both the dorsal columns and the spinothala-
mic tracts. However, due to the slower conduction velocities
and temporal dispersion within the spinothalamic tracts, the
activity in them does not contribute significantly to the earliest
SEP components generated in the brain. The rostral SEPs used
for IOM are mediated almost entirely by the dorsal columns
(51,62), and therefore directly assess only that portion of the
spinal cord, which is supplied by the posterior spinal arteries.
Since intraoperative spinal cord damage during spinal surgery
may be on a vascular basis (63), SEP monitoring may fail to
detect ischemic spinal cord compromise that involves the ante-
rior spinal artery territory but spares the dorsal columns; SEPs
may also remain unchanged following mechanical trauma lim-
ited to the anterior spinal cord. SEP monitoring may therefore
fail to detect damage that is limited to the motor tracts in the
anterolateral funiculus of the spinal cord (64–66). Fortunately
such “false- negative” SEP studies are rare; surgical situations
that compromise the anterior spinal cord usually compromise
the posterior spinal cord as well. MEP monitoring was devel-
oped to directly monitor the central motor pathways.
Even if MEPs are employed during surgery on the spinal
cord, SEPs should still be monitored. Because of the anatomical
separation between the somatosensory and motor pathways
and their different blood supplies, the somatosensory pathways
can be damaged without effects on the motor pathways (67,68),
and SEPs may show changes in the absence of MEP changes
during IOM (69).
MOTOR EVOKED POTENTIALS
MEP monitoring is used to directly assess the motor pathways
within the brain and spinal cord during surgery. If the dura is
opened and motor cortex is accessible, MEPs can be elicited by
direct cortical stimulation, but most often the brain is stimu-
lated through the intact skull using TES. As discussed in
Chapter 52, transcranial magnetic stimulation is of limited use
in anesthetized patients.
Stimulation of the spinal cord rostral to the region that is at
risk coupled with recording of responses from peripheral
nerves in the legs has been proposed as another way of moni-
toring the spinal cord motor tracts (70), but the responses
obtained in this way predominantly reflect retrograde conduc-
tion within somatosensory fibers in the dorsal columns (71)
and therefore, like SEPs, they may fail to detect intraoperative
damage to the corticospinal tracts (72). Responses recorded
from muscles may be mediated both by descending motor
tracts and by retrograde conduction in somatosensory tracts
with reflex connections to the alpha motor neurons (73–75);
thus, they do not convey the desired specificity. The only way to
be sure that the signals being monitored are mediated solely by
the motor tracts within the spinal cord is to stimulate rostral to
the somatosensory synapses in the dorsal column nuclei (76).
Components, Generators, and Recording Electrodes
MEPs to TES can be recorded from the spinal cord or from mus-
cles. Myogenic MEPs (“M-waves”) (Fig. 38.20, left) are more sen-
sitive to anesthetic effects than are the other EPs commonly used
for IOM (Fig. 38.5) due to anesthetic effects at the anterior horn
cell synapse, and total intravenous anesthesia (e.g., a combination

Figure 38.20 M-waves and D-waves

recorded during resection of a recurrent
glioma of the cervical spinal cord in a
22-year-old woman. Left: Consecutive
recordings of M-waves in right hand mus-
cles to multiple-pulse TES with the anode
on the left. Right: D-waves recorded from
an epidural electrode placed over the
spinal cord caudal to the tumor to single-
pulse TES with the anode on the left.
D-waves were recorded more frequently
than M-waves; the D-wave runs included
in this figure are those that were recorded
closest in time to the M-waves that are
shown. M-waves are single sweeps;
D-waves are averages of the responses to
25 stimuli. Tumor resection was com-
pleted at 10:34.
 Chapter 38
of propofol and narcotic infusions) with avoidance of any inhala-
tional agents may be required to record MEPs in some patients.
M-waves will also be attenuated or eliminated by neuromuscular
blockade (NMB). They may be recordable under partial NMB,
but that NMB should be maintained by a continuous infusion of
paralytic drug rather than by intermittent bolus doses (77).
MEPs recorded from the spinal cord ( “D-waves”) (Fig.
38.20, right) are relatively insensitive to anesthetic effects, since
there are no intervening synapses between the site of corti-
cospinal tract stimulation in the brain and the spinal cord
recording site. They are also more consistent from run to run
and can be recorded in some patients in whom M- waves are not
identifiable due to preexisting pathology or anesthetic effects.
However, invasive electrodes must be placed close to the spinal
cord to record them, and they may miss unilateral deficits if the
motor tracts in the brain are stimulated bilaterally (78). Also,
M- waves can be used to monitor the lower spinal cord where
D- waves are not usable, may show adverse changes earlier than
D- waves when the motor tracts are compromised, and may be
monitorable in some patients in whom preexisting spinal cord
pathology has desynchronized the descending corticospinal
tract volley so that there are no identifiable D- waves (79).
M- waves can be recorded from surface or needle electrodes
placed in or over the muscles of interest, and reflect the com-
pound motor action potentials elicited within the muscles.
D- waves can be recorded using an epidural recording electrode
placed percutaneously via a Touhey needle, or by electrodes
placed within the surgical field close to the spinal cord. They are
generated by propagating action potentials within corticospinal
tract axons.
Recording Techniques
To record M- waves to TES, a high- intensity multipulse stimula-
tor is used to produce multiple closely spaced descending vol-
leys in the corticospinal tract, so that temporal summation of
the excitatory postsynaptic potentials that they produce can fire
the anterior horn cells (see Chapter 52). Typical train parame-
ters (69) are 3 to 7 stimulus pulses per train and interpulse
intervals of 2 to 4 msec within the train. Wide-open filters (e.g.,
5 to 3000 Hz) permit recording of both lower and higher fre-
quency components of the MEP waveform.
Only a small percentage of the alpha motor neuron pool is
activated with each TES stimulus, and the subset of alpha motor
neurons that are activated varies from run to run. Therefore, M-
waves typically vary substantially in waveshape from trial to trial
(Fig. 38.20, left), and signal averaging should not be employed.
Nor is it required, since the single-trial myogenic MEPs are
typically quite large, often hundreds of microvolts in amplitude.
D-waves are much smaller, and signal averaging of a small
number of sweeps (e.g., 5 to 25) is useful to obtain a clearer and
more consistent waveform. Since there is no intervening synapse,
pulse train stimulation is not used. This makes it less likely that
the stimulation will cause undesirable patient movement.
D-waves can be elicited by stimulation of subcortical white
matter (80); therefore, M-waves to multipulse stimulation can
also be elicited by stimulation of subcortical white matter. As the
TES stimulus intensity is increased, intraparenchymal current
■ Intraoperative Evoked Potential Monitoring 783
densities capable of stimulating corticospinal tract axons are
produced further and further away from the surface of the head
and may be able to stimulate the motor pathways as far caudally
as the medulla (81). During IOM of the brainstem and of the
corticospinal tracts within the cerebrum, this could prevent
recognition of surgery-related motor tract compromise if the
motor tracts were stimulated caudal to that dysfunction.
Therefore, when monitoring these surgeries, the TES stimulus
intensity should be kept relatively low, for example, to a level that
produces MEPs in one limb but not in multiple limbs.
Assessment of IOM Data
D- waves are highly consistent from run to run (Fig. 38.20,
right) and are evaluated by criteria similar to those used for sen-
sory EPs, for example, a 50% decrease in amplitude or a 10%
increase in latency would constitute a significant change.
M-waves typically show substantial run-to-run variability
(Fig. 38.20, left), and uniform use of a 50% amplitude criterion
would lead to an unacceptably high level of false alarms. In many
centers, the alarm criterion is complete disappearance of the
MEP, although a higher percentage amplitude criterion (e.g., a
75% or a 90% drop in amplitude) may also be used. When eval-
uating M-waves, the anesthetic regimen should be taken into
account. Increases in the anesthetic doses, especially the addition
of an inhalational agent to an intravenous regimen, may cause
substantial amplitude attenuation or eliminate the MEPs (Fig.
38.5). In addition, during surgeries lasting several hours MEPs
amplitudes tend to decrease, and thresholds to elicit MEPs tend
to rise, a phenomenon known as “anesthetic fade” (82,83). This
may require reassessment of the MEP baseline.
Since TES preferentially stimulates the brain under the anode,
it predominantly elicits M-waves in muscles contralateral to the
anode (see Chapter 52). M-waves may also be present in muscles
contralateral to the TES cathode, but these cathodal stimulation
elicited MEPs (CSE-MEPs) may disappear due to anesthetic-
related changes in cortical excitability, in the absence of corti-
cospinal tract compromise. Therefore, adverse changes in
CSE-MEPs should not be interpreted as definite evidence of spinal
cord compromise. Instead, MEPs elicited by anodal stimulation
should be used to more accurately assess the motor pathways to
the limb(s) in which the CSE-MEPs had deteriorated (84).
Causes of Adverse EP Changes
The causes of adverse MEP changes are similar to those
described for SEPs, including direct mechanical or thermal tis-
sue injury and ischemia/infarction of neural or muscle tissue
due to compromise of its blood supply. Apparent significant EP
attenuations due to changes in the anesthetic regimen are even
more common for M- wave monitoring than for SEP monitor-
ing. If the spinal cord below the neck is at risk (as in scoliosis or
aortic surgery), then MEPs recorded from upper limb muscles
can be used as controls to identify anesthetic effects.
Typical Intraoperative Applications
MEPs are used to monitor the condition of the corticospinal
tracts during surgery on the brain, brainstem, or spinal cord
that places those structures at risk, during orthopedic surgery
784 Part V ■ Complementary and Special Techniques
Figure 38.21 Consecutive MEP runs recorded during the same operation as the data shown in Fig. 38.19. Shown here are
the last four MEP runs recorded from the abductor hallucis and tibialis anterior muscles to contralateral anodal TES before
the adverse changes (top) and the first four runs recorded after them (bottom). The MEPs recorded from the hand muscles
did not change.
that poses a risk to the spinal cord (Fig. 38.21), or during sur-
gery on the aorta or on other blood vessels that supply the brain
and spinal cord. They can also be used to assess the function of
cranial nerves or spinal nerve roots by recording the MEPs in
the muscles that they innervate following stimulation of the
corticospinal and corticobulbar tracts within the brain.
As noted above, the somatosensory pathways may be dam-
aged without MEP changes, so SEP monitoring should also be
employed when MEPs are used to monitor the spinal cord. Also,
monitoring both SEPs and MEPs provides a measure of redun-
dancy for monitoring the integrity of the spinal cord; two inde-
pendent monitoring techniques are employed, each capable of
detecting most cases of spinal cord compromise in case the
other technique cannot provide effective IOM due to preexist-
ing neurologic compromise, anesthetic effects, NMB, exces-
sively noisy data, or other technical problems (85).
REFERENCES
1. Epstein CM, Boor DR. Principles of signal analysis and averaging.
Neurol Clin. 1988;6:649–656.
2. American Clinical Neurophysiology Society. Guideline 9C: guide-
lines on short- latency auditory evoked potentials. J Clin
Neurophysiol. 2006;23:157–167.
3. American Clinical Neurophysiology Society. Guideline 9D: guide-
lines on short-latency somatosensory evoked potentials. J Clin
Neurophysiol. 2006;23:168–179.
4. Legatt AD. Intraoperative neurophysiologic monitoring. In: Frost
EAM, ed. Clinical Anesthesia in Neurosurgery. 2nd ed. Boston, MA:
Butterworth-Heinemann; 1991:63–127.
5. Legatt AD. Mechanisms of intraoperative brainstem auditory
evoked potential changes. J Clin Neurophysiol. 2002;19:396–408.
6. Yvert B, Crouzeix A, Bertrand O, et al. Multiple supratemporal
sources of magnetic and electric auditory evoked middle latency
components in humans. Cereb Cortex. 2001;11:411–423.
7. Schneider G, Hollweck R, Ningler M, et al. Detection of conscious-
ness by electroencephalogram and auditory evoked potentials.
Anesthesiology. 2005;103:934–943.
8. Legatt AD. Brainstem auditory evoked potentials: methodology,
interpretation, and clinical application. In: Aminoff MJ, ed.
Electrodiagnosis in Clinical Neurology. 5th ed. New York, NY:
Churchill Livingstone; 2005:489–523.
9. Banoub M, Tetzlaff JE, Schubert A. Pharmacologic and physiologic
influences affecting sensory evoked potentials: implications for
perioperative monitoring. Anesthesiology. 2003;99:716–737.
10. Stockard JJ, Pope-Stockard JE, Sharbrough FW. Brainstem audi-
tory evoked potentials in neurology: methodology, interpretation,
and clinical application. In: Aminoff MJ, ed. Electrodiagnosis in
Clinical Neurology. 3rd ed. New York, NY: Churchill Livingstone;
1992:503–536.
11. Litscher G. Continuous brainstem auditory evoked potential moni-
toring during nocturnal sleep. Int J Neurosci. 1995;82:135–142.
12. Markand ON, Lee BI, Warren C, et al. Effects of hypothermia on
brainstem auditory evoked potentials in humans. Ann Neurol.
1987;22:507–513.
13. Rodriguez RA, Edmonds HL Jr, Auden SM, et al. Auditory brain-
stem evoked responses and temperature monitoring during pedi-
atric cardiopulmonary bypass. Can J Anaesth. 1999;46:832–839.
14. Jewett DL, Williston JS. Auditory- evoked far fields averaged from
the scalp of humans. Brain. 1971;94:681–696.
15. Chiappa K, Gladstone KJ, Young RR. Brain stem auditory evoked
responses. Studies of waveform variations in 50 normal human
subjects. Arch Neurol. 1979;36:81–87.
16. Legatt AD, Arezzo JC, Vaughan HG Jr. The anatomic and physio-
logic bases of brain stem auditory evoked potentials. Neurol Clin.
1988;6:681–704.
 Chapter 38
17. Gersdorff MCH. Simultaneous recordings of human auditory poten-
tials: transtympanic electrocochleography (ECoG) and brainstem-
evoked responses (BER). Arch Otorhinolaryngol. 1982;234:15–20.
18. Hughes JR, Fino JJ. A review of generators of the brainstem audi-
tory evoked potential: contribution of an experimental study.
J Clin Neurophysiol. 1985;2:355–381.
19. Grandori F. Field analysis of auditory evoked brainstem potentials.
Hear Res. 1986;21:51–58.
20. Raudzens PA, Shetter AG. Intraoperative monitoring of brain-
stem auditory evoked potentials. J Neurosurg. 1982;57:341–348.
21. Legatt AD, Pedley TA, Emerson RG, et al. Electrophysiological moni-
toring of seventh and eighth nerve function during surgery for
acoustic neuromas. Electroencephalogr Clin Neurophysiol. 1986;64:30P.
22. Strominger NL. The origins, course and distribution of the dorsal
and intermediate acoustic striae in the rhesus monkey. J Comp
Neurol. 1973;147:209–234.
23. Strominger NL, Nelson LR, Dougherty WJ. Second- order auditory
pathways in the chimpanzee. J Comp Neurol. 1977;172:349–366.
24. Strominger NL, Strominger AI. Ascending brain stem projections
of the anteroventral cochlear nucleus in the rhesus monkey. J
Comp Neurol. 1971;143:217–242.
25. Brown RH Jr, Chiappa KH, Brooks E. Brain stem auditory evoked
responses in 22 patients with intrinsic brain stem lesions: implica-
tions for clinical interpretations. Electroencephalogr Clin Neuro-
physiol. 1981;52:38P.
26. Faught E, Oh SJ. Brainstem auditory evoked responses in brain-
stem infarction. Stroke. 1985;16:701–705.
27. Oh SJ, Kuba T, Soyer A, et al. Lateralization of brainstem lesions by
brainstem auditory evoked potentials. Neurology. 1981;31:14–18.
28. Stockard JJ, Rossiter VS. Clinical and pathologic correlates of brain
stem auditory response abnormalities. Neurology. 1977;27:316–325.
29. Starr A, Hamilton AE. Correlation between confirmed sites of neu-
rological lesions and abnormalities of far- field auditory brainstem
responses. Electroencephalogr Clin Neurophysiol. 1976;41:595–608.
30. Stockard JJ, Stockard JE, Sharbrough FW. Detection and localiza-
tion of occult lesions with brainstem auditory responses. Mayo
Clin Proc. 1977;52:761–769.
31. Hirsch BE, Durrant JD, Yetiser S, et al. Localizing retrocochlear
hearing loss. Am J Otol. 1996;17:537–546.
32. York DH. Correlation between a unilateral midbrain-pontine
lesion and abnormalities of brain-stem auditory evoked potential.
Electroencephalogr Clin Neurophysiol. 1986;65:282–288.
33. Scaioli V, Savoiardo M, Bussone G, et al. Brain-stem auditory
evoked potentials (BAEPs) and magnetic resonance imaging (MRI)
in a case of facial myokymia. Electroencephalogr Clin Neurophysiol.
1988;71:153–156.
34. Zanette G, Carteri A, Cusumano S. Reappearance of brain- stem
auditory evoked potentials after surgical treatment of a brain-stem
hemorrhage: contributions to the question of wave generation.
Electroencephalogr Clin Neurophysiol. 1990;77:140–144.
35. Fischer C, Bognar L, Turjman F, et al. Auditory evoked potentials
in a patient with a unilateral lesion of the inferior colliculus and
medial geniculate body. Electroencephalogr Clin Neurophysiol.
1995;96:261–267.
36. Emerson RG, Brooks EB, Parker SW, et al. Effects of click polarity on
brainstem auditory evoked potentials in normal subjects and
patients: unexpected sensitivity of wave V. Ann N Y Acad Sci. 1982;
388:710–721.
37. Schwartz DM, Morris MD, Spydell JD, et al. Influence of click
polarity on the brain- stem auditory evoked response (BAER)
revisited. Electroencephalogr Clin Neurophysiol. 1990;77:445–457.
■ Intraoperative Evoked Potential Monitoring 785
38. Little JR, Lesser RP, Lueders H, et al. Brain stem auditory evoked
potentials in posterior circulation surgery. Neurosurgery. 1983;12:
496–502.
39. Roeser RJ, Clark JL. Clinical masking. In: Roeser RJ, Valente M,
Hosford-Dunn H, eds. Audiology: Diagnosis. New York, NY: Thieme;
2000:253–279.
40. Møller AR, Jannetta PJ. Monitoring auditory functions during cra-
nial nerve microvascular decompression operations by direct
recording from the eighth nerve. J Neurosurg. 1983;59:493–499.
41. Kim HN, Kim YH, Park IY, et al. Variability of the surgical anatomy
of the neurovascular complex of the cerebellopontine angle. Ann
Otol Rhinol Laryngol. 1990;99:288–296.
42. Nadol JB Jr, Levine R, Ojemann RG, et al. Preservation of hearing in
surgical removal of acoustic neuromas of the internal auditory canal
and cerebellar pontine angle. Laryngoscope. 1987;97:1287–1294.
43. Sekiya T, Møller AR. Avulsion rupture of the internal auditory
artery during operations in the cerebellopontine angle: a study in
monkeys. Neurosurgery. 1987;21:631–637.
44. Levine RA, Ronner SF, Ojemann RG. Auditory evoked potential
and other neurophysiologic monitoring techniques during tumor
surgery in the cerebellopontine angle. In: Loftus CM, Traynelis VC,
eds. Intraoperative Monitoring Techniques in Neurosurgery. New
York, NY: McGraw-Hill; 1994:175–191.
45. Radtke RA, Erwin CW, Wilkins RH. Intraoperative brainstem
auditory evoked potentials: significant decrease in postoperative
morbidity. Neurology. 1989;39:187–191.
46. Kimura J, ed. Electrodiagnosis in Diseases of Nerve and Muscle:
Principles and Practice. 3rd ed. Oxford: University Press; 2001.
47. Pelosi L, Cracco JB, Cracco RQ, et al. Comparison of scalp distri-
bution of short latency somatosensory evoked potentials (SSEPs)
to stimulation of different nerves in the lower extremity.
Electroencephalogr Clin Neurophysiol. 1988;71:422–428.
48. Fried S, Legatt AD. The utility of a forehead- to-inion derivation in
recording the subcortical far- field potential (P14) during median
nerve somatosensory evoked potential (SEP) testing. J Clin
Neurophysiol. In press.
49. American Clinical Neurophysiology Society. Guideline five: guide-
lines for standard electrode position nomenclature. J Clin
Neurophysiol. 1996;23:107–110.
50. Desmedt JE, Cheron G. Non-cephalic reference recording of early
somatosensory potentials to finger stimulation in adult or aging
normal man: differentiation of widespread N18 and contralateral
N20 from prerolandic P22 and N30 components. Electroencephalogr
Clin Neurophysiol. 1981;52:553–570.
51. Emerson RG. Anatomic and physiologic bases of posterior tibial
nerve somatosensory evoked potentials. Neurol Clin. 1988;6:
735–749.
52. Allison T, Goff WR, Williamson PD, et al. On the neural origin of
early components of the human somatosensory evoked potential. In:
Desmedt JE, ed. Clinical Uses of Cerebral, Brainstem and Spinal
Somatosensory Evoked Potentials. Progress in Clinical Neurophysiology.
Vol 7. Basel: S Karger; 1980:51–68.
53. Broughton R, Rasmussen T, Branch C. Scalp and direct cortical
recordings of somatosensory evoked potentials in man (circa 1967).
Can J Psychol. 1981;35:136–158.
54. Lueders H, Lesser RP, Hahn J, et al. Cortical somatosensory evoked
potentials in response to hand stimulation. J Neurosurg. 1983;58:
885–894.
55. Gregorie EM, Goldring S. Localization of function in the excision
of lesions from the sensorimotor region. J Neurosurg. 1984;61:
1047–1054.
786 Part V ■ Complementary and Special Techniques
56. King RB, Schell G. Cortical localization and monitoring during
cerebral operations. J Neurosurg. 1987;67:210–219.
57. Wood CC, Spencer DD, Allison T, et al. Localization of human sen-
sorimotor cortex during surgery by cortical surface recording of
somatosensory evoked potentials. J Neurosurg. 1988;68:99–111.
58. Legatt AD, Kader A. Topography of the initial cortical component
of the median nerve somatosensory evoked potential: relationship
to central sulcus anatomy. J Clin Neurophysiol. 2000;17:321–325.
59. Yamada T. The anatomic and physiologic bases of median nerve
somatosensory evoked potentials. Neurol Clin. 1988;6:705–733.
60. O’Brien MF, Lenke LG, Bridwell KH, et al. Evoked potential mon-
itoring of the upper extremities during thoracic and lumbar spinal
deformity surgery: a prospective study. J Spinal Disord. 1994;7:
277–284.
61. Chung I, Glow JA, Dimopoulos V, et al. Upper- limb somatosen-
sory evoked potential monitoring in lumbosacral spine surgery: a
prognostic marker for position- related ulnar nerve injury. Spine J.
2009;9:287–295.
62. Cusick JF, Myklebust JB, Larson SJ, et al. Spinal cord evaluation by
cortical evoked responses. Arch Neurol. 1979;36:140–143.
63. Machida M, Weinstein SL, Yamada T, et al. Dissociation of muscle
action potentials and spinal somatosensory evoked potentials after
ischemic damage of spinal cord. Spine. 1988;13:1119–1124.
64. Ben-David B, Haller G, Taylor P. Anterior spinal fusion complicated
by paraplegia: a case report of a false-negative somatosensory-
evoked potential. Spine. 1987;12:536–539.
65. Zornow MH, Grafe MR, Tybor C, et al. Preservation of evoked
potentials in a case of anterior spinal artery syndrome. Electroen-
cephalogr Clin Neurophysiol. 1990;77:137–139.
66. Jones SJ, Buonamassa S, Crockard HA. Two cases of quadriparesis
following anterior cervical discectomy, with normal perioperative
somatosensory evoked potentials. J Neurol Neurosurg Psychiatry.
2003;74:273–276.
67. Ben- David B, Taylor PD, Haller GS. Posterior spinal fusion com-
plicated by posterior column injury. A case report of a false-
negative wake-up test. Spine. 1987;12:540–543.
68. Chatrian G-E, Berger MS, Wirch AL. Discrepancy between intraop-
erative SSEP’s and postoperative function. J Neurosurg. 1988;69:
450–454.
69. Legatt AD. Current practice of motor evoked potential monitor-
ing: results of a survey. J Clin Neurophysiol. 2002;19:454–460.
70. Owen JH, Laschinger J, Bridwell K, et al. Sensitivity and specificity
of somatosensory and neurogenic- motor evoked potentials in ani-
mals and humans. Spine. 1988;13:1111–1118.
71. Toleikis JR, Skelly JP, Carlvin AO, et al. Spinally elicited peripheral
nerve responses are sensory rather than motor. Clin Neurophysiol.
2000;111:736–742.
72. Minahan RE, Sepkuty JP, Lesser RP, et al. Anterior spinal cord
injury with preserved neurogenic “motor” evoked potentials. Clin
Neurophysiol. 2001;112:1442–1450.
73. Machida M, Weinstein SL, Yamada T, et al. Monitoring of motor
action potentials after stimulation of the spinal cord. J Bone Joint
Surg. 1988;70:911–918.
74. Mochida K, Shinomiya K, Komori H, et al. A new method of mul-
tisegment motor pathway monitoring using muscle potentials after
train spinal stimulation. Spine. 1995;20:2240–2246.
75. Su CF, Haghighi SS, Oro JJ, et al. “Backfiring” in spinal cord moni-
toring. High thoracic spinal cord stimulation evokes sciatic response
by antidromic sensory pathway conduction, not motor tract conduc-
tion. Spine. 1992;17:504–508.
76. Legatt AD, Ellen R. Grass lecture: motor evoked potential monitor-
ing. Am J Electroneurodiagnostic Technol. 2004;44:223–243.
77. Sloan T, Heyer EJ. Anesthesia for intraoperative neurophysiologic
monitoring of the spinal cord. J Clin Neurophysiol. 2002;19:430–443.
78. Burke D, Hicks RG. Surgical monitoring of motor pathways. J Clin
Neurophysiol. 1998;15:194–205.
79. Deletis V, Kothbauer K. Intraoperative neurophysiology of the cor-
ticospinal tract. In: Stålberg E, Sharma HS, Olsson Y, eds. Spinal
Cord Monitoring. Wein: Springer; 1998:421–444.
80. Amassian VE, Stewart M, Quirk GJ, et al. Physiological basis of motor
effects of a transient stimulus to cerebral cortex. Neurosurgery. 1987;
20:74–93.
81. Rothwell J, Burke D, Hicks R, et al. Transcranial electrical stimula-
tion of the motor cortex in man: further evidence for the site of
activation. J Physiol. 1994;481:243–250.
82. Lyon R, Feiner J, Lieberman JA. Progressive suppression of motor
evoked potentials during general anesthesia: the phenomenon of
“anesthetic fade.” J Neurosurg Anesthesiol. 2005;17:13–19.
83. MacDonald DB, Al Zayed Z, Khoudeir I, et al. Monitoring scoliosis
surgery with combined multiple pulse transcranial electric motor
and cortical somatosensory- evoked potentials from the lower and
upper extremities. Spine. 2003;28:194–203.
84. Legatt AD. MEPs elicited by cathodal stimulation during transcra-
nial electrical stimulation- MEP monitoring. Neurology. 2006;
66(suppl 2):A68.
85. Nagle KJ, Emerson RG, Adams DC, et al. Intraoperative monitoring
of motor evoked potentials: a review of 116 cases. Neurology. 1996;47:
999–1004.

Monitoring EEG during Carotid Surgery
INTRODUCTION
The studies from the North American Symptomatic Carotid
Endarterectomy Trial Collaborators (1) and the European
Carotid Surgery Trialists’ Collaboration Group (2) have shown
the benefit of carotid endarterectomy (CEA) to prevent stroke,
especially for patients with high-grade carotid artery stenosis. A
more recent study has reaffirmed the effectiveness of CEA,
especially in symptomatic patients with severe carotid stenosis
( 70%) (3). Patients with asymptomatic stenosis in the 50% to
60% range are in general uncertain candidates for CEA, but
may well benefit, especially for patients with additional medical
conditions having a higher stroke risk. The CEA benefit is sig-
nificantly less for asymptomatic patients (4,5), but CEA may
still be applicable for asymptomatic patients with a high degree
of stenosis (2), plaque with ulceration (6), and an occluded con-
tralateral carotid artery (7).
Because CEA always carries a risk of morbidity and mortal-
ity (allowable rates of stroke and death rate of less than 6%),
CEA for asymptomatic patients should only be considered if
surgery could be done by an expert surgeon with track record
of low complication rates of less than 3%. CEAs are generally
not indicated for patients with less than 50% symptomatic or
less than 60% asymptomatic stenosis (8).
CEA can be performed under general or regional anesthesia
(RA). During CEA, the internal carotid artery must be cross-
clamped; therefore, the brain becomes subject to ischemia or
hypoperfusion. The brain can tolerate more than 1 hour if cere-
bral blood flow is greater than 15 mL/100 g/min, but the severe
infarct would occur within 4 minutes with zero blood flow. A
number of monitoring techniques are available for examining the
cerebral perfusion and detecting the risk of cerebral ischemia.
These include EEG, somatosensory-evoked potential (SSEP)
(median nerve stimulation), transcranial Doppler, and oxygen
saturation of hemoglobin using near-infrared spectroscopy,
stump-pressure measurement, and cerebral blood flow by Xenon-
133. In order to protect brain from ischemic damage during cross-
clamping, placement of a catheter to shunt blood around the
surgical site has been used. Because the shunt placement itself car-
ries a significant risk of embolization as a result of dislodging ath-
erosclerotic emboli coming from the vessel wall, only patients with
a potentially high risk of developing ischemic brain damage are
selected for shunt placement. Determining which patients require
shunts depend on the multiple factors. EEG has been the most
commonly used monitoring method for shunt determination but
other tests are complimentary or may be more sensitive and accu-
rate for predicting ischemic insult in some cases.
CHAPTER
THORU YAMADA AND MALCOLM YEH 39
Whether to use a shunt has long been argued by many inves-
tigators. Some have proposed the use of shunt in all CEAs (9)
while others have stressed the increase of stroke incidence with
the use of shunt and have argued that the shunt should not be
used at all (10,11). Most CEAs, however, have been done with
selective shunting based on EEG or other monitoring methods.
Irrespective of the shunt used routinely, selectively or not at all,
the incidences of perioperative stroke were overall similar, 3%
(1). In three groups of studies, however, the incidence of stroke
was highest among those using a shunt routinely and lowest
among those never using a shunt. These data suggest that intra-
operative or postoperative stroke is more likely due to emboli
rather than hypoperfusion or ischemia and thus raises a ques-
tion of the value to shunting and monitoring (12). Indeed in a
study using RA (patients were awake during the CEA), only one
of 21 neurologic deficits (a total of 345 patients) was attributable
to cerebral anoxia and remaining 20 patients were due to
thrombo-embolic origin (13). Another large scale of review
study (3062 CEAs) showed that intraoperative ischemia, throm-
bosis, embolization, and intracranial hemorrhage accounted for
47 of the 63 deficits (14). Of these, thrombo-embolic events were
the most common cause of neurologic deficit accounting for
45%. Nonetheless a substantial number of patients would suffer
from ischemic stroke unless appropriate protection is provided.
Moreover, in some cases, postoperative stroke could be extensive
resulting in global cerebral dysfunction. This is where various
monitoring methods can be used to warn of an impending
ischemic event. One study comparing selective shunting based
on EEG (311 patients) and a nonshunt group (591 patients) in a
single institute reported much higher incidence of postoperative
stroke in the nonshunt group (13 patients) as compared to a
selective shunt group (1 patient) (15).
ANESTHESIA
Commonly used anesthesia for CEA surgery is gaseous anesthe-
sia with isoflurane ranging between 0.4% and 0.6% with 40%
to 60% N2O and oxygen. This combination may be supple-
mented by sufentanil or remifentanil (16,17). An alternative
anesthetic regimen consists of nitrous oxide in oxygen and
propofol after induction with etomidate and fentanyl (17).
Other commonly used inhalation agents include desflurane,
enflurane, and sevoflurane. Overall different anesthetic agents
show a similar EEG pattern when used at a concentration below
their minimal alveolar concentrations (MAC) level (the level
necessary to prevent movement to painful stimuli in about 50%
of subjects).
787
788 Part V ■ Complementary and Special Techniques
EEG Change during Induction
Most anesthesia inductions are carried out by etomidate, propo-
fol, or thiopental via intravenous administration. This initially
produces an EEG showing beta activity with a decrease in the
alpha rhythm. The beta activity rapidly becomes more wide-
spread, with increases in amplitude. This is followed by a slowing
in the beta frequency that approaches the alpha frequency activ-
ity. During this phase, the beta or alpha activity may be inter-
mixed with a burst of high amplitude, intermittent rhythmic
delta activity, which is often of frontal dominance, that is, FIRDA
(frontal intermittent delta activity) before a steady-state anes-
thetic pattern is established (Figs. 39.1A, B and 39.2A,B) (18).
EEG Patterns at Sub-MAC Concentration
At this light level of steady-state anesthesia, a widespread ante-
riorly dominant rhythm (WAR) pattern, usually in lower beta
or alpha frequency activity, is seen with all of the above
described agents (18). The frequency of this pattern tends to
slow with increasing concentration of these agents. This anes-
thetic WAR pattern is strikingly similar to the alpha-coma pat-
tern often seen in a postanoxic cerebral insult (Figs. 39.1D,
39.2D, 39.3A, and 39.4A). However, the alpha frequency WAR
pattern may be simply due to slowing of the beta frequency
activity induced by anesthesia and may approach alpha or even
theta frequency activity.
In addition to the WAR pattern, anterior maximum inter-
mittent slow (AIS) waves may appear (Figs. 39.1C and 39.2C)
(18). AIS waves are commonly diphasic having a sharply con-
toured negative–positive configuration with a duration of less
than a second occurring singly or in brief trains.
Another pattern during this level of anesthesia is widespread
persistent slow (WPS) waves (Figs. 39.1C and 39.2C) (18). The
WPS waves consist of polymorphic slow waves with a duration
usually exceeding 1 second. This pattern is least obvious with
halothane and enflurane and is most prominent with isoflurane
(18). With deepening of the anesthetic level, there is an increase
in the amplitude and wave length of the WPS pattern and a
decrease in WAR pattern.
EEG Pattern w ith Supra-MAC Concentration
The EEG patterns of different anesthetic agents are similar when
the concentrations are at sub-MAC level. But the EEG patterns
are different when the concentration levels are above MAC. For
instance, enflurane elicits spike and wave at 1.5 MAC and a burst
suppression pattern that may include a spike component at
MAC levels greater than 2. At these levels even seizure activity
may be elicited when activated by hyperventilation (19).
In contrast, isoflurane tends to exert an antiepileptic effect
producing an intermittent suppression pattern at 1.5 MAC and
continuous suppression at MAC above 2. Halothane also has an
antiepileptic property but may not produce EEG inactivity even
with concentrations as high as 4 MAC (18). Desflurane and
isoflurane tend to rapidly produce EEG patterns of suppression
or burst suppression at supra-MAC levels. Adding nitrous oxide
to isoflurane or desflurane may potentiate the EEG suppression
at a lower concentration of the primary anesthetic agent (18).
EEG Pattern During Recovery from Anesthesia
When discontinuing anesthetics at the sub-MAC level, the WAR
pattern decreases in its amplitude and increases in frequency,
which subsequently changes to the beta range activity. There is
also a decrease of the WPS pattern before the patient regains
consciousness (18). During the transition from the anesthetic
state to the arousal state, bursts of intermittent delta activity,
resembling a FIRDA pattern, may appear which are similar to
but less prominent than those seen during induction.
Other Factors That Affect the EEG
The decreasing PaCO 2 level below 40 mm Hg or decreasing
the amount of stimulation that would be painful if given dur-
ing awake state would produce a pattern suggestive of a deeper
level of anesthesia, that is, increase of slow waves, especially
during light sub-MAC level of anesthesia. Severe blood pres-
sure drops are associated with increases of slow waves and
eventual flattening of the EEG activity. Hypothermia slows the
EEG frequency spectrum with gradual reduction of amplitude
and eventually leads to EEG “flattening” at about 20 C to
25 C. However, electrocerebral inactivity created by severe
hypothermia could act as cerebral protective mechanism.
Hypothermia also accentuates the EEG effect of anesthetic
agents (18).
EEG MONITORING
EEG has been used for monitoring the brain function during
CEA for many years and probably has been the most commonly
used method for the detection of cerebral ischemia that would
determine the need for selective shunting. EEG can be recorded
easily in the operating room despite an electrically hostile envi-
ronment where multiple electrical instruments are connected
to or working nearby the patients and surgical personnel. Most
EEG monitoring are performed under general anesthesia and
its activity may vary depending on the type and depth of anes-
thesia, blood pressure, body temperature, or other physiological
factors such as PaCO2 and PaO2 level. These variables should be
thoroughly understood by the monitoring team.
Recording Technique
Because EEG changes during the CEA can be diffuse or focal,
the electrodes across the entire scalp, using 16 EEG data chan-
nels, are commonly used. It is preferable to use collodion, which
ensures the secure attachment of electrodes to the scalp during
the surgery.
The anterior–posterior longitudinal bipolar (so-called
banana) montage is commonly used because this montage
allows easy detection of hemispheric asymmetry.
The filter setting is usually the same as of routine EEG with
the high (low pass) filter of 70 Hz. Some laboratories may
choose the low (high pass) filter setting of 0.3 Hz and high (low
pass) filter setting of 35 Hz. Because of the electrically “noisy”
operating room environment, 60 Hz contamination onto the
EEG recording may be inevitable. The monitoring team should
try to find the source of 60-Hz artifacts and to eliminate the
Figure 39.1 EEG examples of the anesthetic effect using isoflurane and nitrous oxide. The patient is a 50-year-old woman
having a history of TIA symptoms with intermittent partial visual loss in the left eye with right extremity weakness.
A: Within 5 seconds after IV propofol injection, the EEG starts to show diffuse delta with superimposed beta activity. (This
patient had preexisting beta activity prior to the induction.) B: Subsequent polymorphic delta activity with superimposed
beta activity is seen.
(continued on next page)
 Figure 39.1 (continued) This was followed by C that shows a decrease of delta activity along with the widespread ante-
riorly dominant rhythm (WAR) consisting of dominant alpha and beta activities. In addition there were occasional anterior
dominant intermittent slow (AIS) and widespread (irregular) persistent slow (WPS) patterns. D: When the anesthesia was
stabilized at sub-MAC level, AIS and WPS patterns became less frequent and less prominent, and WAR pattern dominated
throughout the procedure.
790
Figure 39.2 Another EEG example of anesthetic effect using desflurane and nitrous oxide. The EEG monitors a right CEA
in a 66-year-old man without a history of TIA or stroke but having a 70% to 80% stenosis on right and 50% to 60% steno-
sis on left carotid artery. A and B: Within 5 seconds after the IV propofol injection, the EEG started to show diffuse theta
activity mixed with beta activity, which was followed by an increase of irregular delta as well as frontal dominant rhythmic
delta activity (FIRDA) within 30 to 40 seconds after the induction. (continued on next page)
 Figure 39.2 (continued) C: About a minute after the induction, delta slow waves became more dominant with irregular
delta (WPS) along with a frontal dominant intermittent delta (AIS) pattern. D: Once the anesthesia was stabilized at the
sub-MAC level, the EEG was dominated by diffuse alpha and slow beta activity with minimal theta–delta components,
resembling “alpha-coma.”
792
Figure 39.3 A 69-year-old man without a history of TIA or stroke has a 90% stenosis of left carotid and a patent right
carotid artery. A: Within several seconds after cross-clamping the left carotid artery, there was a slight reduction and slow-
ing of the alpha activity, noted more prominently in the left hemisphere. Note the dominant WAR pattern with alpha activ-
ity resembling an “alpha-coma” prior to the clamp. B: The suppression of EEG activity in the left hemisphere became more
prominent within 20 seconds after carotid clamp. This persists until the shunt was placed. (continued on next page)
794 Part V ■ Complementary and Special Techniques

Figure 39.3 (continued) Within a few seconds after the shunt was placed, the EEG started to show recovery (C,D). It
took more than 3 minutes to recover fully to the preclamp EEG (E).

Figure 39.3 (continued)
artifacts by technical modifications if possible. If 60-Hz
artifacts cannot be eliminated, the use of 60-Hz notch filter
is allowed.
The sensitivity setting is usually 7 or 5 V/mm. It is not
uncommon for the EEG amplitude to be exceedingly low under
anesthesia, requiring a sensitivity of 3 V/mm.
The sweep speed is usually a 10-second display on a full
video screen but may be adjusted to 5 seconds or 20 to 30 sec-
onds screen depending on the interpretation needs and prefer-
ence of electroencephalographer.
A preoperative, awake resting EEG can be recorded in the
operating room before anesthesia induction. This allows
inspection of any preexisting EEG abnormalities before anes-
thesia. At least 5 to 10 minutes of baseline preclamp EEG is
essential while the patient is under stable anesthesia to appreci-
ate any clamp-associated EEG changes. After the clamp is
applied, the EEG of the preclamp state may be displayed on half
of the video screen, so that the postclamp on-going EEG can be
directly compared with preclamp EEG. Although most clamp-
related EEG changes occur within 20 to 30 seconds after the
clamp, the EEG should be monitored throughout the procedure
because EEG changes may occur in the midst of procedure, long
after the clamp has been applied. Possible causes for the late
EEG changes include hypoperfusion secondary to a lowered
blood pressure or embolism, or shunt malformation if a shunt
is used.
Chapter 39 ■ Monitoring EEG during Carotid Surgery 795
Preclamp Focal EEG Abnormalities in
Relationship to the Anesthesia
The majority of EEG patterns during CEA show a symmetric
baseline pattern before the clamping of the carotid artery.
However, depending on case selection, 30% to 40% of preclamp
EEG may show focal abnormality of varying degree (20). The
focal abnormality may consist of unilateral reduction of the
WAR amplitude. This may be accompanied by polymorphic
delta activity on the side of the reduced amplitude. Most focal
EEG abnormalities under anesthetic usually correlate with a
preoperative waking EEG abnormalities. However, anesthesia
may activate an abnormality that was either inapparent or less
apparent during the waking state (18). Conversely, the anesthe-
sia may obscure a focal abnormality in some cases that was
apparent in EEG recordings during the awake state.
In patients with a lesion in the anterior head region where
the alpha rhythm tends to be unaffected in the awake state, the
anesthesia may bring out a reduction of the WAR pattern
along with decreased beta and increased polymorphic slowing
in the anterior head region. Preexisting intermittent rhythmic
delta in the temporal region in the awake state may become
more prominent and persistent, along with a reduction of the
WAR pattern (18). A more posterior lesion having decreased
alpha activity on the side of lesion may leave the anesthetic
WAR pattern symmetrical without focal slowing (18).
Preexisting nonlateralizing FIRDA or persistent generalized
 Figure 39.4 A 77-year-old man without a history of TIA or stroke has occlusion of right carotid and an 80% stenosis of
left carotid artery. A: The EEG started to show an increase of theta and delta range activity bilaterally, which is more promi-
nently noted on the left hemisphere within 5 seconds after the clamp. B: Within 20 seconds after the clamp, the EEG
showed diffuse suppression and loss of fast activity bilaterally which was to a greater degree over the left hemisphere (B).
796
Figure 39.4 (continued) C: Within a few seconds after the shunt was placed the EEG started to recover. D: Full recovery
was achieved at about 1 minute after the shunt was placed.

797
798 Part V ■ Complementary and Special Techniques
slowing during awake state may be observed and cannot be
distinguished from the anesthesia-induced slowing that com-
monly occurs during induction or during the recovery phase
of anesthesia (18).
EEG Changes after Cross-Clamping
the Carotid Artery
Most preclamp EEG patterns have a symmetrical pattern with-
out focal features and the development of an asymmetric pat-
tern after the clamp can be easily recognized. The
clamp-associated EEG change occurs within the first minute in
the majority of patients ( 80%) with most (69%) appearing
within 20 seconds after the clamp (21). The most common and
sensitive EEG change indicating cerebral ischemia is the decrease
of alpha and beta frequency activity (WAR pattern) on the
clamped side (Figs. 39.3A,B and 39.4A,B). Fortunately the most
common EEG pattern under stable sub-MAC level of anesthesia
consists of abundant alpha and beta activity (WAR pattern).
Further progress of ischemia is associated with an increase of
delta amplitude or slower frequencies . The most severe degree
of EEG change is “flattening” of the EEG pattern with depres-
sion of all activity including delta and faster frequency compo-
nents (Figs. 39.3C and 39.4B). These changes occur primarily
on the hemisphere ipsilateral to the side of clamp but a bilateral
change could occur if the blood flow of contralateral hemi-
sphere depends on the collateral circulation from the ipsilateral
hemisphere. This may occur when occlusion or severe stenosis
of the contralateral carotid artery exists. When bilateral changes
occur, it may be difficult to differentiate if changes are due to
global ischemia or systemic factors such as a change in anes-
thetic level, blood pressure, temperature, or PaCO2/O2 level.
However, the bilateral change secondary to hypoperfusion usu-
ally shows a greater EEG change where more “flattening” of
EEG activity occurs in the presence of slower delta and/or
greater depression of fast activity over the ipsilateral hemi-
sphere (Fig. 39.4B,C).
After the shunt is placed, the EEG recovers gradually to the
baseline state (preclamp EEG) and may take several minutes to
full recovery. It appears that the longer the clamp time before
shunt placement occurs, the slower the recovery time. In a small
number of patients, the EEG improves but never recovers to the
preclamp state. Most of these patients wake up with no neuro-
logic deficit. If any, the deficit is minor or transient. In the
majority of shunted or nonshunted cases, the EEG remains
symmetric throughout the procedure. In a small percentage of
patients, transient and focal EEG changes may appear during
the middle of the procedure, long after the cross-clamping has
occurred. These transient and focal EEG changes have been
thought due to the asymmetrical effects of changing level of
anesthesia or resurgence of preexisting focal abnormality. These
transient changes have no neurologic consequence (20). This
finding may be due to decreased blood pressure. Because the
threshold of blood pressure to maintain the appropriate cere-
bral perfusion is not known and varies individually, slight
reductions of systemic blood pressure may result in focal areas
of ischemia bringing out EEG changes especially when the
blood pressure at the time of clamp is close to the threshold
level. Raising the blood pressure usually corrects the adverse
EEG change. Another possibility of late and persistent EEG
change may be due to embolization especially during dissection
or at the time of clamp release which will likely result in neuro-
logic deficit.
A late EEG change may also occur in shunted patient. If this
is not due to blood pressure, anesthesia, or embolization, the
shunt malfunction should be considered
Predictability of EEG Change
The overall incidence of clamp-related EEG changes has been
estimated to be about 10% to 30% (20–23). Most studies have
agreed that the incidence of clamp-related EEG changes was
higher in patients when contralateral carotid artery was
occluded. When comparing patients with a patent and occluded
contralateral carotid artery in a total of 564 patients from a sin-
gle institution, a significantly greater incidence of ipsilateral
EEG change was found in patients with an occluded artery
(39%, 22 patients out of 57 patients) than in patients with
patent artery (16%, 80 patients out of 479 patients) showing
statistically significant difference (P 0.001) (24). The greater
incidence of clamp-associated EEG changes was found with
90% than 50% stenosis of contralateral carotid artery (25).
Other studies also agreed that 30% to 40% of the patients with
contralateral occlusion required shunt based on EEG findings
(26,27).
Another study also reported a 48% incidence of clamp-
related EEG change with contralateral carotid occlusion com-
pared to 18% for stenosis and 21% for normal carotid artery
(28). Clamp-related EEG changes are also more common in
patients when contralateral abnormalities on the preoperative
awake EEG exist than in patients with only ipsilateral or diffuse
abnormalities (21) or patients with previous history of stroke
(29). It should be noted that there are still substantial numbers
of patients who have a complete occlusion of the contralateral
carotid artery that do not show EEG changes when clamping of
the ipsilateral carotid artery occurs. This is likely due to ade-
quate collateral circulation. In such cases, the surgeon may pre-
fer to shunt even if there is no EEG change.
It is intuitive to expect that the greater degree of ipsilateral
stenosis is associated with a lower chance of EEG change,
because the collateral flow from the contralateral hemisphere
would be well established in patients with severe stenosis than
in patients with a patent ipsilateral carotid artery. However, one
study did not support this expectation (15).
CEA RESULTS WITH AN EEG MONITORING
FOR THE SHUNT NEED
There are three types of surgical preferences for CEA: one is a
surgical protocol using shunts routinely (9,30), another proto-
col will not use a shunt at all (21,31), still another protocol
places a shunt selectively by one or more monitoring techniques
(20,32). The reported incidences of perioperative stroke were
not much different among these three protocols but the overall
incidence tends to be highest among those using a shunt rou-
tinely and the lowest among those who do not use a shunt at all.

This suggests that perioperative stroke are more likely due to
embolism rather than cerebral ischemia, and thus the necessity
of any type of monitoring can be questioned. Only a large-scale,
prospective study in a single institution can make accurate
comparisons of these three methods of surgery. Nonetheless,
many studies agree that CEAs without shunt appear to be asso-
ciated with a substantially higher risk of stroke than CEAs with-
out selective shunting (15,23,33). In a retrospective review of
902 consecutive CEAs from a single institute, 591 CEAs were
performed without shunting and also without EEG monitor-
ing. This was compared to 311 CEAs that were done with selec-
tive shunting based on the EEG findings (15). This study
reported that clamp-related EEG changes occurred in 40
patients (12.8%) and a postoperative stroke occurring in only
one patient (0.32%) in the selective shunted group, whereas in
the nonshunted group (no EEG monitoring), 13 patients
(2.19%) had postoperative stroke (P 0.05).
In another large-scale study involving 1661 CEAs, EEG
changes were noted in 366 procedures (22%) (23). A perioper-
ative stroke occurred in five patients (0.03%). Of the five
patients, four had positive EEG finding and only one had false-
negative EEG finding. The authors concluded that EEG moni-
toring accurately identified the patients who did not need
shunting. One study reports less satisfactory results of EEG
monitoring, in which 70 patients (15%) of 458 CEAs showed
EEG changes and received shunts (29). In this study, 15 patients
were found to have immediate stroke or transient neurologic
deficits. Ten of these 15 patients showed no EEG changes after
the clamp. In the remaining five patients, EEG changes
occurred but resolved after the shunt placement or raising
blood pressure. All 15 EEG studies were re-reviewed and veri-
fied by an experienced electroencephalographer.
One of the reasons for a false-negative EEG finding is due
to inability to detect subcortical/capsular or small cortical
ischemia by visual inspection of the EEG. In this sense,
comonitoring SSEP may be useful because SSEP can detect
capsular ischemia. Another reason may be due to embolic
stroke, which may occur in the recovery room after EEG mon-
itoring has ended. The EEG monitoring also showed a consid-
erably high incidence of false-positive findings. Of 176 CEA
patients with EEG monitoring but without shunt, postopera-
tive stroke occurred in only 2 (9%) of the 22 patients with
clamp-related major EEG changes and none of the 33 patients
with moderate EEG changes (21). Another similar study (with
EEG monitoring but without shunt use) also reported a low
incidence of perioperative stroke associated with positive EEG
changes; of 72 CEAs with moderate or major EEG changes
from a total of 293 CEAs, only 6 (8%) of them resulted in
postoperative stroke (34).
THE INCIDENCE OF STROKE AFTER CEA
An intraoperative stroke rate was reported to be 2.1% in the
North American Symptomatic Carotid Endarterectomy Trial
(1) study, in which no uniform method of intraoperative cere-
bral monitoring or protection was used. An overall incidence
of perioperative stroke rate was reported to be 2% to 6% (8).
Chapter 39 ■ Monitoring EEG during Carotid Surgery 799
With EEG monitoring and selective shunting, the incidence of
stroke appeared to be lower than those without monitoring.
One study reported that the incidence of stroke was 2.19% in
nonmonitored/nonshunted group versus 0.32% in selective
shunting based on EEG (P 0.05) (15). Other studies using
EEG monitoring reported 0.4% (35) and 0.03% (23) stroke
rate. Another recent study also reported that selective shunting
based on EEG and SSEP findings resulted in lower stroke rate
of 1% (2 out of 194 patients) in selective shunted group com-
pared with 4% (47 out of 1217 patients) of routinely shunted
group (36).
The incidences of perioperative stroke with and without the
presence of contralateral carotid occlusion or stenosis have
been in dispute. In a large-scale study involving 1661 CEAs, the
statistically significant increases of stroke rate were associated
with contralateral carotid occlusion (1.8%) while the overall
incidence of stroke rate was 0.03% (23). In this study, the
stroke rate was also higher in patients who had EEG change
(1.1%) and greatest in patients who had both EEG as well as
contralateral occlusion (3.3%). Another study also reported
the increase of stroke rate from 1.8% to 7.8% with contralat-
eral occlusion and 1.9% to 4.0% with history of stroke (37). A
third study also reported a greater incidence of perioperative
stroke in patients with than without history of stroke (3% vs.
0.3%) (27). However, there were studies reporting no differ-
ence in stroke rate irrespective of presence or absence of
carotid occlusion (24,38).
LONG-TERM EEG CHANGE AFTER CEA
Thus far little attention has been paid to the long-term outcome
of EEG after the successful CEA. Besides preventing stroke, CEA
has been reported to improve cerebral circulation (39–41) and
cognitive brain function (42). In 28 patients with symptomatic
high-grade carotid stenosis, quantitative EEG (q-EEG) showed
an increased mean frequency 1 Hz (P 0.01) postopera-
tively. The Mini mental test and Set Test for verbal frequency
had a positive correlation with the q-EEG changes (43).
Another study also showed an increased mean frequency of the
alpha band and peak frequency of the alpha after CEA in 166
patients with 70% stenosis of the internal carotid artery (44).
In general, patients with occluded contralateral carotid artery
improved more than other patients.
REGIONAL ANESTHESIA FOR AWAKE CEA
AND EEG CORRELATES
The advantage of regional anesthesia (RA) instead of general
anesthesia is that the patient’s neurologic status can be assessed
during the cross-clamping period. RA can be performed in
most patients, if so elected, unless a significant contraindication
exist such as patient’s refusal, cognitive limitation, anxiety, or
language barrier. RA is accomplished by superficial cervical
block by local anesthetics. Occasionally small doses of intra-
venous narcotic or midazolam may be used, but the patient
must remain awake and lucid during the procedure.
800 Part V ■ Complementary and Special Techniques
A relatively simple neurologic examination involving speech,
sensory and motor and general cognitive functions can be
tested when the patient is awake during the procedure. Perhaps
this is the most reliable method to determine the necessity of
shunt placement.
Based on EEG findings, the rate of shunt requirement under
general anesthesia ranges from about 10% to 30% (15,45–47).
This rate is substantially higher than that of RA with the
neurologic examination in an awake patient, which is about 5%
to 15% (13,46–49). Does this mean that the EEG under general
anesthesia is too sensitive to warn the ischemic change, result-
ing in excessive false-positive finding?
However, studies comparing EEG findings with cases using
neurologic examination in an awake patient during CEA
showed results contrary to this notion. Of 89 patients who
underwent awake CEA with EEG monitoring (45), the con-
cordance rate of both normal EEG and neurologic examina-
tion was 76.4% (68 patients) and the concordance rate of both
abnormal EEG and neurologic examination was 12.4%
(11 patients). The total concordance rate was thus 88.8%. The
remaining 12.2% (10 patients) showed a discrepancy between
the EEG and the neurologic examination. Abnormal EEG with
normal examination, that is, false-positive finding, was seen in
6.7% (6 patients) and normal EEG with abnormal neurologic
examination, that is, false-negative finding, was seen in 4.5%
(4 patients) (50). Another study showed a much higher false-
negative rate of 40.6% where only 19 (59.4%) of 32 patients
showed concordant abnormalities in both EEG and neuro-
logic examination (51). The false-positive findings were found
in three patients (1.0%) in this study. Another study also
reported a high incidence of false-negative EEG changes; of
134 awake CEAs, 13 patients required shunt based on neuro-
logic changes but 4 (30.5%) of these did not show EEG
changes (48). Increased false-negative results for EEG in
awake CEAs may be due to some difficulty in assessing awake
patients, because the EEG recording may have the extra noise
of increased muscle or other artifacts, and at times low-
amplitude alpha rhythm or exceedingly low-amplitude back-
ground activity in awake patients who are anxious. The most
sensitive EEG parameter is beta activity, which is commonly
seen in general anesthesia. This may not be present in the
awake patient (Figs. 39.5A through D). Also some false-
negative EEG monitors may be due to the inability to detect
abnormalities when ischemia or hypoperfusion is limited to
capsular or small cortical/subcortical region. One study raised
a possibility that the awake state may make brain less vulner-
able to hypoperfusion and cerebral ischemia than anesthetic
state (50). This appears to contradict the general assumption
that anesthesia, especially deep anesthesia with burst suppres-
sion, is protective from cerebral ischemic by decreased metab-
olism. However, under general anesthesia, the relationship
between EEG and postoperative neurologic deficit appeared
to be opposite since more false-positive ( 90%) than false-
negative ( 10%) EEG findings are noted (21,34) (see also
“Incidence of Stroke”). These two studies suggested that the
EEG under general anesthesia is overly sensitive for predicting
cerebral ischemia.
ANESTHESIA-INDUCED EEG BURST
SUPPRESSION PATTERN FOR CEAS
The barbiturate-induced burst suppression has been known to
protect against cerebral ischemia by decreasing cerebral metab-
olism during carotid clamping (52–54). One of the earliest stud-
ies using barbiturate for cerebral protection during CEAs
involved seven patients whose EEG changes after cross-clamping
could not be corrected by raising systolic blood pressure (55).
They were placed in a barbiturate-induced burst-suppression
pattern with suppression intervals of 15 to 30 seconds. Despite
significant clamp-related EEG changes including the one with
occluded contralateral carotid artery showing severe depression
of EEG activity, none of these seven patients developed neuro-
logic deficit.
Disadvantages of barbiturate-induced burst suppression
include hemodynamic instability such as intraoperative
hypotension and prolonged anesthesia effect after surgery,
which makes early postoperative neurologic evaluation diffi-
cult. Instead of barbiturate, the more recent studies used
Propofol to induce an EEG burst suppression pattern. Unlike
barbiturate, Propofol does not impair the autoregulation of the
hemodynamic state. Also recovery from anesthesia is quicker
than for barbiturates. Using a propofol-induced burst suppres-
sion pattern, one study reported one major stroke, one postop-
erative death due to myocardial infarction, and two patients
with TIA (transient ischemic attack) out of 100 consecutive
CEA patients (56). In another study from 102 consecutive
CEAs, etomidate-induced burst suppression was used in 20
patients who had EEG changes but without shunt. Of these 20
patients, only one patient developed perioperative stroke (57).
A larger population study is needed to validate the effects of
deep anesthesia to the point of burst suppression pattern on
EEG as a protective measure against brain ischemia.
OTHER MONITORING METHODS FOR CEAS
IN RELATIONSHIP WITH EEG
Cerebral Oxygen Saturation Measurement
Cerebral oximetry using near-infrared spectroscopy measures
SaO2 saturation. Two sensors are placed on the right and left
forehead. The decrease of cerebral oxygen saturation is reported
to be accompanied by a significant decrease in the mean middle
cerebral artery blood velocity, even with minimal change of the
SaO2 (58). A fall of 5% or more in cerebral oxygen saturation
following cross-clamping application was accompanied by a
decrease in mean middle cerebral artery blood velocity of at
least 60% (58).
An earlier study showed fairly good correlation between
change in SaO2 saturation and SSEP (59) and/or EEG changes
(60). The decrease of cerebral oxygen saturation was also
correlated with neurologic symptoms in awake CEA
patients (61). Overall cerebral oxygen saturation less than 60%
and/or a greater than 15% to 20% decrease in SaO2 from the
preclamp baseline appeared to correlate with a prediction of
ischemic compromise using EEG as a gold standard (61,62).

Conversely, cerebral oxygen saturations greater than 61% with
desaturations less than 6.9% to 7.3% of the preclamp baseline
was considered to be in a safe range (62). More recent studies,
however, showed poorer concordance between SaO 2 and
EEG/SSEP findings. Of 323 CEAs under general anesthesia, 24
patients (7.4%) showed significant discrepancies between SaO2
and EEG/SSEP measurement (63); 16 patients showed no sig-
nificant EEG/SSEP change but profound changes occurred in
SaO2, and seven patients showed no change in SaO2 but pro-
found change in EEG/SSEP. The sensitivity of SaO2 compared
with EEG/SSEP was 68% and the specificity was 94%. This gave
a positive-prediction value of 47% and negative-prediction
value of 98% (63). This study concluded that relying on SaO2
alone for selection of shunting is potentially dangerous.
Another recent study also showed positive prediction of 33%
and negative-prediction value of 100% (64). Other studies also
discourage the use of SaO2 as a sole guide for determining shunt
placement (65,66). The overall threshold for the identification
of ischemia has been difficult to define (59).
Carotid Stump Pressure (CSP)
Measurement of stump pressure was one of the earliest tech-
niques used in an attempt to measure the adequacy of collateral
flow after carotid clamping. The residual pressure in the distal
portion of the internal carotid system following temporary
occlusion of the common and external carotid arteries has been
thought to reflect arterial pressure through the circle of Willis and
thus to indicate the efficacy of collateral blood supply to the
hemisphere of the occluded vessel (67). Although earlier reports
doubted its usefulness (20,68), later studies obtained close corre-
lation of CSP with EEG changes to indicate the patency of the
contralateral carotid artery. For example, one study reported EEG
changes in 58% of cases where CSP was 25 mm Hg, 32% of
cases with CSP between 25 and 50 mm Hg, and only 4% of cases
when CSP exceeded 50 mm Hg (69). Based on neurologic exam-
ination in 474 patients with awake CEA, another study showed
false-negative rate requiring shunt placement was only 1.0%
using 40 mm Hg systolic as a threshold for the need of shunting
(70). If these 474 CEAs had been performed using general anes-
thesia the shunt determination rate based on CSP would have
been 29% of patients for CSP 50 mm Hg and 15% of patients
for CSP 40 mm Hg. A more recent and larger scale study (1135
patients) concluded that 45 mm Hg CSP was a reliable predic-
tion of adequate cerebral perfusion (71). In another study of 48
patients undergoing awake CEA, CSP provided similar accuracy
with transcranial Doppler sonography and near-infrared spec-
troscopy for the detection of cerebral ischemia (72). However, in
another study of 100 patients under general anesthesia, 34
patients had CSP 50 mm Hg and only 4 of these needed shunt-
ing based on EEG change (64). Conversely 2 of 66 patients who
had SP 50 mm Hg, 2 patients needed shunting. If a shunting
policy has been used on a CSP of 50 mm Hg, 30 patients would
have been shunted unnecessarily (positive predictive value of
12%), whereas CSP 50 mm Hg would determine that a shunt
was not required in 64 of 66 patients (negative predictive value of
97%). The authors concluded that CSP should not be used as the
sole predictor for shunting for CEA (64).
Chapter 39 ■ Monitoring EEG during Carotid Surgery 801
Transcranial Doppler (TCD)
TCD monitoring of middle cerebral artery flow velocity can
also assess the dependence of the ipsilateral hemisphere on the
middle cerebral artery blood supply (25,73,74). Stenosis or
occlusion of the contralateral carotid artery was associated with
a large middle cerebral artery velocity decrease upon clamping
(25). TCD can also detect signals representing embolic flow
after clamp release (73). Increased numbers of emboli on
closure may warn of impending luminal thrombosis (75). TCD
may also be useful to identify postoperative risk of hyperperfu-
sion syndrome (76). Using a 75% reduction of mean blood flow
of middle cerebral artery as the threshold for shunt placement,
one study showed one true-positive and four false-positive and
no false-negative results from 91 patients (77). In this study
EEG yielded one false-negative and no false-positive results,
and SEP had no false-positive or false-negative results. Another
study also showed higher false-positive results in TCD (78). Of
85 patients, 13 showed a drop of 70% and only one of them
required shunt based on EEG and SSEP. The authors concluded
that using a drop of 70% or more of the systolic blood flow
velocity in the middle cerebral artery during internal carotid
artery cross-clamping would lead to excessive unnecessary
shunting. On the other hand, another study reported a dra-
matic reduction of TCD in association with normal EEG results
in three patients with a resulting transient postoperative neuro-
logic deficit (79).
Overall TCD appears to show more false-positive finding for
the prediction of ischemia. Another problem of TCD is the rela-
tively high rate of unsuccessful recordings ranging from 20% to
40% (72,77). Perhaps TCD is more useful in detecting embolic
and cerebral hyperperfusion after endarterectomies (80,81).
Xenon-133 Washout Study
The xenon washout study is an accurate method of measuring
cerebral blood flow. Xenon-133 is injected into either internal
carotid artery above the plaque or the common carotid artery
below the plaque with external carotid artery occlusion. EEG
changes have been found to occur with blood flow at or below
18 to 20 mL/100 g brain tissue/min (20). The method, however,
is not widely available and is less commonly used for CEAs.
Somatosensory-Evoked Potential (SSEP)
SSEP is dependent upon the integrity of sensory pathways start-
ing from the peripheral nerve to the sensory cortex via the
internal capsule and thalamus. Unlike EEG that is not sensitive
to detect small subcortical ischemia, SSEP can reflect ischemic
changes occurring in small areas of the capsular region, which
is important for neurologic function. However, one should real-
ize that the area covered by SSEPs is relatively small as com-
pared to EEG.
One of the disadvantages of SSEP is the technical challenge
of recording such small potentials in an electrically hostile
environment common to an operating room. One encounters
technical difficulties more often with SSEP recordings than
with EEG recording. Electrical noise or artifacts are relatively
easy to recognize in the EEG recording because of its continu-
ous nature. Artifacts or electrical noise that are often difficult
 Figure 39.5 An 83-year-old woman with history of stroke and TIAs affecting the left hemisphere has a 90% stenosis of
the left carotid artery. CEA was performed during the awake state with regional anesthesia. A: The EEG showed preexisting
delta activity over the left hemisphere. B: Within 30 seconds after the clamp, there were increases of delta slow waves and
decreases of alpha-theta background activity over the left hemisphere.
802
 Chapter 39 ■ Monitoring EEG during Carotid Surgery 803

Figure 39.5 (continued) C: This persisted until a shunt was placed. Despite obvious EEG changes, the patient remained
asymptomatic during the clamp period. D: Final recovery was achieved about 3 minutes after the shunt was placed.
804 Part V ■ Complementary and Special Techniques
to identify can contaminate SSEP responses. These artifacts in
turn can change the response making it appear as if real
pathology is present. Unlike EEG that shows changes instanta-
neously, SSEP recordings require averaging the responses.
Thus, there is a delay of a few minutes to yield one response
and an additional few minutes delay to verify the response
changes. Another disadvantage of SSEP is that the recording
often cannot be performed in the patients who have peripheral
neuropathy or impaired peripheral nerve conduction of the
upper extremities, which is not uncommon in the elderly
patient’s population. Nonetheless, SSEP monitoring has shown
favorable results for the purpose of monitoring to detect cere-
bral ischemia.
Comparing the SSEP, EEG, and TCD in 156 patients, EEG
yielded one false-negative and no false-positive results, and
TCD showed four false-positive and no false-negative results,
whereas SEP had no false-positive or false-negative results (77).
The criteria used for determining the need for shunting was
greater than 50% amplitude reduction of the N20/P25 potential
using median nerve SSEP. Using the same criteria, another
study also supported the SEP to be most reliable in warning of
ischemia, as compared to TCD or spectral edge frequency
analysis of EEG (82). The same criteria could be used in
patients who had history of previous stroke (83).
SSEP monitoring in addition to EEG may indeed help to
identify the subcortical ischemia not detected by EEG, which in
turn would promote the accuracy of warning for cerebral
ischemia.
Quantitative EEG Analysis (q-EEG)
The visually assessed EEG changes indicative of cerebral
ischemia are reflected by a reduction or amplitude decrease of
fast activity and an increase of delta/theta bands activity. The
reliable assessment and the accuracy are dependent on the skill
and experience of electroencephalographers. The type, magni-
tude, and duration of EEG changes needed to accurately predict
cerebral ischemia have not been clearly established. Thus the
precise EEG criteria for shunt requirement are still a matter of
debate. Also the reasons for false-positive or false-negative EEG
results have not been clarified.
Since most EEGs are now performed by digital recording
techniques, various quantitive EEG (q-EEG) analyses using
appropriate algorithms are possible, either on- or off-line.
These include power spectrum frequency analysis, spectral edge
frequency parameter, brain symmetric index (BSI), bispectral
index profile (BIS), and brain mapping. It is hoped that q-EEG
will provide a more accurate and objective measure to reflect
the subtle EEG changes indicative of cerebral ischemia with the
goal of identifying when to use a shunt with more accuracy.
This, in turn, is expected to reduce the false-positive and false-
negative EEG finding for the prediction of perioperative neuro-
logic deficit.
A fast Fourier transform (FFT) is used to generate spectral
power of specified or pre-selected frequency bands. The den-
sity spectral array (DSA) and/or spectral edge frequency (SEF)
derived from FFT data have been used as monitors during
CEA. The DSA is a three-dimensional data display designed to
express the power spectrum amplitude of a range of frequen-
cies along a time axis. The X-axis is designated as the time
scale and the Y-axis expresses both the power scale by color-
coded differential or gray scale differential and frequency by
height along the Y-axis. The SEF is expressed as a percentage
of frequency power at 90% to 97% of total power and is
defined to be the highest frequency component of the EEG
visible in the current spectrum. Using DSA and SEF on a sin-
gle channel of EEG (without conventional EEG recording and
without shunt placement irrespective of q-EEG findings), one
earlier study showed all 7 out of 70 neurologically intact
patients preoperatively who had the changes in SEF developed
postoperative neurologic deficit (84). The criteria used
included a decrease in high frequencies by SEF lasting longer
than 10 minutes where the amplitude was less than 50% of its
average value over the preceding moments. In the same study,
however, 31 patients who had preoperative neurologic deficit
showed one false-positive and one false-negative result using
the same criteria.
One study compared visually inspected EEG with q-EEG
analyses, which included total power, SEF, spectral variance, and
log spectral variance. Overall q-EEG showed significant changes
only in patients who had severe degree of slow waves (continu-
ous delta or persistent amplitude attenuation, or both) that was
evident visually as an EEG abnormality (85). Of the various
q-EEG parameters, total power was most sensitive and SEF was
least sensitive in correlating with EEG findings. Another study
also reported that the total power spectrum was more sensitive
than SEF as an indicator of cerebral ischemia (86). In another
study, two electroencephalographers independently interpreted
DSA and conventional EEG from 103 CEA patients under gen-
eral anesthesia (87). One interpreted the significant DSA
change in 21 of 29 patients (sensitivity of 72% and specificity of
96%). The other identified significant DSA change in 16 of
27 patients (sensitivity of 59% and specificity of 96%). Failure
of DSA to detect potential ischemia was greater in patients who
had mild degrees of EEG change. The authors concluded that
DSA did not reliably detect mild EEG pattern changes indica-
tive of cerebral ischemia and is not a reliable substitute for
16-channel conventional EEG (87).
Separating fast (factor 1: beta and alpha) and slow (factor 2:
theta and delta) band frequencies obtained slightly better
results of power spectrum analysis (88). Decrease of factor 1
and increase of factor 2 was considered indicative of cerebral
ischemia. Minor ischemia was distinguished by decrease of fac-
tor 1. Of 16 patients who had significant decrease of factor 1,
15 patients correlated with EEG that showed a visual change
and were subsequently shunted. The factor 2 was, however, not
revealing. Another study measured the 8 to 15 Hz band power
before and after the carotid clamp and calculated the reduction
ratio of this band power after the clamp that reflected the
degree of power change and designated it as “the desynchro-
nization index” (D-index) (89). In 47 consecutive patients with
CEAs, the major EEG changes (11 patients) found were severe
depression or loss of 8- to 15-Hz activity and/or at least
twofolds increase of slow delta ( 1 Hz) noted on visual inspec-
tion. This was consistently associated with a D-index of greater

than 65% (mean of 76.85%). The authors concluded that the
D-index of greater than 65% correctly identified the patient
with cerebral hypoxic risk. The moderate EEG changes
(4 patients) with decreased but persisting 8- to 15-Hz activity
and/or persistent increase of delta activity ( 1 Hz) were asso-
ciated with mean D-index of 40.23%. Because two patients of
this group awoke without any neurologic deficit despite no
shunting used, the potential risk of D-index reduction of about
65% to 50% could not be determined.
Comparing the difference of power spectrum before and
after cross-clamping appears to show a better result in distin-
guishing between shunt and nonshunt group (90). However,
there were some differences in sensitivity between two anesthet-
ics; the difference of absolute power spectrum was better when
propofol anesthesia was used and SEF 90% was a better moni-
tor when isoflurane anesthesia was used.
Bispectral Analysis (BSA)
The power spectral analysis quantifies only power distribution as
a function of frequency, ignoring phase information. It also
assumes that the signal arises from a linear process and ignores
any nonlinear process of interaction between the components of
a signal that are manifested as phase coupling. BSA is a signal-
processing technique that determines both EEG linear (frequency
and power) and nonlinear components (harmonics) which
quantitates the interfrequency phase coupling of EEG signals.
BSA has primarily been used by anesthesiologists to monitor
the depth of anesthesia by placing a few electrodes over the fore-
head (Fp1, Fp2, F3, F4) usually referenced to the Cz electrode
(88,91,92). Earlier studies suggested potential usefulness of BSA
as an application for CEA to determine the adequacy of cerebral
perfusion (93,94). However, the more recent studies have
doubted the reliability of BSA in detecting cerebral ischemia.
Of the 52 patients monitored with BSA during awake CEA,
5 patients showed clinical evidence of cortical ischemia during
cross-clamping of the carotid but there was no change in BSA
value in these five patients (95). In another recent study, BSA
increased in 47%, decreased in 25%, and unchanged in 28% of
36 patients tested during first 3 minutes after cross-clamping of
the carotid (96). Because of this paradoxical increase of BSA in
some patients, the authors suggested that caution should be used
in the interpretation of BSA value during CEA.
Brain Symmetric Index (BSI)
The BSI is determined by the relative difference of the average
spectral density of the right and left hemispheres in the fre-
quency range from 1 to 25 Hz. The perfect symmetry of all
change is 0, whereas BSI = 1 indicates maximum asymmetry.
The authors who introduced BSI found that a change in the BSI
smaller than 0.03 during cross-clamping corresponded with no
change in visual EEG analysis and a change of BSI equal to or
greater than 0.06 showed EEG change that warranted shunting
(97). Because BSI depends only on the asymmetry of the EEG
spectrum spatial brain symmetric index (sBSI) and is insensi-
tive for temporary changes in the EEG, the same author intro-
duced temporal brain symmetry index (tBSI) (98,99). The
authors found that diffuse changes noticed by visual inspection
Chapter 39 ■ Monitoring EEG during Carotid Surgery 805
of EEG were reflected by changes on the temporal symmetry
measure with tBSI greater than 0.02. A large-scale clinical study
is needed to find out if the above BSI parameters accurately pre-
dict cerebral ischemia or provide more sensitive information
with respect to the criteria for shunt requirement.
Computed Topographic Brain Mapping (CTBM)
CTBM is based on power/amplitude spectrum derived from
Fourier transform measured at multiple scalp electrodes. This is
displayed on two-dimensional scalp figure. There have been a
few studies claiming that CTBM is superior to 16-channel stan-
dard EEG in detecting cerebral ischemia.
One study involving 46 consecutive CEA patients showed
that CTBM changes secondary to ischemia were noted in
23 patients, whereas visually inspected conventional EEG
showed changes only in 13 patients (100). The ischemic
change detected by CTBM was significantly greater in
patients with previous stroke (6 out of 7 patients) than the
patients without stroke (17 out of 39 patients). CTBM
changes were also correlated with average stump pressure of
less than 38 mm Hg, compared to 57 mm Hg in patients
without CTBM change (P 0.05). Overall, 51 of 138 CTMB
studies (37%) showed “abnormalities,” whereas standard
EEG showed abnormality in only 23 studies (17%). The
authors concluded that CTBM was much more sensitive
than the 16-channel standard EEG in showing electrophysi-
ologic changes, particularly during times of carotid cross-
clamping.
Another CTBM study for CEA showed that 10 patients
(15%) of 65 CEAs had ischemic change after cross-clamping
(101). The changes were resolved after shunt placement in all
patients. This study also showed a higher incidence of ischemic
change in patients who had abnormal preoperative CTBM as
compared to those with normal CTBM (102).
Although both studies claimed that CTBM was more sensi-
tive for selective shunting than EEG and predicted more accu-
rately postoperative neurologic deficits, no specific and
objective criteria for selective shunting was provided. Also the
number of patients involved in these studies were small. A
larger scale study is needed for further validation of CTBM as a
monitor for CEA.
Summary of q-EEG Assessment
It is hoped that q-EEG will reveal the EEG change not detected
by visual inspection and more accurately predicts cerebral
ischemia. However, none of the q-EEG parameters studied so
far have shown greater accuracy than conventional visually
inspected EEG in predicting cerebral ischemia, that is, no false-
positive or negative finding for perioperative neurologic
deficits. Another shortcoming of q-EEG is the artifact contam-
ination, which cannot be readily identified and may lead to
erroneous conclusions if based solely on q-EEG data. Although
the above discussed q-EEG parameters may assist in the visual
assessment of raw EEG data for selective shunting in CEA, at
present none of the q-EEG parameters can replace the visual
inspection of 16-channel EEG interpreted by a qualified elec-
troencephalographer.
806 Part V ■ Complementary and Special Techniques
APPENDIX
CEA and Carotid Angioplasty and Stenting (CAS)
More than 10 years has passed since the introduction of CAS
and it continues to grow. The surge of CAS has dramatically
reduced the request of EEG monitoring for carotid surgery in
many EEG laboratories. The advantages of CAS are less costly
than CEA and can be applied to the patients having high cervi-
cal and surgically inaccessible stenosis, or patients who have
significant medical conditions that would contraindicate CEA
performed under general anesthesia such as recent myocardia
infarction, unstable angina, or uncontrolled congestive heart
failure.
In a recent large-scale multicenter randomized study
involving 3182 patients, CAS seemed to carry a slightly higher
risk for stroke within 30 days after the procedure as compared
with CEA (8.2% vs. 6.2%, P = 0.04), whereas the rates of non-
fatal myocardial infarction or death within 30 days did not
differ significantly (P = 0.47) (103). Another large-scale study
of randomized controlled trials from 3182 patients showed
that CAS had no significant increase of any stroke or major
disabling stroked compared to CAE (104). The risks of death
and nonfatal myocardial infarction were the same between
CAE and CAS. The large scale of asymptomatic patients is not
available, but one study of 326 patients showed no statistical
difference between CAS and CEA: stroke 2 patients (1.7%)
versus 2 patients (1%), 2 myocardial infarction (1.7%) versus
3 (1.5%) and 1 death (0.8%) versus 0 death, respectively
(104). Recent long-term follow-up study comparing CEA and
CAS revealed that significantly higher rate of re-stenosis
( 70%) (6 of 32 vs. 0 of 29) occurred after CAS compared
with CEA (105). Large-scale multicenter trials are needed for
more accurate assessment of asymptomatic patients and long-
term follow-up after CAS. In recent years, there has been an
increasing trend for the use of CEA, perhaps due to more
careful selection of CAS candidates.
REFERENCES
1. North American Symptomatic Carotid Endarterectomy Trial
Collaborators. Beneficial effect of carotid endarterectomy in
symptomatic patients with high-grade carotid stenosis. N Eng
J Med. 1991;325:445–453.
2. Europian Carotid Surgery Trialists’ Colloboration Group. MRC
Europe Carotid Surgery Trial: interim results for symptomatic
patients with severe (70%–99%) or with mid (0%–29%) carotid
stenosis. Lancet. 1991;337:1235–1243.
3. Barnett HJM, Meldrum H, Eliaziw M, et al. Treatment of sympto-
matic arteriosclerotic carotid artery disease. In: Barnett HJM,
Bogousslavsky J, Meldrum H, eds. Advances in Neurology Ischemic
Stroke, Vol. 92. Philadelphia: Lippincott William & Wilkins;
2003:307–317.
4. Hobson RW, Weiss DG, Fields, WS, et al., for the Veterans Affairs
Cooperative Study Group. Efficiency of carotid endarterectomy of
asymptomatic carotid stenosis. N Eng J Med. 1993;328:221–227.
5. Executive Committee for the Asymptomatic Carotid Athero-
sclerosis Study. Endarterectomy for asymptomatic carotid artery
stenosis. JAMA. 1995;273:1421–1428.
6. Norris JW, Zhu CZ, Borenstein NM, et al. Vascular risks of asymp-
tomatic carotid stenosis. Stroke. 1992;22:1485–1490.
7. Rutgers DR, Klijn CJM, Kappelle, et al. Sustained bilateral hemo-
dynamic benefit of contralateral carotid endarterectomy in patient
with symptomatic internal carotid artery occlusion. Stroke. 2001;
32:728–734.
8. Findlay JM, Marchak BE, Pelz DM, et al. Carotid endarterectomy:
a review. Can J Neurol Sci. 2004;31:22–36.
9. Giannotta SL, Dicks RE, Kindt GW. Carotid endarterectomy:
technical improvements. Neurosurgery. 1980;7:309–312.
10. Ott DA, Cooley DA, Chapa L, et al. Carotid endarterectomy with-
out temporary intraluminal shunt. Ann Sug. 1980;191:708–714.
11. Baker WH, LiHooy FN, Hayes AC, et al. Carotid endarterectomy
without a shunt: the control series. J Vasc Surg. 1984;1:50–56.
12. Ferguson GG. Intra-operative monitoring and internal shunt: are
they necessary in carotid endarterectomy? Stroke. 1982;13:287–289.
13. Steed DL, Peitzman AB, Grundy BL, et al. Causes of stroke in
carotid endarterectomy. Surgery. 1982;92:634–641.
14. Riles TS, Imparato AM, Jacobowitz GR, et al. The cause of periop-
erative stroke after carotid endarterectomy. J Vasc Surg. 1994;19:
206–216.
15. Pletis KA, Loubser P, Mizrahi EM. Continuous electroencephalo-
graphic monitoring and selective shunting reduce neurologic mor-
bidity rate in carotid endarterectomy. J Vasc Surg. 1997;25:620–628.
16. Dogle PW, Cole JP, Leary TM, et al. A comparison of remifentanil
and fentanyl in patients undergoing carotid endarterectomy. Eur
J Anesthesiol. 2001;18(1):13–19.
17. Laman DM, vander Reijden CS, Wieneke GH, et al. EEG evidence
for shunt requirement during carotid endarterectomy: optimal
EEG derivation with respect to frequency bands and anesthetic
regimen. J Clin Neurophysiol. 2001;18(4):353–363.
18. Blume WT, Sharbrough FW. EEG monitoring during carotid
endarterectomy and open heart surgery. In: Niedermeyer E, Lopes
de Silva F, eds. Electroencephalogrphy, Basic Principles, Clinical
Application and Related Fields. 5th ed. Philadelphia: Lippincott
Williams & Wilkins; 2005:815–828.
19. Stockard JJ, Bickford RG, Myers RR. The neurophysiology of anes-
thesia. In: Gordon E, ed. A Basis and Practice of Neuroanesthesia.
Amsterdam: Excepta Medica; 1975:3–46.
20. Sundt TM, Sharbough FW, Piepgras DG, et al. Correlation of cere-
bral blood flow and electroencephalographic changes during
carotid endarterectomy. Mayo Clin Proc. 1981;56:533–543.
21. Blume WT, Ferguson GG, McNeill DK. Significance of EEG
changes at carotid endarterectomy. Stroke. 1986;17:891–897.
22. Chiappa KH, Burke SR, Young RR. Results of electroencephalo-
graphic monitoring during 367 carotid endarterectomies. Stroke.
1979;10:381–388.
23. Pinkerton JA Jr. EEG as a criteria for shunt in carotid endarterec-
tomy. Ann Vasc Surg. 2002;16(6):756–761.
24. Schneider JR, Droste JS, Schendler N, et al. Carotid endarterec-
tomy with routine electroencephalography and selective shunting:
influence of contralateral internal carotid artery occlusion and
utility in prevention of perioperative strokes. J Vas Surg.
2002;35(6):1114–1122.
25. Arnold M, Sturzenegger M, Schaffler L, et al. Continuous intraop-
erative monitoring of middle cerebral artery blood flow velocities
and electroencephalography during carotid endarterectomy. A
comparison of the two methods to detect cerebral ischemia.
Stroke. 1997;28(7):1345–1350.
26. Whittemore AD, Kauffman JL, Kohler TR, et al. Routine electroen-
cephalographic (EEG) monitoring during carotid endarterectomy.
Ann Surg. 1983;197:707–713.

27. Green RM, Messick WJ, Ricotta JJ, et al. Benefits, short-comings,
and costs of EEG monitoring. Ann Surg. 1985;201:785–792.
28. Cherry KJ, Roland CF, Hallett JW Jr, et al. Stump pressure, the con-
tralateral carotid artery, and electroencephalographic changes. Am
J Surg. 1991;162(2):185–188.
29. Kresowik TF, Worsey MJ, Khonry MD, et al. Limitations of elec-
troencephalographic monitoring in the detection of cerebral
ischemia accompanying carotid endarterectomy. J Vasc Surg.
1991;13:439–443.
30. Thompson JE, Austin DJ, Patman RD. Carotid endarterectomy for
cerebrovascular insufficiency: long-term results in 592 patients
followed up to thirteen years. Ann Surg. 1970;172:663–679.
31. Whitney DG, Kahn EM, Estes JW, et al. Carotid artery surgery
without temporary indwelling shunt: 1917 consecutive proce-
dures. Arch Surg. 1980;115:1393–1399.
32. Ojemann RG, Crowell RM, Robertson GH, et al. Surgical treat-
ment of extracranial carotid occlusive disease. Clin Neurosurg.
1975;22:214–263.
33. Lesage R, Paris E, Koskas F, et al. Surgical reconstruction of the
internal carotid artery with contralateral occlusion without use of
a shunt. Ann Vasc Surg. 1991;52:55–60.
34. Redekop G, Ferguson G. Correlation of contralateral stenosis and
intraoperative electroencephalographic changes with risk of stroke
during carotid endarterectomy. Neurosurgery. 1992;30:191–194.
35. Archie JP Jr. A fifteen-year experience with carotid endarterectomy
after a formal operative protocol requiring highly frequent patch
angioplasty. J Vasc Surg. 2000;31:724–735.
36. Woodworth GF, McGirt MJ, Than KD, et al. Selective versus rou-
tine intraoperative shunting during carotid endarterectomy, a
multi-variant outcome analysis. Neurosurgery. 2007;61(6):
1176–1177.
37. Imparato AM, Ramirez A, Riles T, et al. Cerebral protection in
carotid surgery. Arch Surg. 1982;117:1073–1078.
38. Grego F, Antonello M, Lepidi S, et al. Is contralateral carotid artery
occlusion a risk factor for carotid endarterectomy? Ann Vasc Surg.
2005;19(16):882–889.
39. Cikrit DG, Burt RW, Dalring EG, et al. Acetazolaminde enhanced
single photo emission computed tomography (SPECT) evaluation
of cerebral perfusion before and after carotid endarterectomy.
J Vas Surg. 1992;15:747–754.
40. Araki CT, Babikian VL, Cantelmo NL, et al. Cerebrovascular
hemodynamic changes associated with carotid endarterectomy.
J Vasc Surg. 1991;13:854–860.
41. Blankensteijn JD, Van der Grand J, Mali WPthM, et al. Flow vol-
ume changes in major cerebral arteries before and after carotid
endarterectomy: an MR angiography study. Eur J Vas Endovasc
Surg. 1997;14:446–450.
42. Irvine CD, Gardner FV, Davies AH, et al. Cognitive testing in
patients undergoing carotid endarterectomy, Eur J Vas Endosvasc
Surg. 1998;15:195–204.
43. Ucle’s P, Almarcegui S, Lorente S, et al. Evaluation of cerebral func-
tion after carotid endarterectomy. J Clin Neruophysiol.
1997;14(3):242–249.
44. Vriens EM, Wieneke GA, Van Huffelen AC, et al. Increase in alpha
rhythm frequency after carotid endarterectomy. Clin Neurophysiol.
2000;111(8):1505–1513.
45. Stoughton J, Nath RL, Abbott WM. Comparison of simultaneous
electroencephalographic and mental status monitoring during
carotid endarterectomy with regional anesthesia. J Vasc Surg.
1998;28:1014–1023.
46. Cheng MA, Theard MD, Templehoff R. Anesthesia for carotid
endarterectomy: a survey. J Neurosurg Anesthesiol. 1997;9:211–216.
Chapter 39 ■ Monitoring EEG during Carotid Surgery 807
47. Lawrence PF, Alves JC, Jicha D, et al. Incidence, timing and causes
of cerebral ischemia during carotid endarterectomy with regional
anesthesia. J Vasc Surg. 1998;27:329–337.
48. Evans WE, Hayes JP, Waltke EA, et al. Optimal cerebral monitoring
during carotid endarterectomy: neurologic response under local
anesthesia. J Vasc Surg. 1985;2:775–777.
49. Santamaria G, Birtti RD, Tescione M, et al. Comparison between
local and general anesthesia for carotid endarterectomy: a retro-
spective analysis. Minerva Anesthesiol. 2004;72(12):1013–1014.
50. Illig KA, Sternbach Y, Zhang R, et al. EEG changes during awake
carotid endarterectomy. Ann Vasc Surg. 2002;16:6–11.
51. Hans SS, Jareupoon O. Prospective evaluation of electroen-
cephalography, cerebral artery stump pressure and neurologic
change during 314consequitive carotid endarterectomies per-
formed in awake patients. J Vasc Surg. 2007;45(3):511–515.
52. Kassell NF, Hitchon PW, Gerk MK, et al. Alteration in cerebral
blood flow, oxygen metabolism and electrical activity produced by
high dose sodium thiopental. Neurosurgery. 1980;7:598–603.
53. Spetzler RF, Martin N, Hadley MN, et al. Microsurgical
endarterectomy under barbiturate protection: a prospective study.
J Neurosurg. 1986;65(1):63–73.
54. Frawley JE, Hicks RG, Gray U, et al. Carotid endarterectomy with-
out shunt for symptomatic lesions associated with contralateral
severe stenosis or occlusion. J Vasc Surg. 1996;23(3):421–427.
55. Gross CE, Adams HP, Sokoll MD, et al. Use of anticoagulants, elec-
troencephalographic monitoring and barbiturate cerebral protec-
tion in carotid endarterectomy, Neurosurgery, 1981; 9(1): 1-5.
56. D’Angelo V, Catapano G, Bozzini V, et al. Propofol EEG burst
suppression in carotid endarterectomy. J Neurosurg Sci. 2001;
45:159–162.
57. Melgar MA, Mariwalla N, Madhusudan H. Carotid endarterec-
tomy without shunt: the role of cerebral metabolic protection.
Neurol Res. 2005;27(8):850–856.
58. Williams IM, Picton A, Farrell A, et al. Light-reflective cerebral
oximetry and jugular bulb venous oxygen saturation during
carotid endarterectomy. Br J Surg. 1994;81:1291–1295.
59. Beese U, Langeer H, Lang W, et al. Comparison of near-infrared
spectroscopy and somatosensory evoked potentials for the detec-
tion of cerebral ischemia during carotid endarterectomy. Stroke.
1998;29:2032–2037.
60. Ferrari M, Zanette E, Sideri G, et al. Effects of carotid compression, as
assessed by near infrared spectroscopy, upon cerebral blood volume
and hemoglobin oxygen saturation. J R Soc Med. 1987;80:83–87.
61. Samara SK, Dy EA, Welch K, et al. Evaluation of a cerebral oxime-
ter as a monitor of cerebral ischemia during carotid endarterec-
tomy. Anesthesiology. 2000;93:964–970.
62. Oyama H, Endoh O, Iizuka H, et al. The effectiveness of regional
cerebral oxygen saturation monitoring using near-infrared spec-
troscopy in carotid endarterectomy. J Clin Neurosci.
2003;10(1):79–83.
63. Friedell ML, Clark JM, Graham DA, et al. Cerebral oximetry does
not correlate with electroencephalography and somatosensory
evoked potential in determining the need for shunting during
carotid endarterectomy. J Vasc Surg. 2008;48:601–606.
64. Botes K, Le Roux DA, Van Marle J. Cerebral monitoring during
carotid endarterectomy – a comparison between electroen-
cephalography, transcranial cerebral oximetry and carotid stump
pressure. S Afr J Surg. 2007;45(2):43–46.
65. Rigamonti A, Scandroglio M, Minicucci F, et al. A clinical evalua-
tion of near-infrared cerebral oximetry in the awake patients to
monitor cerebral perfusion during carotid endarterectomy. J Clin
Anesth. 2005;17(6):426–430.
808 Part V ■ Complementary and Special Techniques
66. Vets P, ten Broeke P, Adriaensen H, et al. Cerebral oximetry in
patients undergoing carotid endarterectomy: preliminary results.
Acta Anesthesiol Belg. 2004;55(3):215–220.
67. Wilkinson HA, Wright RL, Sweet WH. Correlation of reduction in
pressure and angiographic cross-filling with tolerance of carotid
occlusion. J Neurosurg. 1965;22:241–245.
68. Beebe HG, Starr C, Slack D. Carotid artery stump pressure: its vari-
ability when measured serially. J Cardiovasc Surg. 1989;30:419–423.
69. Harada RN, Comerota AJ, Good GM, et al. Stump pressure, elec-
troencephalographic changes, and the contralateral carotid artery:
another look at selective shunting. Am J Surg. 1995;170:148–153.
70. Calligaro KD, Dougherty MJ. Correlation of carotid artery stump
pressure an neurological changes during 474 carotid endarterec-
tomies performed in awake patients. J Vas Surg. 2005;42(4):684–689.
71. Jacob T, Hingorani A, Ascher F. Carotid artery Stump Pressure
(CASP) in 1135 consecutive endarterectomies under general anes-
thesia: an old method that survived the test of time. J Cardiovasc
Surg. 2007;48(6):677–681.
72. Moritz S, Kasprzak P, Arlt M, et al. Accuracy of cerebral monitor-
ing in detecting cerebral ischemia during carotid endarterec-
tomies: a comparison of transcranial Doppler sonography,
near-infrared spectroscopy, stump pressure, and somatosensory
evoked potentials. Anesthesiology. 2007;107(4):563–569.
73. Fiori L, Parenti G, Marconi, F. Combined transcranial Doppler and
electrophysiologic monitoring for carotid endarterectomy.
J Neurosurg Anesthesiol. 1977;9(1):11–16.
74. McCarthy RJ, Nasr MK, McAteer P, et al. Physiological advantages
of cerebral blood flow during carotid endarterectomy under local
anesthesia. A randomized clinical trial. Eur J Vas Endvasc Surg.
2002;24(3):215–221.
75. Ackerstaff RG, Moons KG, van de Vlasakker CJ, et al. Association
of intraoperative transcranial Doppler monitoring variables with
stroke from carotid endarterectomy. Stroke. 2000;31:1817–1823.
76. Dalman JE, Beenokkers ICM, Moll FL, et al. Transcranial Doppler
monitoring during carotid endarterectomy helps to identify
patients at risk of postoperative hyperperfusion. Eur J Vas
Endovasc Surg. 1999;18:222–227.
77. Rowed DW, Houlden DA, Burkholder LM, et al. Comparison of
monitoring techniques for intraoperative cerebral ischemia. Can J
Neurol Sci. 2005;32(1):131–132.
78. Lacroix H, Beyens G, Van Hemelrijck J, et al. Is transcranial
Doppler useful in the detection of internal carotid artery cross-
clamp intolerance? Cardiovasc Surg. 1999;7(2):203–207.
79. Costin M, Rampersad A, Soloman RA, et al. Cerebral injury pre-
dicted by transcranial Doppler ultrasonography but not electroen-
cephalography during carotid endarterectomy. J Neurosurg
Anesthesiol. 2002;14(4):287–292.
80. Dunne VG, Besser M, Ma WJ. Transcranial Doppler in carotid
endarterectomy. J Clin Neurosci. 2001;8(2):140–145.
81. Totaro R, Varroni A, Gizzi E, et al. Transcranial Doppler sonogra-
phy in the pre-intra- and post-operative evaluation of 85 patients
undergoing carotid endarterectomy. Clin Ter. 1998;149:267–270.
82. Stejskal L, Kramar F, Ostry S, et al. Experience of 500 case of neu-
rophysiological monitoring in carotid endarterectomy. Acta
Neuroclir (Wein). 2007;149(7): 681–688.
83. Manninen PH, Tan TK, Sarjeant RM. Somatosensory evoked
potential monitoring during carotid endarterectomy in patients
with a stroke. Anesth Analg. 2001;93(1):39–44.
84. Rampil IJ, Holzer JA, Quest DQ, et al. Prognostic value of comput-
erized EEG analysis during carotid endarterectomy. Anesth Analg.
1983;62:186–192.
85. Young WL, Moberg RS, Ornstein E, et al. Electroencephalographic
monitoring for ischemia during carotid endarterectomy: visual
versus computer analysis. J Clin Mont. 1988;4:78–85.
86. Hanowell LH, Soriano S, Bennett HL. EEG power changes are
more sensitive than spectral edge frequency variation for detec-
tion of cerebral ischemia during carotid artery surgery: a
prospective assessment of processed EEG monitoring.
J Cardiothorac Vasc Anesth. 1992;6:292–294.
87. Kearse LA, Martin D, McPeck K, et al. Computer-derived density
spectral array in detection of mild analog electroencephalo-
graphic ischemic pattern changes during carotid enterectomy.
J Neurosurg. 1993;78:884–890.
88. Visser GH, Wieneke GH, van Huffelen AC. Carotid endarterec-
tomy monitoring: patterns of spectral EEG changes due to
carotid artery clamping. Clinical Neurophysiol. 1999;110:286–
294.
89. Cursi M, Meraviglia MV, Fanelli GF, et al. Electroencephalographic
background desynchronization during cerebral blood flow reduc-
tion. Clin Neurophysiol. 2005;116(11):2577–2785.
90. Larman DM, Wieneke GH, van Duijin H, et al. Q-EEG changes
during carotid clamping in carotid endarterectomy: spectral edge
frequency parameters and relative band power parameters. J Clin
Neurophysiol. 2005;22(4):244–252.
91. Liu J, Singh H, White PF. Electroencephalographic bispectral
analysis predicts the depth of midazolam induced sedation.
Anesthesiology. 1996;84(1):64–69.
92. Hoffman WE, Zsigmond E, Albrecht RF. The bispectral index
during induction of anesthesia with midazolam and propofol.
J Neurosurg Anesthesiol. 1996;8(1):15–20.
93. el-Dawlatly AA. EEG bispectral index during carotid endarterec-
tomy. Middle East J Anesthesiol. 2003;17(2):287–293.
94. M’erat S, L’evecque JP, Le Gulluche, et al. BIS monitoring may
allow the detection of serve cerebral ischemia. Can J Anesth.
2001;48(11):1055–1060.
95. Deogaonkar A, Vivar R, Bullock RE, et al. Bispectral index mon-
itoring may not reliably indicate cerebral ischemia during awake
carotid endarterectomy. Br J Anesth. 2005;94(4):559–560.
96. Bonhomme V, Desiron Q, Lemineur T, et al. Bispectral index pro-
file during carotid cross clamping. J Neurosurg Anesthesiol. 2007;
19(1):49–55.
97. Van Putten MJAM, Peters JM, Mulder SM, et al. A brain symme-
try index (BSI) for online EEG monitoring in carotid endarterec-
tomy. Clin Neurophysiol. 2004;115:1189–1194.
98. Van Putten MJAM. Extended BSI for continuous EEG monitoring
in carotid endarterectomy. Clin Neurophysiol. 2006;117:2661–2666.
99. Van Putten MJAM. The revised brain symmetry index. Clin
Neurophysiol. 2007;118:2362–2367.
100. Ahn SS, Jordan SE, Nuwer MR, et al. Computed electroen-
cephalographic topographic brain mapping. A new and accurate
monitor of cerebral circulation and function for patients having
carotid endarterectomy. J Vasc Surg. 1988;8:247–254.
101. Larkin M. Should endarterectomy decision be based on non-
invasive imaging? Lancet. 2001;357:1343.
102. Elmore JR, Eldrup-Jorgensen J, Leschey WH, et al. Computerized
topographic brain mapping during carotid endarterectomy. Arch
Surg. 1990;125:734–738.
103. Murad MH, Flynn DN, Elamin MB, et al. Endarterectomy vs
stenting for carotid artery stenosis: a systematic review and meta-
analysis. J Vasc Surg. 2008;48(2):487–493.
104. Tang GL, Matsumura JS, Morasch MO, et al. Carotid angioplasty
and stenting vs. carotid endarterectomy for treatment of asymp-
tomatic disease: single-center experience. Arch Surg. 2008;143(7):
653–658.
105. Steinbauer MG, Pfister K, Greindl M, et al. Alert for increased
long-term follow-up after carotid artery stenting: results of a
prospective, randomized, single-center trial of carotid artery
stenting vs. carotid endarterectomy. J Vasc Surg. 2008;48(1):93–98.

ANTON KAMP, GERT PFURTSCHELLER, GÜNTER EDLINGER, AND FERNANDO H. LOPES DA SILVA
P
olygraphy denotes the simultaneous recording of several
physiologic and/or behavioral variables. The main rea-
sons for the simultaneous recording of several variables
are to obtain information on behavioral aspects and to differen-
tiate artifacts in the electroencephalographic (EEG) data. These
objectives usually do not require precise representation; in many
instances, the relevant information concerns only the occur-
rence of a certain phenomenon or is easily obtained from clearly
discernible characteristics of a variable. Therefore, most poly-
graphic data of interest in EEG studies can be obtained using
simple recording methods that allow appreciable distortion of
the original data. If, however, the polygraphic variables are of
primary concern, then sophisticated and precise recording
methods are necessary. In view of the techniques used and the
interpretation of the recorded data, these methods go far beyond
the simple polygraphic methodology commonly applied to EEG
studies. This chapter discusses these simple methods and pres-
ents a number of examples of variables that are of interest in
certain EEG studies. A classic survey of variables of interest in
polygraphic studies and of corresponding recording methods
can be found in Manual of Psychophysiologic Methods, edited by
Venables and Martin (1).
In practice, polygraphic recordings are made with an EEG
apparatus; this may be of primary interest for determining the
temporal relations between the EEG and the other signals,
which reflect different physiologic functions and/or behavioral
states. Unfortunately, the frequency characteristics and the
magnitude ranges of many signals of interest for polygraphic
studies fall outside those provided by a conventional EEG
recording system; moreover, they might not be recordable due
to the electrical characteristics of the input circuit of the EEG
recorder. Such signals, therefore, require special provisions,
such as the use of specialized preamplifiers or input couplers to
obtain an adaptation or a conversion; in this way, the recording
of such signals can be carried out with the EEG apparatus.
Many variables of interest in polygraphic studies, such as blood
pressure, respiratory parameters, temperature, and electroder-
mal signals, vary slowly as a function of time; therefore, their
recording requires highly sensitive universal direct current
(DC) amplifiers that are equipped with means of sensitivity
control and adjustable high- and low-pass filters for selection of
the appropriate frequency response. Modern EEG recorders
have low sensitivity auxiliary input terminals that also permit
DC recording; these inputs can be used to record other signals
of sufficiently large amplitude (e.g., in Fig. 40.8, the traces indi-
cated by EDG, STIM, BUTTON PRESS, and TIME CODE).
When the EEG is used to record different physiologic and/or
CHAPTER
Polygraphy
40
behavioral variables simultaneously, employing separate
recording systems, the time relations between the various types
of signals must be preserved by using a form of time indexing
or time marking on both recording media.
CARDIOVASCULAR VARIABLES
Electrocardiogram and Heart Rate
There are several reasons for recording the electrocardiogram
(ECG) simultaneously with the EEG. It may be desirable in spe-
cific cardio- or cerebrovascular studies. In most cases, however,
the ECG recording is not intended to carry out a vascular study
but only serves as an indicator of ECG artifacts in EEG records
or as a general parameter of vegetative functions; in these latter
circumstances, one is mainly interested in the heart rate (HR).
An ECG can be recorded perfectly using an EEG system because
the electrical characteristics of its input circuit and the provi-
sions commonly available for adjustment of frequency response
and gain are adequate. The bandwidth required for appropriate
ECG recording goes from 0.8 to 60 Hz; the recording sensitivity
required is approximately 1 mV/cm using conventional ECG
electrode placements. When the ECG electrodes are placed on
the chest wall, a higher sensitivity may be necessary. The sub-
ject’s behavioral activities might lead to artifacts in the ECG
record, owing to muscular activity or electrode motion. The lat-
ter can be reduced significantly by using an appropriate type of
electrode, such as cup electrodes with a jelly bridge between
skin and electrode surface. Interference caused by electromyo-
graphic potentials can be minimized by choosing electrode
positions carefully and lowering the high-frequency response
of the recording system (20 Hz; 3 dB) in order to attenuate the
high-frequency electromyogram (EMG) potentials. High-
frequency filtering can be obtained by means of the EEG appa-
ratus’s adjustable high-frequency filters.
HR recording is best carried out by using a series of pulses
generated at the top of clearly distinguishable R waves of the
ECG. If, however, owing to less favorable electrode positions,
the R wave cannot be easily distinguished, extra signal process-
ing must be applied. This processing may consist of high-pass
filtering and/or the introduction of a refractory period, during
which the instrument is insensitive. Commercially available
heart frequency meters or HR counters usually have such pro-
visions. HR measurements may be given in terms of the num-
ber of beats over a certain period of time (HR) or in terms of
heart period (HP), the average interval between a number of
successive heartbeats. Because HR and HP are reciprocal, the
809
810 Part V ■ Complementary and Special Techniques

Figure 40.2 Principle of finger photophethysmography for transmis-

sion and reflexion of light.

in polygraphy. Figure 40.2 shows the principle of finger photo-

Figure 40.1 Method for recording instantaneous heart rate (HR); the
momentary heart period (HP) value, transformed into HR, is plotted in
relation to an adjustable preset mean HR value.
instrument, although calibrated in terms of HR, may have a
meter deflection or another output signal proportional to T int,
the interval between successive heartbeats. In the available
instruments, HR is more commonly presented than HP. In
some applications of polygraphy, the main interest is not in the
nominal HR but rather in HR changes and the relation to other
physiologic, psychological, or behavioral variables. So that rela-
tively small HR changes can be distinguished, the recording is
best carried out with a preadjusted preset HR value. The model
given in Figure 40.1 demonstrates a simple method for sub-
tracting a preset value from the HR meter’s electrical output.
plethysmography. Two photosensors measure the transmission
and reflection of the light emitted from a light-emitting diode.
The fraction of transmitted light through the tissue and the frac-
tion of reflected light from the tissue depend on the amount of
blood in the tissue. Extensive discussions of the measuring prin-
ciples, amplifier recorder requirements, and recorded waveforms
have been provided by Lader (3) (pneumatic plethysmography)
and Weiman (4) (photoplethysmography).
Impedance plethysmography of the thorax for impedance
cardiography is the basis for noninvasive beat-to-beat monitor-
ing of the stroke volume (5). An electric current is introduced
into the thorax and the corresponding voltage is measured. The
ratio of voltage to current yields the impedance (Z) that varies
(in a very simplified model) with the amount and distribution
of blood in the thorax. Based on the ECG, the phonocardiogram
(PCG), and impedance cardiogram (ICG), the stroke volume
can be determined noninvasively (Fig. 40.3).

Plethysmography
Plethysmography is the measurement of the variations in organ
or limb volume due to changes in the quantity of blood it con-
tains. Because such volume changes are related to increased or
decreased blood flow, plethysmographic methods can be used to
obtain estimates of the mean blood flow rate and of pulsatile and
transient flow changes. Plethysmography may be of interest in
psychophysiologic studies because mental processes and behav-
ioral responses are often accompanied by changes in such cardio-
vascular parameters as blood flow, accompanied by measurable
changes in limb volume. Continuous measurement of the latter
is known as pulse volume plethysmography; under certain
restricted conditions, an index of the blood flow rate can also be
obtained in this way (2). In most psychophysiologic studies, how-
ever, the variable of interest relates to changes in blood volume
and in blood volume pulses. The most common methods for Figure 40.3 Principle of impedance cardiography. (Adapted from
measuring limb volume changes are pneumatic and photoelec- Gratze G, Fortin J, Holler A, et al. A software package for noninvasive,
tric. Pneumatic methods are more complicated and are not real-time beat-to-beat monitoring of stroke volume, blood pressure,
suited to psychophysiologic studies. Therefore, although provid- total peripheral resistance and for assessment of autonomic function.
ing more precise information, they are much less frequently used Comp Biol Med. 1998;28:121–142.)

Figure 40.4 Block diagram of a system for noninvasive continuous
recording of blood pressure based on the Penas principle. (From
Wesseling KH, van Bemmel RA, van Dieren A, et al. Two methods for
the assessment of hemodynamic parameters for epidemiology. Acta
Cardiol. 1978;33:84–87.)
Blood Pressure
The catheter-manometer system is, at present, the fundamental
method for continuous accurate measurement of the full arte-
rial pressure waveform. It is, however, an invasive procedure
and should be avoided unless the introduction of a catheter
into an artery is absolutely necessary. The Riva–Rocci–
Korotkoff method (using an upper arm cuff and a stethoscope)
is noninvasive and commonly used, but it does not provide
continuous blood pressure information. However, most auto-
matic methods developed for determining blood pressure have
been based primarily on the Riva–Rocci–Korotkoff method.
For instance, Roy and Weiss (6) described a technique for pro-
viding intermittent determinations of the systolic and diastolic
blood pressure obtained over several heartbeats. Such systems
have also become commercially available. The method indi-
cated above, however, is sensitive to motion artifacts and not
continuous.
Chapter 40 ■ Polygraphy 811
Penaz (7) developed an important improvement in the non-
invasive determination of blood pressure, using continuous
measurement of the blood pressure in the finger. This method
uses a finger cuff. By means of a servosystem the cuff pressure is
maintained equal to the arterial pressure. This is achieved by
minimizing arterial diameter changes using a photoelectric
plethysmographic feedback method. The working principle is
shown in the block diagram of Figure 40.4. The further develop-
ment and evaluation of methods based on this approach (8,9)
are of great interest to those interested in the continuous meas-
urement and recording of beat-to-beat diastolic, systolic, and
mean arterial pressure. This is indicated in Figure 40.5 by the
similarity between continuous blood pressure and curves
recorded simultaneously by way of noninvasive and invasive
methods.
The noninvasive blood pressure measurement was first
applied during anesthesia (10) and was also used for long-term
sleep monitoring in patients with systemic hypertension and
sleep-related breathing disorders (11). The new system gives
valuable results if the position of the finger cuff is carefully con-
trolled. An example of blood pressure recordings combined
with respiratory measurements is displayed in Figure 40.6. A
review of noninvasive continuous blood pressure measurements
is found in Ruddel and Curio (12) and more recently in Parati
et al. (13).
The capacity for measuring acute and immediate changes in
autonomic, EEG, and hemodynamic physiologic variables dur-
ing different sleep stages on a continuous basis has played an
important role in enabling us to understand the interplay
between changes in EEG and changes in circulatory variables
and in autonomic neural functions. In this way the possibility
of recording simultaneously with the EEG, HR variability, and
blood pressure (BP), among other cardiovascular physiologic
variables, has advanced our understanding of mechanisms link-
ing sleep and cardiovascular physiology (14).
Figure 40.5 Example of a comparison of
continuous recordings of blood pressure,
simultaneously obtained by an invasive
(intraarterial) and a noninvasive method. The
latter was performed according to the method
introduced by Penaz.
812 Part V ■ Complementary and Special Techniques

Figure 40.6 Recording of a patient with obstructive apneas and systemic hypertension. The trace of nasal airflow (NAF) shows
complete cessation of respiratory flow, whereas rib cage (RC) and abdominal (Abd) movement show obstructive efforts. Noninvasive
(FINA-PRES) and invasive BP were recorded in parallel. SaO2, oxygen saturation; micro, snoring noise. (From Penzel T, Ducke E,
Peter JJ, et al. Noninvasive monitoring of blood pressure in a sleep laboratory. In: Ruddel H, Curio I, eds. Non-invasive Continuous
Blood Pressure Measurement. Frankfurt am Maim: Peter Lang; 1991.)

Respiration obtained on frequency and on depth of chest movements. Other

The polygraphic recording of respiration patterns is usually car-
ried out to obtain information on frequency or changes in
inhalation depth. Changes in thoracic volume due to respira-
tory movements are usually estimated in terms of changes in
the perimeter of the chest; this can be measured easily using
transducers working on the principle of a strain gauge. This
type of respiration transducer, which is commercially available,
consists of a tube with an inner diameter of a few millimeters;
the tube is elastic and is filled with an electrically conductive
liquid substance with measurable electrical resistance. The tube
forms part of a belt strapped around the chest; changes in
resistance resulting from changes in chest perimeter are meas-
ured by means of the Wheatstone bridge circuit. The rather
smooth slow respiration rhythm can be recorded with most
EEGs without further measures; in case of insufficient fre-
quency response of the apparatus, an alternating current (AC)
voltage can be applied to the measuring bridge instead of a DC
voltage. If respiration must be recorded from moving subjects,
chest size changes not related to respiration can easily occur; in
these cases other methods should be chosen.
A simple and easily implemented solution is to use a thermis-
tor or equivalent temperature-sensitive device placed in the
mouth and/or nostrils. Such a device works as a respiration trans-
ducer by signaling the characteristic differences in temperature of
the inspired and expired air. With such methods, information is
methods should be used to obtain exact information concerning
respiratory volume.
For information concerning respiratory volume and thoracic
and abdominal diameter, variations can be measured by using
elastic strips provided with strain gauges that encircle the thorax
and abdomen at the level of the nipple and umbilicus, respec-
tively. Using both measurements, the respiratory volume can be
calculated. This method of spirometry determines the contribu-
tions of the chest and abdomen separately, then adds them
together to mimic the total spirometric volume. As the chest and
abdominal volumes change during breathing, changes in electri-
cal impedance of the bands are related to changes in the spiro-
metric volume contributions using a calibration and gain
adjustment procedure. A combination of techniques is com-
monly used, namely the measurement of oral and nasal airflow
and, in parallel, of abdominal and thoracic movements
(Fig. 40.4). The best method to analyze the respiratory effort is
the use of an esophageal pressure probe (15).
Pulse oximetry is a noninvasive method to measure the arte-
rial hemoglobin oxygen saturation (SaO2). Continuous pulse
and oxygen saturation measurements are obtained by ear,
finger, or soft probes (Fig. 40.4 contains an example). The
widely used pulse oximeter quantifies the SaO2 as a percentage
based on spectrophometric and photoelectric plethysmography
(16). The SaO2 measurement yields information about the

effectiveness of respiration and is recommended for all types of
sleep monitoring (17).
Electrodermal Activities
The variations of skin electrical properties in relation to psycho-
logical variables, commonly known as the galvanic skin response
(GSR) or psychogalvanic response (PGR), consist of changes of
the electrical conductivity of the skin or of the electrical skin
potential, which can be measured by means of electrodes placed
on the palms or the soles in reference to an electrode placed
elsewhere, such as on the back of the hand.
For such measurements, the skin should be intact; when plac-
ing the electrodes, the skin must not be abraded. The skin poten-
tial can be recorded easily, but it is unstable and of little use;
therefore, skin electrical conductance is much more useful in
this respect. It can be estimated by applying a constant voltage
across two electrodes and by measuring the resulting electrical
current. This is, of course, equivalent to estimating the skin elec-
trical resistance by applying a constant current through the elec-
trodes and measuring the voltage across the electrodes. Skin
resistance can vary considerably among subjects, assuming val-
ues from kilo- to megaohms. Transient skin resistance responses
related to sudden changes in psychological state are in the order
of 100 ohms. In practice, it is preferable to measure skin electri-
cal conductance by applying a constant voltage, instead of meas-
uring resistance by applying a constant electrical current. The
circuit of Figure 40.7A illustrates a constant voltage measuring
procedure. An example of electrodermal responses recorded
during the performance of a contingent negative variation
(CNV) paradigm from a normal subject is shown in Figure 40.8.
Figure 40.7 A: Schematic diagram of a basic electronic circuit for the
recording of electrodermal conductance using a constant voltage
source. B: Bridge circuit for the recording of relatively small changes in
electrodermal conductance.
Chapter 40 ■ Polygraphy 813
Classic extensive basic and practical information about electro-
dermal response recording can be found in the publications by
Venables and Martin (18) and Montagu (19) on skin resistance
potential. Interesting applications of the GSR in psychophysio-
logic studies have been published by Deschaumes-Molinaro et
al. (20) and Vernet-Maury et al. (21).
Eye Movements
In various behavioral studies, particularly in sleep research for
the recognition of sleep stages, eye movement recording (i.e., by
means of the electro-oculogram, EOG) is necessary. Eye move-
ment recording is also useful in EEG recording for identifying
eye movement artifacts and studying lambda waves. An example
of the former is the recording of eye movement in relation to
that of the CNV; in this case, it is possible either to reject the EEG
epochs, where the amount of eye movement exceeds a certain
predetermined threshold in order to carry out selective averag-
ing (22) or to average the EOG along with the CNV to obtain an
indication of the reliability of the latter (see also Chapter 36).
Preferably, the method chosen for identifying eye movement
artifacts will make use of the electrical field generated by the
eyes. The EOG is, in general, easy to measure. The usual princi-
ple of EOG measurement is demonstrated by the model in
Figure 40.8 Polygraphic recording obtained during a contingent nega-
tive variation (CNV) investigation carried out in a normal subject. EEG
activities were recorded from three direct-coupled CNV derivations (FP2,
C4, and C3 against linked mastoids) and six anteroposterior symmetrical
derivations. Other variables recorded were electrodermogram (EDG),
electrocardiogram (ECG), instantaneous heart rate, electro-oculogram
horizontal eye movements [EOG (HOR)], CNV stimulus presentation
[STIM (CNV)], the button press, and a time code (1-second intervals).
814 Part V ■ Complementary and Special Techniques
Figure 40.9 Positions of EOG electrodes for recording horizontal and
vertical eye movements.
Figure 40.9. As a result of the corneoretinal standing potential
(the cornea is positive relative to the fundus), a DC potential dif-
ference can be measured either between the pair of electrodes
(EH) placed in a horizontal plane near the canthi of the eyes or
between the two electrodes (EV) placed in a vertical plane,
depending on the position of the eyeballs. Any change in eyeball
position results in a corresponding change of these two potential
differences. DC recording is necessary to measure exact eye posi-
tions, whereas AC recording suffices for determining changes in
eye position.
To carry out a DC recording of the EOG, nonpolarizable
electrodes must be used, and the drift of the electrodes’ offset
potential should be taken into account. To obtain sufficiently
high recording sensitivity, the electrodes must be placed as close
to the eye as possible. To record horizontal movements, the
electrodes should be placed near the external canthus of each
eye; for vertical movements, the electrodes should be placed
closely above and below one or both eyes. The EOG measured
in this way has an amplitude of about 20 V per degree of eye-
ball rotation. A frequency response up to about 30 Hz is ade-
quate to record the most rapid eye movements, and thus an
EEG recording channel can be used without further provision
(see, e.g., trace EOG (HOR) in Fig. 40.8). Other principles for
monitoring eye movement in clinical EEG are the measurement
of changes in impedance related to eye movements (23) and use
of a pressure transducer system (24). In the latter system, a thin
membrane covering one end of a tube held in a spectacle frame
is adjusted to touch the closed eyelid lightly. A pressure trans-
ducer connected at the other end of the tube detects pressure
changes from eye movements. Robinson (25) describes eye
movement recording systems that also provide information on
exact eye position; eye position is determined from the voltage
generated by an alternating magnetic field in a coil embedded
in a scleral contact lens worn by the subject. A similar photo-
electric device (26) consists of four small infrared-detecting
cells mounted on a light spectacle frame together with a minia-
ture infrared (9000Å)-emitting diode. This system, which pre-
serves maximum vision field size, has a resolution less than 1
minute of arc, a bandwidth of 1000 Hz, and 5% linearity of the
maximum range. Currently, systems are available for recording
and analysis of a wide range of eye movements, both saccadic
and smooth pursuit movements, where these are measured
using a scleral reflection technique (IRIS instrument) (27).
Saccades are rapid eye movements that move the line of sight
between successive points of fixation; they are among the best
understood of movements, possessing dynamic properties that
are easily measured (28). Saccades have become a popular
means to study motor control and cognitive functions, particu-
larly in conjunction with other techniques such as EEG, evoked
potentials, functional imaging, and transcranial magnetic
stimulation.
MUSCLE ACTIVITY AND BODY MOVEMENTS
EMG Activity
Part of the electrical activity of muscles (EMG) can be recorded
easily, by means of either surface electrodes placed on the sur-
rounding skin or needle electrodes inserted into muscle. The
frequency range of EMG potentials, particularly those recorded
by means of intramuscular needles, goes far beyond the fre-
quency response of most EEG recording systems. Even an EEG
apparatus having an ink-jet writing system with a frequency
response of up to 1000 Hz does not provide a faithful represen-
tation of EMG potentials. However, using a filter giving the
highest possible EEG frequency response (usually 70 to 100 Hz),
the recording of EMG activity in most instances adequately indi-
cates the presence of muscular activity and may even provide a
rough quantitative measure of the amount of such activity. In
certain applications, the latter can be better obtained by passing
the EMG potentials through a two-way rectifier, the output of
which is integrated with a specified time window. In this way, a
rectified smoothed EMG record is obtained. It is important to
recognize that the recording of EMG along with EEG signals
may be necessary for the interpretation of some specific features
of the latter, particularly in those applications such as EEG-
based brain–computer interfaces that rely on automated meas-
urements of EEG features (29).
In behavioral studies, EMG activity is often used to monitor
the onset of certain motor activities such as limb movements.
Another way of determining limb movements and tremor (30)
makes use of displacement transducers or accelerometers. A
special form of tremor is the microtremor (microvibration),
which is a phenomenon of the sensorimotor system that is
modulated by central mechanisms (31). This microtremor is in
the frequency range of 8 to 12 Hz (amplitudes of 1 to 10 m)
and can be recorded by an accelerometer fixed, for example, at
the wrist. The measurement range of such an accelerometer
should be between 5 and 5g., such that the microtremor in
the range of 10 mV/g can be recorded. Using accelerometry and
EMG recordings of the forearm, the characteristics of physio-
logical tremor have been studied (32). In addition to the
microtremor, ballistic movements also be measured with this
technique.
Body Movements
The detection of limb or whole-body movements can be of
interest in sleep studies. With a simple wrist actigraph, a dis-
crimination between sleep and wakefulness is possible (33).
Actigraphy has been found to be reliable for evaluating sleep
patterns in patients with insomnia, for studying the effect of

treatments designed to improve sleep, and in the diagnosis of
circadian rhythm disorders, including shift work (34). Such a
wrist actigraph can be used not only in the sleep laboratory but
also with outpatients to obtain a picture of the degree of sleep
disturbances within, for example, a 24-hour period (17). The
recording of movements along with the EEG can also be impor-
tant in differentiating authentic EEG activity from movement
artifacts (35), which, on the basis of their waveforms and ampli-
tude, cannot easily be identified as noncerebral. When monitor-
ing movements of epileptic patients to determine the
occurrence of seizures during the night, transducers indicating
global body movements may be used. With this objective, con-
ventional pressure or displacement transducers or other special
methods developed for the detection of whole-body (36) or
limb movements (37) have been used.
Another method used for recording body movements is the
static charge-sensitive bed, which consists of a mattress with
two electrically active layers (38) with the use of filters; the bal-
listocardiogram, respiratory signals, and movement signals can
be differentiated.
For movement quantification in epileptic seizures advanced
video analysis methods can be applied (39). Markers at land-
mark points are attached to the patient. Then EEG is acquired
and the movement of the body parts is monitored using special
cameras. Quantified motion trajectories of body parts can be
extracted based on the fiducial markers. The trajectories reflect
the motion pattern of patients during seizures yielding addi-
tional movement information that cannot be obtained from
standard video-EEG analysis.
Temperature
Temperature can be measured with sensors placed on the skin
or with a rectal probe. Temperature measurements are espe-
cially important in sleep studies. Detailed studies of rectal tem-
perature recordings over 24 hours are reported by Stephan and
Dorow (40). Experiments with long-term isolation of subjects
have revealed a relationship between the body temperature and
the duration and stage of sleep (41). To assess core temperature,
for instance during monitoring under anesthesia, the recording
of tympanic temperature may be carried out.
REFERENCES
1. Venables PH, Martin I, eds. A Manual of Psychophysiological
Methods. Amsterdam: North Holland; 1967.
2. Melrose DG, Lynn RB, Rainbow RLG, et al. A sensitive digital
plethysmograph. Lancet. 1954;1:810–812.
3. Lader MH. Pneumatic plethysmography. In: Venables PH, Martin
I, eds. A Manual of Psychophysiological Methods. Amsterdam:
North Holland; 1967:159–183.
4. Weiman J. Photoplethysmography. In: Venables PH, Martin I, eds.
A Manual of Psychophysiological Methods. Amsterdam: North
Holland; 1967:185–217.
5. Gratze G, Fortin J, Holler A, et al. A software package for noninva-
sive, real-time beat-to-beat monitoring of stroke volume, blood
pressure, total peripheral resistance and for assessment of auto-
nomic function. Comp Biol Med. 1998;28:121–142.
Chapter 40 ■ Polygraphy 815
6. Roy R, Weiss M. Automatic blood pressure indicator. IRE Trans
Biomed Electron 1962;9:244–246.
7. Penaz J. Photoelectric measurement of blood pressure volume and
flow in the finger. In: Alben R, Vogt W, Helbig W, eds. Digest of the
10th International Conference on Medicine and Biological
Engineering. Dresden; 1973:104.
8. Reeben V, Epler M. Detection and many-side use of signals from
controlled counter pressure finger cuffs. Proc Biocapt Paris.
1975;1:265–270.
9. Wesseling KH, van Bemmel RA, van Dieren A, et al. Two methods
for the assessment of hemodynamic parameters for epidemiology.
Acta Cardiol. 1978;33:84–87.
10. Wesseling KH, Settels JJ, de Wit B. The measurement of continu-
ous finger arterial pressure noninvasively in stationary subjects. In:
Schmidt TH, Dembroski TM, Blumchen G, eds. Biological and
Psychological Factors in Cardiovascular Disease. Berlin: Springer-
Verlag; 1986:355–375.
11. Penzel T, Ducke E, Peter JJ, et al. Non-invasive monitoring of blood
pressure in a sleep laboratory. In: Ruddel H, Curio I, eds. Non-
invasive Continuous Blood Pressure Measurement. Frankfurt am
Main: Peter Lang; 1991.
12. Ruddel H, Curio I. Non-invasive Continuous Blood Pressure
Measurement. Frankfurt am Main: Peter Lang; 1991.
13. Parati G, Ongaro G, Bilo G, et al. Non-invasive beat-to-beat blood
pressure monitoring: new developments. Blood Press Monit.
2003;8(1):31–36.
14. Murali NS, Svatikova A, Somers VK. Cardiovascular physiology
and sleep. Front Biosci. 2003;8:s636–s652.
15. Roberts S, Davies WL. Comparison of simple (slope based) and
more complex (simple syntactic) algorithms for central apnoea
detection. In: Chase MH, Lydic R, O’Connor C, eds. Sleep Research.
Los Angeles: Brain Information Service/Brain Research Inst;
1989:398.
16. West P, George CF, Kryger MH. Dynamic in vivo response charac-
teristics of three oximeters: Hewlett-Packard 47201A, Biox III, and
Nellcor N-100. Sleep. 1987;10:263–271.
17. Penzel T, Stephan K, Kubicki S, et al. Integrated sleep analysis, with
emphasis on automatic methods. In: Degen R, Roding EA, eds.
Epilepsy, Sleep and Sleep Deprivation. 2nd ed. (Epilepsy Res. Suppl.
2). Amsterdam: Elsevier; 1991:177–204.
18. Venables PH, Martin I. Skin resistance and skin potential. In:
Venables PH, Manin I, eds. A Manual of Psychophysiological
Methods. Amsterdam: North Holland; 1967:53–102.
19. Montagu JD. The psycho-galvanic reflex: a comparison of d.c. and
a.c. methods of measurement. J Psychosom Res. 1964;8:49–65.
20. Deschaumes-Molinaro C, Dittmar A, Vernet-Maury E. Autonomic
nervous system response patterns correlate with mental imagery.
Physiol Behav. 1992;51(5):1021–1027.
21. Vernet-Maury E, Alaoui-Ismaili O, Dittmar A, et al. Basic emo-
tions induced by odorants: a new approach based on autonomic
pattern results. J Auton Nerv Syst. 1999;75(2–3):176–183.
22. Papakostopoulos D, Winter A, Newton P. New techniques for the
control of eye potential artifacts in multichannel CNV recordings.
Electroencephalogr Clin Neurophysiol. 1973;34:651–653.
23. Sullivan GH, Weltman G. Impedance oculograph—a new tech-
nique. J Appl Physiol. 1963;18:215.
24. Winter AL, Kellenyi L. Eye movement monitoring. Proc
Electrophysiol Technol Assoc. 1971;18:121.
25. Robinson DA. A method of measuring eye movement using a scle-
ral search coil in a magnetic field. IEEE Trans Biomed Eng BME.
1963;10:137–145.
816 Part V ■ Complementary and Special Techniques
26. Gauthier GM, Volle M. Two-dimensional eye movement monitor
for clinical and laboratory recordings. Electroencephalogr Clin
Neurophysiol. 1975;39:285–291.
27. Muir SR, MacAskill MR, Herron D, et al. EMMA—an eye move-
ment measurement and analysis system. Australas Phys Eng Sci
Med. 2003;26(1):18–24.
28. Leigh RJ, Kennard C. Using saccades as a research tool in the clin-
ical neurosciences. Brain. 2003;127(3):460–477.
29. Goncharova II, McFarland DJ, Vaughan TM, et al. EMG contami-
nation of EEG: spectral and topographical characteristics. Clin
Neurophysiol. 2003;114(9):1580–1593.
30. Oppel F, Umbach WW. A quantitative measurement of tremor.
Electroencephalogr Clin Neurophysiol. 1977;43:885–888.
31. Burne JA, Lippold OC, Pryor M. Proprioceptors and normal
tremor. J Physiol. 1984;348:559–572.
32. Elble RJ. Characteristics of physiologic tremor in young and eld-
erly adults. Clin Neurophysiol. 2003;114(4):624–635.
33. Sadeh AJ, Alster J, Urbach D, et al. Actigraphically based automatic
bedtime sleep-wake scoring: validity and clinical applications. J
Ambul Monit.1989;2:209–216.
34. Ancoli-Israel S, Cole R, Alessi C, et al. The role of actigraphy in the
study of sleep and circadian rhythms. Sleep. 2003;26(3):342–392.
35. Buchthal F, Dahl E, Projaborg W. Simultaneous recording of accel-
eration and brain waves. Electroencephalogr. Clin Neurophysiol.
1973;34:550–552.
36. van Nimwegen C, Boter J, van Eijnsbergen B. A method of detect-
ing epileptic seizures. Epilepsia. 1975;16:689–692.
37. Kripke DR, Mullaney DJ, Messin S, et al. Wrist actigraphic meas-
ures of sleep and rhythms. Electroencephalogr Clin Neurophysiol.
1978;44:674–676.
38. Alihanka J, Vaahtoranta K, Saarikivi J. A new method of long-term
monitoring of the ballistocardiogram, heart rate, and respiration.
Am J Physiol. 1981;240:348–392.
39. Li Z, da Silva AM, Cunha JPS. Movement quantification in epilep-
tic seizures: a new approach to video-EEG analysis. IEEE Trans
Biomed Eng. 2002;49:565–573.
40. Stephan K, Dorow R. Circadian variations of core body tempera-
ture, performance and subjective ratings of fatigue in “morning”
and “evening” types. In: Redfern PA, Campall IC, Davies JA, eds.
Circadian Rhythms in the Central Nervous System. Weinheim:
Verlagsgesellschaft, 1985:233–236.
41. Zulley J, Wever RA, Aschoff J. The dependence of onset and dura-
tion of sleep on the circadian rhythm of rectum temperature.
Pflugers Arch. 1981;391:314–318.

Polysomnography:
Technical and Clinical Aspects
SUDHANSU CHOKROVERTY, RODNEY RADTKE, AND JANET MULLINGTON
P
olysomnography (PSG) is the single most important lab-
oratory technique for assessment of sleep and its disor-
ders. PSG consists of recordings of multiple physiologic
characteristics during sleep, whereas polygraphy indicates the
recording of similar characteristics at any time during the day.
An understanding of the importance of the laboratory evalua-
tion of sleep and its disorders has been evolving slowly, but in
the last century great advances have been made in this direc-
tion. The discoveries of the human electroencephalogram (EEG)
by Berger (1) in 1929 and rapid eye movements (REMs) during
sleep by Aserinsky and Kleitman (2) in 1953 are the real driving
forces behind this understanding. In 1974, Holland and his col-
leagues (3) during a presentation at the 14th annual meeting of
the Association of the Psychophysiological Study of Sleep (later
named American Sleep Disorders Association, which was
finally changed to the American Academy of Sleep Medicine
[AASM]) coined the term PSG. This chapter is divided into sev-
eral parts: The initial part includes a brief review of the histori-
cal milestones, functional neuroanatomy of sleep, physiologic
changes emphasizing those pertinent to overnight PSG inter-
pretation and clinical relevance as well as homeostatic and
circadian factors, and functions of sleep. Next we deal with
laboratory procedures including PSG recording and scoring
techniques, indications for PSG, video PSG, ambulatory and
computerized PSG, artifacts during PSG recording, and pitfalls
of PSG. We then discuss clinical considerations, briefly describ-
ing clinical presentation, diagnosis, and treatment but mainly
focusing on PSG findings in common sleep disorders as well as
sleep-related movement disorders and neurologic disorders,
and sleep-related epilepsies. The final part addresses related
laboratory procedures for assessment of sleep including multi-
ple sleep latency test (MSLT), maintenance of wakefulness test
(MWT), and actigraphy.
HISTORICAL MILESTONES
In 1875, Richard Caton (4), an English neurophysiologist, first
recorded electrical activities from rabbit and monkey brains.
It was, however, Hans Berger (1), an Austrian psychiatrist, who
in 1924 first obtained electrical activities from the scalp surface
of a 17-year-old young man with a skull defect and published his
findings in 1929. Although sleep was known since antiquity,
brain activities in sleep could not have been recorded before the
discovery of EEG. It was in 1937 that Loomis and his colleagues
classified sleep into five stages (A–E) based on the EEG activities
CHAPTER
41
during the deepening stages of sleep (5). It is interesting to note
that Kohlschutter (6,7), a 19th century German physiologist,
thought that sleep was deepest in the first few hours and became
lighter as time went on, and he also described the varying
arousal thresholds throughout the night. The REMs of sleep
were not known in those days. In 1953, Aserinsky and Kleitman
(2) obtained characteristic REMs during sleep using surface
electrodes over the eye lids. Over the 4-year period, Loomis and
his colleagues (5) published 12 papers, describing the oscillating
nature of EEG along with five distinct stages of sleep as well as
arousals manifested by increased body movements and respira-
tory effort. In 1957, Dement and Kleitman (8) described sleep
evolving through non-rapid eye movement (NREM) and REM
sleep states in a cyclic manner throughout the night. All the
components of REM sleep, however, have not been described by
them until Jouvet and Michel (9) in 1959 observed markedly
decreased muscle tone during REM sleep in cats. In 1961, Berger
(10) recorded markedly decreased muscle tone from extrinsic
laryngeal muscles during human REM sleep. In order to stan-
dardize the scoring of different stages of sleep, an ad hoc com-
mittee led by Rechtschaffen and Kales (R–K) (11) in 1968
produced the now-famous sleep scoring technical manual (The
R–K Scoring Technique). This remained the “gold standard” until
the AASM published the AASM Manual for the Scoring of Sleep
and Associated Events (12), which modified the R and K tech-
nique and extended the scoring rules. The R and K sleep scoring
manual (11) was devised only for normal sleeping adults, but
later infant sleep scoring manual was developed. R and K sleep
scoring technique is based on three physiologic characteristics:
EEG, electro-oculography (EOG), and electromyography
(EMG). PSG, however, includes more than just sleep staging,
and other physiologic characteristics such as respiration, limb
muscle activity, blood oxygen saturation, electrocardiogram
(ECG), body position, snoring, and other special recordings
have been incorporated, which are described later in this chap-
ter. Since the discovery of REM sleep, dream was thought to be
associated with REM sleep during 80% of the time. In 1868
Griesinger (13) suggested that dreaming was associated with eye
movements and in 1895, Freud (14) indicated that dreaming was
associated with relaxation of the major muscles of the body.
Amazingly, centuries ago (ca. 1000 BC), Upanishads (15), the
ancient Indian text of Hindu religion, sought to divide human
existence into four states: the waking, the dreaming, the deep
dreamless sleep, and the superconscious (“the very self”). This is
reminiscent of modern classification of three states of existence.
817
818 Part V ■ Complementary and Special Techniques
FUNCTIONAL NEUROANATOMY OF SLEEP
The control of sleep is quite complex. Whereas for the most part
during the whole sleep period we are in an unconscious state,
changes occur that are physiologically quite distinct. During slow
wave sleep, the cortical neurons fire in relative synchrony, pro-
ducing what is described as slow wave sleep, so named after
the polygraphic pattern of this brain activity. During REM sleep,
the eyes show phasic, rapid conjugate movements, whereas at the
same time skeletal muscle tone is at its lowest levels. Brain activ-
ity during REM sleep is rapid and desynchronous, almost like
waking, which is why it is also referred to as “paradoxical sleep.”
ASCENDING AROUSAL SYSTEM
There are two major branches to the ascending arousal or retic-
ular activating system (ARAS) (16). One is an ascending path-
way from pedunculopontine (PPT) and laterodorsal tegmental
(LDT) nuclei to the reticular nucleus and thalamic relay nuclei
in the thalamus. These thalamic relay neurons transmit signals
to the cortex. Neurons in the PPT and LDT fire most rapidly
during REM and wakefulness, and are at their most inactive
state during NREM.
The second branch of the ARAS comes from locus coeruleus
(LC), dorsal and median raphe nuclei, periaqueductal gray mat-
ter, and tuberomammillary neurons. Cortical input goes through
lateral hypothalamic neurons and basal forebrain neurons. The
neuron groups in the two branches of the ARAS are shown in
Figure 41.1 along with their respective neurotransmitters (17).
The Hypocretin (Orexin) peptidergic system (18) located in
the lateral hypothalamic and perifornical regions with its wide-
spread ascending and descending projections is thought to play
an important role in the control of arousal and wakefulness. A
reduction of activities of the hypocretin projections to the LC,
midline raphe, mesopontine, and posterior hypothalamic and
tuberomedullary regions will cause sleepiness.
REM-Generating Neurons
The existence of REM sleep-generating neurons in the pons has
been proved by transection experiments in cats through differ-
ent regions of the midbrain, pons, and medulla (19) (Fig. 41.2).
A transection of the pontomesencephalic (A) and pon-
tomedullary (B) junctions produced an isolated pons that shows
all the signs of REM sleep (see Fig. 41.2). There are mainly two
animal models available to explain the mechanism of REM
sleep. McCarley–Hobson (20) reciprocal interaction model
based on reciprocal interactions of REM-on and REM-off neu-
rons is the earliest and most well-known model. A reciprocal
interaction in the brainstem between REM-on neurons (the
cholinergic pedunculopontine tegmental and LDT nuclei in the
pontomesencephalic region) and REM-off neurons (the amin-
ergic LC and dorsal raphe [DR] nuclei) is responsible for REM
generation and maintenance. The role of gamma-aminobutyric
acid (GABA) in the REM sleep generation has been emphasized
in the latest modification of the reciprocal interaction model by
McCarley (20). Lu and co-workers (21) described a “flip-flop”
switch interaction model in rats between GABAergic REM-off
Figure 41.1 The ascending arousal system. Two major pathways are
shown. One (shown in light gray) providing upper brainstem input to
the thalamic-relay nuclei and to the reticular nucleus of the thalamus
coming from the pedunculopontine and laterodorsal tegmental
(PPT/ LDT) nuclei, which are acetylcholine (ACh)-producing neuronal
groups. The second major group of neurons (shown in dark gray) come
from the noradrenergic (NA) locus coeruleus (LC), serotoninergic (5-HT)
dorsal and median raphe nuclei, dopaminergic (DA) periaquiductal gray
matter (vPAG), and histaminergic (His) tuberomamillary neurons (TMN).
Additional cortical input merges from the basal forebrain (BF) neurons
containing GABA or ACh, and by lateral hypothalamic (LH) peptidergic
neurons that contain melanin-concentrating hormone (MCH) or orexin
(hypocretin) (ORX). (Reproduced from Saper CB, Scammell TE, Lu J.
Hypothalamic regulation of sleep and circadian rhythms. Nature. 2005;
437:1257–1263, with permission.)
neurons in the deep mesencephalon, ventral periaqueductal
gray, and lateral pontine tegmentum, and GABAergic REM-on
neurons in the sublaterodorsal (SLD) nucleus (equivalent to
perilocus coeruleus alpha in the cat) and the dorsal extension of
the SLD named precoeruleus to explain REM sleep mechanism.
Ascending glutamatergic projections from precoeruleus neurons
to the medial septum are responsible for the hippocampal EEG
theta rhythm during REM sleep. Muscle atonia during REM
sleep is related to descending glutamatergic projections from the
ventral SLD directly to the spinal interneurons inhibiting spinal
ventral horn cells by both glycinergic and GABAergic mecha-
nisms. Cholinergic and aminergic neurons play a modulatory
role in this model and are not part of the “flip-flop” switch.
Brooks and Peever (22) recently challenged the glycinergic and
GABAergic neurochemical mechanisms of REM motor atonia
based on experimental evidence in rats that REM atonia per-
sisted even when glycine and GABA receptors were blocked.
These authors suggested that multiple biochemical pathways are
responsible for controlling muscle tone in REM sleep.
 Chapter 41
Schematic: Sagittal Section of Brain Stem
C B A noted in the respiratory, cardiovascular, gastrointestinal, and
Spinal Cord Medulla Pons Midbrain Forebrain
Figure 41.2 Schematic sagittal section of the brainstem of the cat.
A: Junction of midbrain and pons. B: Junction of pons and medulla.
C: Junction of medulla and spinal cord. (Reproduced from Chokroverty
S. Sleep Disorders Medicine: Basic Science, Technical Considerations and
Clinical Aspects. 3rd ed. Philadelphia: Saunders/ Elsevier; 2009, with
permission.)
Anatomical Substrates for NREM Sleep
NREM sleep-generating neurons are located primarily in the
ventrolateral preoptic area (VLPO) of the hypothalamus, the
basal forebrain area as well as the solitary nucleus region of
the medulla. The reticular nucleus of the thalamus is thought to
be responsible for sleep spindle generation. Both passive and
active mechanisms play a role in the generation of sleep. The con-
temporary theory (23) for the mechanism of NREM sleep sug-
gests a reciprocal interaction between two antagonistic neurons
in the VLPO region of the anterior hypothalamus and wake-pro-
moting neurons in the tuberomammillary nuclei of the posterior
hypothalamus and the hypocretinergic neurons in the lateral
hypothalamus, as well as the LC, DR nuclei, basal forebrain, and
mesopontine tegmentum. The extended VLPO input to LC and
DR is thought to be involved in regulating REM sleep. Reciprocal
interaction between sleep-promoting neurons in the region of
the solitary nucleus and the wake-promoting neurons within the
ARAS of the brainstem independent of the reciprocal interaction
of the neurons of the forebrain also plays a minor role in the gen-
eration of NREM sleep. It has been suggested that adenosine, a
neuromodulator, may act as a physiologic sleep factor modulat-
ing the somnogenic effects of prolonged wakefulness (24).
PHYSIOLOGIC CHANGES DURING SLEEP
A vast number of physiologic changes take place during sleep in
humans, affecting almost every system in the body (Table 41.1)
(25). It is important to have a basic knowledge about these
changes during sleep and how they affect various sleep disorders.
These changes may be documented in the Polysomnography. A
case in point is obstructive sleep apnea syndrome (OSAS) causing
dramatic changes in the respiratory control of the upper airway
muscles during sleep, directing our attention to a very important
■ Polysomnography: Technical and Clinical Aspects 819
pathophysiologic mechanism and the therapeutic intervention for
this disorder. These physiologic changes are most commonly
endocrine systems. Respiration is controlled by the automatic or
metabolic and behavioral systems complemented by the third sys-
tem known as the arousal system or the system for wakefulness
stimulus. These respiratory systems receive inputs from various
peripheral and central neural structures to maintain acid–base
regulation and respiratory homeostasis. The location of the respi-
ratory neurons makes them easily vulnerable to a variety of neu-
rologic disorders, particularly those involving the brainstem.
Some conditions may affect control of breathing only during
sleep, causing undesirable, often catastrophic, results including
cardiorespiratory failure or even sudden death. The two control
systems (metabolic and voluntary) are active during wakefulness,
but during NREM sleep, the voluntary system is inactive and res-
piration is entirely dependent upon the metabolic controller. The
behavioral mechanism is probably responsible for controlling
breathing, at least in part during REM sleep. Tidal volume and
alveolar ventilation decrease during sleep. Arterial oxygen tension
is slightly decreased and arterial carbon dioxide tension is
increased during both NREM and REM sleep. Hypoxic ventila-
tory response is impaired in NREM sleep in adult men but not in
women. Hypoxic ventilatory response during REM sleep is signif-
icantly decreased in both sexes. Hypercapnic ventilatory response
is also decreased during NREM and further decreased during
REM sleep. Thus, ventilation is unstable during sleep and few
periods of apneas may occur particularly at sleep onset and dur-
ing REM sleep in normal individuals. Respiratory homeostasis is
thus relatively unprotected during sleep, making those individuals
with intrinsic respiratory disease, such as chronic obstructive pul-
monary disease (COPD) or bronchial asthma, highly vulnerable
for respiratory failure during sleep.
As a result of increased parasympathetic and decreased sym-
pathetic activity during sleep, heart rate, blood pressure, cardiac
output, and peripheral vascular resistance decrease during
NREM sleep and decrease still further during REM sleep. In
REM sleep, however, there is an intermittent activation of the
sympathetic nervous system accounting for rapid fluctuations
in blood pressure and heart rate. Cardiac output falls progres-
sively during sleep and the greatest decrement occurs during
the last sleep cycle, particularly during the last REM sleep cycle
early in the morning. This may explain why normal individuals
and patients with cardiopulmonary disease are most likely to
die during the early morning hours. Cerebral blood flow and
cerebral metabolic rates for glucose and oxygen decrease during
NREM sleep but increase to that of the waking values during
REM sleep. Because of these hemodynamic and sympathetic
changes during REM sleep during the last third of the sleep
cycle in the early hours of the morning, there could be increased
platelet aggregability, plaque rupture, and coronary arterial
spasms possibly triggering thrombotic events causing myocar-
dial infarction, ventricular arrhythmias, or even sudden cardiac
death. These circadian variations in cardiovascular and cere-
brovascular events with the highest rates of events occurring
during the early morning hours have been documented by
meta-analysis of epidemiologic studies.
820 Part V ■ Complementary and Special Techniques

Tabl e 41 . 1

Physiologic Changes during Wakefulness, NREM Sleep, and REM Sleep

Physiology Wakefulness NREM Sleep REM Sleep

Parasympathetic activity
Sympathetic activity Decreases or variable ( )
Heart rate Normal sinus rhythm Bradycardia Bradytachyarrhythmia
Blood pressure Normal Decreases Variable
Cardiac output Normal Decreases Decreases further
Peripheral vascular resistance Normal Normal or decreases slightly Decreases further
Respiratory rate Normal Decreases Variable; apneas may occur
Alveolar ventilation Normal Decreases Decreases further
Upper airway muscle tone Decreases or absent
Upper airway resistance
Hypoxic and hypercapnic Normal Decreases Decreases further
ventilatory responses
Cerebral blood flowa
Thermoregulation –
Gastric acid secretion Normal Variable Variable
Gastric motility Normal Decreases Decreases
Swallowing Normal Decreases Decreases
Salivary flow Normal Decreases Decreases
Migrating motor complex Normal Slow velocity Slow velocity
(a special type of intestinal
motor activity)
Penile or clitoral tumescence Normal Normal Markedly increased

aThere is, in general, global decrease in cerebral blood flow with regional variation during NREM sleep, but this is not homogeneous. It may not decrease in some
areas and may even show phasic increase in certain areas.
NREM, non-rapid eye movement; REM, rapid eye movement; , mild; , moderate; , marked; , very marked; , absent.
Reproduced from Chokroverty S. Sleep and its disorders. In: Bradley WG, Daroff RB, Fenichel GM, et al., eds. Neurology in Clinical Practice. Philadelphia: Elsevier;
2008:1960, with permission.

There are profound changes in endocrine secretions during
sleep. There is a pulsatile increase of growth hormone during
NREM sleep in the first third of the normal sleep period. Prolactin
secretion also rises 30 to 90 minutes after sleep onset. Cortisol
secretion is inhibited by sleep, whereas thyroid-stimulating
hormone reaches a peak in the evening and then decreases
throughout the night. Testosterone levels in men increase during
sleep, rising from low levels at 8 PM to peak levels at 8 AM, but there
is no clear relationship noted between levels of gonadotrophic
hormones and sleep–wake cycle in children or adults. Melatonin,
the hormone of darkness released by the pineal gland, reaches its
highest secretion level between 3 AM and 5 AM and then decreases
to low levels during the day. Thermoregulation is maintained
during NREM sleep, but it is nonexistent in REM sleep. Body tem-
perature begins to fall at sleep onset and reaches its lowest point
during the third sleep cycle.
Homeostatic and Circadian Factors
and Functions of Sleep
The normal sleep–wake rhythm is controlled by both the circa-
dian and the homeostatic systems. The circadian system is
entrained by the normal light–dark cycle and the hormone con-
trolling the circadian timing of physiologic systems is the pineal
hormone, melatonin. Bright light suppresses melatonin and it is
released during the dark phase. Human circadian rhythm has a
cycle close to 24 hours (24.2). The paired suprachiasmatic
nuclei (SCNs) of the hypothalamus above the optic chiasm
serve as the master biologic clock. SCN receives afferent infor-
mation from the retinohypothalamic tract sending signals to
multiple synaptic pathways including other parts of the hypo-
thalamus and the pineal gland where melatonin is released.
Remarkable progress including identification of many genes
and their protein products within the circadian clock in both
 Chapter 41
fruit flies (Drosophila) and mammals has been made (26).
Dysfunction of the circadian rhythm results in several impor-
tant sleep disorders including delayed sleep phase syndrome
and advanced sleep phase syndrome.
During wakefulness, it is thought that a build up of some
sleep factor, or process “S,” occurs and is burnt off or dissipated
at night through sleep (27). In the morning after waking, cog-
nitive and physiologic functioning takes a while to get to a full
waking state due to an inertia from the sleep system (28). The
wake system takes a while to fully prime and synchronize all
components for optimal output. This mechanistic framework
has explanatory utility and does a reasonable job of providing a
framework for much of the physiologic and cognitive data.
Notably, the concept of homeostasis is exemplified by the find-
ing of recovery sleep following a period of sleep deprivation
that is intensified with respect to slow wave activity. This sug-
gests that physiologic systems need slow wave sleep and there-
fore attempt to catch up for what was lost by intensifying and
lengthening the sleep bout as well (29).
The function of sleep remains the greatest biologic mystery
of all times. Sleep is essential and sleep deprivation leads to
impaired attention and decreased performance in addition to
sleepiness. Sleep is thought to have restorative, conservative,
adaptive, thermoregulatory, and consolidative functions, as well
as maintenance of synaptic and neuronal network integrity.
Recent scientific data have strengthened the theory that mem-
ory reinforcement and consolidation take place during sleep.
SLEEP ARCHITECTURE
The term “sleep architecture” refers to the pattern of sleep
stages that cycle through the night in what is referred to as the
REM–NREM cycle. During development, there are changes
that occur in the general pattern of sleep stages, but the adult
shows a pattern of stages N1, N2, and N3 followed by REM
sleep, and the periodicity of this REM–NREM cycle throughout
the night is generally 90 to 120 minutes in duration. The first
REM period of the night is very often brief and may even be
missed in young healthy sleepers.
During infancy, the sleep–wake cycle is approximately 50%
dominated by REM sleep, and the sleep pattern involves fre-
quent napping through the day. As the infant matures, there is
a consolidation of sleep into the night period. During infancy,
the mother’s melatonin rhythm is conveyed to the infant
through breast milk, helping to consolidate the sleep period
into the nocturnal dark phase. The last sleep to be given up dur-
ing the day period is generally the afternoon nap.
POLYSOMNOGAPHIC TECHNIQUE
In 2007, the AASM (12) published a new manual for the scor-
ing of sleep and associated events, almost 40 years after the first
consensus on the technical standards for the field was published
by Rechtschaffen and Kales in 1968 (11). The AASM has a Web
site (http://www.aasmnet.org/) with links to resources for sleep
medicine clinicians, scientists, technical personnel, and labora-
tory managers, and to several position papers.
■ Polysomnography: Technical and Clinical Aspects 821
TECHNICAL PERSONNEL
The hiring of technologists who are mature and competent and
highly motivated is a critical step in the operations of a sleep
laboratory. A feeling of trust between the patient and technolo-
gist is very important for sleeping in an unusual environment
because electrodes and other devices attached to the body can
put the patient in a somewhat uneasy state. Coupled with the
fact that they are in the laboratory for reasons of disturbed sleep
and/or wakefulness, and that continuous positive pressure may
be tested, even on their first night in the laboratory, it is very
important that they have a trusting professional interaction
with the technologist.
In addition to professionalism, maturity, and excellent bed-
side manner, the technologist needs to be able to function well
at night, and, as much as possible, maintain a consistent shift.
Frequent rotations from night to day lead to stress and result in
unnecessary sleep loss and misalignment of biologic rhythms.
This can in turn lead to impaired performance and mood dys-
regulation.
Technologists chosen for performing diagnostic and treat-
ment initiation sleep recordings frequently have full EEG techni-
cian training and/or training as a respiratory therapist. In
addition, they require well-supervised training in specialized
polygraphic recording methods and in visual sleep staging,
including identification and classification of arousals, leg move-
ments, and respiratory disturbance. Technologists are responsible
for the maintenance of equipment and for the identification and
fixing of technical problems at night during a recording if they
should arise. Although most recordings are now performed with
commercially available digital systems, the technologists still
need to quickly and correctly identify problems and swap out
parts as necessary. Therefore, they need to be able to make sure
that adequate supplies and back up equipment are available. The
“American Association of Sleep Technologists” has developed
standards of training, practice, and certification for registration
of technologists in North America (http://www.aastweb.org/).
SETTING: EQUIPMENT AND
RECORDING ROOM
A report for the technologist including the patient’s presenting
symptoms and reason for the referral needs to be provided. The
technologist should also have access to list of medications and
other information pertinent to the care of the patient through
the night in the laboratory. It is most common for patients to be
studied outside of a hospital setting, but even within the hospi-
tal setting, it is important for the technologist to have explicit
instructions for when and whom to call with respect to critical
levels of oximetry, abnormal ECG, or other medical needs of
the patient. As in any hospital ward setting, it is not desirable for
a technologist to work alone, without any back up personnel
immediately available, for safety reasons.
The equipment and technologist are usually housed in a room
separate from that of the sleeping subject. Rather than an institu-
tional-like bedroom, it is preferable for the patient to have an
environment that is as home-like as possible, with a comfortable
822 Part V ■ Complementary and Special Techniques
mattress, a closet, and night stand. The sleeping room should be
quiet and sound attenuated so that noises from outside the sleep
laboratory and technologist’s area do not disturb the sleeping
patient. An intercommunication system for contact with the
patient is essential. Toilet facilities should preferably be attached
to the bedroom and a shower should also be readily available.
Patients are generally asked to report to the laboratory facil-
ity 1 to 3 hours before their normal bedtimes. They are shown
their room and the facility around to let them know they are
appropriate. The technician asks them to change into their bed
clothes and then they begin to attach the electrodes and record-
ing devices. The technologist turns lights out after he/she con-
ducts a test of the equipment and performs biologic calibrations.
Biologic calibrations enable the technologist to demonstrate to
the individual, who subsequently will review the recording, the
normal waking eye movement pattern, and pattern of gritting
teeth, breathing deeply, coughing, and pointing and flexing the
toes and feet. All of these need to be documented by the technol-
ogist at the outset of the recording, and when lights are turned
off for the night, the technologist must document the time, in
the recording.
THE ELECTRODE SETUP
EEG
Sleep EEG has always been based on the standard “international
10 to 20 EEG system” derivations originally defined by Jasper in
1958 (30). According to the most recent AASM scoring guide-
lines (12), the recommended EEG derivations for sleep include
scalp EEG derivations (from the frontal, central, and occipital
regions) linked to the left mastoid (M1) or right mastoid (M2).
Recommendations for sleep stage discrimination are F4–M1,
C4–M1, and O2–M1 with back up electrodes placed at F3, C3,
O1, and M2 to allow for re-referencing to F3–M2, C3–M2, and
O1–M2 if there are problems that develop over the course of
the recording with the right hemisphere selections. An alterna-
tive montage for stage scoring is Fz–Cz, Cz–Oz, and C4–M1 with
back up electrodes Fpz, C3, O1, and M2. The EEG is the most
important physiologic information for the determination of
sleep stages. Of course, for studies where epileptic discharge
during sleep is under investigation, additional electrodes would
be included depending on the brain regions of interest.
EOG
For routine EOG, the AASM scoring manual (12) recommends
referential recordings from each outer canthus (1 cm below the
left and 1 cm above the right) to the ipsilateral mastoid. These
derivations show horizontal out-of-phase potentials very well,
which are used in the identification of REMs associated with
REM sleep. EOG electrodes should be attached with soft tape
made for use on skin, and they should be never attached with col-
lodion-soaked gauze because of the potential for corneal damage
and also because collodion becomes inflexible when dry, so not
only itchy and uncomfortable but prone to lifting and coming off
during sleep. A micropore surgical tape maintains good contact
for more than 48 hours of continuous recording and can be eas-
ily trimmed to needed dimensions. Tape width is usually twice
the diameter of the electrode and about 5 to 8 cm along the wire.
A length of tape added at right angles over the end of the first
tape helps anchor the latter. A high-frequency filter of 35 Hz is
used to reduce EMG artifact on the tracing. The low-frequency
filter is generally selected at 0.3 to 0.5 second (time constant) and
permits recording of both the slow rolling eye movements of
drowsiness and the REMs of REM sleep.
EMG
Mylohyoid muscle EMG is recorded with electrodes placed on
the skin, secured with flexible sticky tape. It is recommended
that three electrodes be attached, one electrode positioned mid-
line just above the inferior rim of the mandible and two other
electrodes each 2 cm below the inferior edge of the mandible, to
the right and left. The technologist can in this way select the
bipolar combination of electrodes giving the best tonic EMG.
Pairs of electrodes have also been used on the jaw edge or over
the masseter muscles to record axial muscle tone (31,32). There
is a great deal of variability in the resting chin tone EMG level
that is related to adipose tissue and age, and therefore it is
important that the technologist adjusts the EMG amplification
such that the background level during drowsiness shows some
baseline elevation so that transition into sleep and particularly
REM is easily discernable.
In addition to submental EMG, muscle activity is routinely
recorded from the legs to document leg muscle activation.
These extra long electrodes are placed on the anterior tibialis
muscles, again fixed with soft flexible skin tape such as microp-
ore, and need to be well fixed with tape so that they don’t come
off with normal leg movements during sleep.
ECG
A single lead that is modified from lead II used in clinical cardi-
ology is used to monitor the heartbeat during sleep, with an elec-
trode placed on the right subclavicle area and left torso over the
lower ribcage beneath the heart. Standard EEG electrodes may be
used, but ECG electrode applications are preferable in that they
are more resistant to artifact. Heart rate, extrasystoles, and other
arrhythmias can be monitored using this lead, but due to differ-
ences in sampling rate and amplifier characteristics, this tracing
is not suitable for detecting PQRST complex abnormalities.
RESPIRATORY MONITORING
The measurement of respiration during sleep is critical to the
diagnosis of sleep disorders and includes chest and abdominal
excursion, most often using inductance plethysmography or
sometimes strain gauges, and also a means of assessing nasal air-
flow. Syndrome definitions and measurement techniques are
described for sleep-disordered breathing in an AASM task force
report (33), and a more recent guidelines to the classification and
scoring of respiratory events was published in the AASM Manual
for the Scoring of Sleep and Associated Events in 2007 (12).
Careful monitoring of both upper airway airflow and thora-
coabdominal movement in the respiratory tracings is used to
categorize central, mixed, and obstructive sleep apnea (OSA),
which is further discussed later. In addition, the monitoring and
 Chapter 41
quantification of blood oxygen levels are necessary and accom-
plished using pulse oximetry. Endoesophageal (intrathoracic)
pressure recording is also sometimes used in the detection of
partial obstructions or upper airway resistance that can exist
without actual apneas or hypopneas. The nasal pressure trans-
ducer (NPT) is recommended for the measurement of subtle
reductions in airflow and can also be used in the identification
of absent airflow, characteristic of apnea.
Airflow
Upper Airway Exchange: Thermistors,
Thermocouples, and Pressure
Upper airway exchange can be monitored by thermistor resist-
ance fluctuations or by thermocouples, both sensitive to temper-
ature changes of air passing in and out through the mouth and
nostrils. Both of these sensors are attached under the nostrils
and can also be designed or applied with an additional sensor
pointing to the mouth, they are useful for documenting mouth
breathing. Thermistors generally use a ceramic or polymer
material and detect temperature changes based on a Wheatstone
bridge using DC voltage. Thermocouples are constructed of dis-
similar metals (e.g., constantin and copper) that generate a
potential in response to temperature change and are therefore
thermoelectric generators. They can be connected directly to the
plug-in box and used with AC voltage for recording.
Nasal Flow Using Nasal Airway Pressure
Nasal flow can be monitored using a simple cannula, typically
used to deliver oxygen, attached to a pressure transducer and
may be recorded using very low-frequency filters with AC or
DC amplifiers. This technique is more informative and reliable
than thermistors and thermocouples and is a recommended
method for the detection of hypopneas (12). With appropriate
filtering of the slow component to reveal the vibration signal
using an AC amplifier, nasal pressure can also be used to record
snoring. The primary concern in using nasal pressure exclu-
sively to measure upper airway flow is that the signal may drop
due to mouth breathing. However, additionally a simple oral
thermistor/thermocouple can be used to resolve this problem.
The pressure transducer can also be used during continuous
positive airway pressure (CPAP) titration to monitor respira-
tory waveform and mask leak.
CAPNOGRAPHY
Capnography refers to the monitoring of the concentration of
carbon dioxide (CO2) in the respiratory gases and is measured
in the breath at the end of complete expiration (end tidal). A
CO2 analyzer is sometimes used to document CO2 retention in
sleep-related respiratory disorders, especially COPD and
hypoventilation syndromes.
END TIDAL CO2
End tidal CO2 refers to the measurement of the CO2 in the air
exhaled from the lungs, and the normal range is 4% to 6%
(equivalent to 35 to 45 mm Hg).
■ Polysomnography: Technical and Clinical Aspects 823
PNEUMOTACHOGRAPHY
This technique is rarely used in diagnostic testing because it
involves the placement of an airtight mask on the face and uses
a pressure transducer system to measure flow rate, tidal volume,
and other respiratory variables. This method is adopted fre-
quently in the sleep testing laboratory during CPAP pressure
titration in order to monitor adequacy of pressure and flow (see
Nasal Pressure heading above).
RESPIRATORY EFFORT
Piezoelectric Strain Gauges
Thoracoabdominal movement is frequently monitored with
piezoelectric crystal sensors incorporated into bands that are
fitted around the torso at the level of the chest and abdomen.
Generally these sensors can be plugged directly into the head-
box and connected to the AC amplifier. Respiratory excursion is
monitored and the traces are used in detecting reductions in
breathing effort, but it does not produce a measured signal,
rather only a relative one.
Inductance Plethysmography
Inductive plethysmography uses wire coils placed around the
chest and abdomen, sometimes woven into a vest, to monitor sep-
arate movements and then add them together, mimicking total
spirometric volume. In contrast to the piezoelectric belts, this
measuring technique is not only adjusted for visual characteristics
of the wave but also calibrated before the sleep period begins. It
produces a measured output for chest movement, abdominal
movement, and total volume. This system has an advantage in
being able to clearly distinguish paradoxical breathing associated
with obstruction in contrast with central cessation of breathing
that is reflected in a loss of signal in all three tracings.
INTERCOSTAL EMG
Long electrode leads are sometimes applied on the lower back
intercostal muscles to aid in the detection of effort during
breathing and the placement is sensitive to muscle activity dur-
ing coughing.
SNORING MONITOR
In addition to the NPT method mentioned earlier in this chap-
ter, both piezo sensors (encased in a rubber disk attached to the
throat and plugged into the headbox) and microphones are
used to monitor snoring. Snoring is associated with reduced
upper airway diameter, and bursts of loud guttural inspiratory
snorting are signs of OSA.
INDIRECT MEASUREMENT
OF BLOOD GASES
Transcutaneous (TcCO2) Measurement
Transcutaneous (TcCO2) measurement is the measurement of
CO2 gas tension of tissue underlying a specialized electrode that
is usually placed on the chest or forearm. The electrode method
824 Part V ■ Complementary and Special Techniques
uses a Stow–Severinghaus glass/ceramic electrochemical sensor,
with a small heater unit to facilitate blood flow. Transcutaneous
monitoring of CO2 tension is more reliable than transcutaneous
measurement of O2 because CO2 diffuses better through the skin.
Finger Oximetry for Oxygen Saturation
Standard polygraphic diagnostic recording must include mon-
itoring of an index of blood oxygen saturation or the percent-
age of available circulating hemoglobin that is saturated with
oxygen (SaO2). Probes are easily attached to the finger with a
specially designed clip or with tape. Measurement is achieved
using an optical device that uses a DC channel involving com-
putation of absorption of certain wavelengths of light.
BODY POSITION
Body position is the useful information to have when interpret-
ing a pattern of respiratory disturbance and can be obtained
visually through infrared camera monitoring and/or with the
aid of a position sensor. Such sensors are available and can be
attached to respiratory bands with accurate output of informa-
tion including prone, supine, right side, and left side.
Table 41.2 lists appropriate filter settings for recording vari-
ous physiologic characteristics in the PSG.
SLEEP STAGING CRITERIA
In 1968, the publication of the sleep scoring atlas by
Rechtschaffen and Kales (11) represented the consensus agree-
ment between sleep researchers of the time to the scoring of sleep
stages. Remarkably, this document remains the dominant force in
sleep scoring and analysis. With the recent publication of the
AASM Manual for the Scoring of Sleep and Associated Events, the
standardization of sleep and event scoring has reached a new
level of consensus (12). The scoring of sleep in this recent docu-
ment varies only limitedly from that published in 1968 (11).
The nomenclature of sleep stages has changed with the recent
AASM scoring manual with NREM stage 1 now called stage N1,
and NREM stage 2 is now stage N2. What had previously been
NREM stages 3 and 4 (or slow wave sleep) is now unified into a
single stage N3. Stage REM is now labeled as stage R.
Tabl e 41 . 2
Filter Settings for PSG Recordings
High-Frequency Time Constant
Characteristics Filter (Hz) (second)
Electroencephalogram 70 or 35 0.4
Electro-oculogram 70 or 35 0.4
Electromyogram 90 0.04
Electrocardiogram 15 0.12
Airflow and effort 15 1 0.1
A sleep study is scored in 30-second epochs with each stage
labeled as stage W (or wakefulness), stage N1, stage N2, stage
N3, or stage R. If two or more stages coexist in an epoch, the
epoch is labeled as the stage that represents the greatest portion
of the epoch. Stage W represents the waking state ranging from
full alertness through early drowsiness. Stage W is scored when
more than 50% of the epoch demonstrates alpha frequency
activity over the occipital region (Fig. 41.3). Stage W is scored
in the absence of alpha activity (which occurs in 10% to 20% of
individuals) if eye blinks, reading eye movements, or irregular
conjugate REMs are identified with accompanying normal or
elevated chin muscle tone. The chin EMG during stage W is
variable but is usually higher than that seen during sleep stages.
Stage N1 represents late drowsiness and light sleep. It is
scored when greater than 50% of an epoch shows alpha attenu-
ation and is replaced by low amplitude, mixed frequency EEG
activity (Fig. 41.4). K-complexes and sleep spindles are absent
by definition. If an individual does not generate alpha activity,
stage N1 is scored when over 50% of the epoch demonstrates
theta range slowing (4 to 7 Hz), vertex sharp waves, or slow eye
movements. In individuals who do generate alpha activity, slow
eye movements are frequently seen before the disappearance of
alpha activity, so individuals who do not generate alpha activity
may have stage N1 scored slightly earlier than those who do
generate alpha activity. During stage N1, chin EMG is variable
but usually lower than that seen in stage W.
Stage N2 represents the predominant sleep stage during an
overnight recording and is scored when there is the appearance
of a K complex or sleep spindle (Fig. 41.5). In sleep medicine, a
K complex is defined as a biphasic negative sharp wave maxi-
mum at the vertex or frontal regions that lasts greater than 0.5
second. Sleep spindles are 11 to 16 Hz sinusoidal activity lasting
at least 0.5 second seen maximally in the central head region.
Less than 20% of an epoch of stage N2 can be delta activity
( 2 Hz, 75 V in amplitude). Stage N2 is scored from the
first epoch of stage N2 until a clear epoch of stage W or another
stage of sleep is identified. Although eye movements are usually
absent in stage N2, slow eye movements can sometimes be seen.
Chin or axial EMG is variable but usually lower than stage W.
Stage N3 (which encompasses both former stages 3 and 4
of R and K criteria (11)) is the deepest stage of sleep and is
Low-Frequency
Filter (Hz) Sensitivity
0.3 5–7 V/mm
0.3 5–7 V/mm
5.0 2–3 V/mm
1.0 1 V/cm to start; adjust
5–7 V/mm; adjust
Chapter 41 ■ Polysomnography: Technical and Clinical Aspects 825

Figure 41.3 The limited PSG montage illustrates

alpha activity (well seen after eye closure) consis-
tent with wakefulness or stage W (30-second
epoch, chin = submental EMG).

Figure 41.4 The limited PSG montage illustrates

the absence of alpha activity and the appearance
of slow eye movements (out-of-phase activity in
channels E1 and E2) consistent with stage N1
sleep (30-second epoch, chin = submental EMG).

Figure 41.5 The limited PSG montage illustrates

sleep spindles (sinusoidal activity in top two
channels) and broad vertex waves consistent
with stage N2 (30-second epoch, chin = submen-
tal EMG).
826 Part V ■ Complementary and Special Techniques
Figure 41.6 The limited PSG montage illus-
trates high-amplitude delta activity ( 2 Hz)
occupying greater than 20% of the epoch and
represents an example of stage N3. Waveforms
that span the two dotted lines in the top two
channels identify activity that is at least 75 V in
amplitude (30-second epoch, chin = submental
EMG).
associated with increasing delta activity. Stage N3 is scored
when 20% or greater of an epoch is delta activity (Fig. 41.6).
For the purposes of sleep scoring, delta activity is defined as
activity of 0.5 to 2 Hz that is at least 75 V in amplitude when
measured over the frontal or central regions. Sleep spindles may
persist into stage N3, but eye movements are usually absent.
Axial EMG is usually lower than stage N2 and may approach
that seen in stage R.
Stage R (formerly stage REM) requires three components for
it to be scored: low-amplitude mixed frequency EEG, low chin
EMG, and the presence of REMs (Fig. 41.7A,B). Phasic EMG
activity may occur, but tonic EMG activity must be at a level
that is as low as, or lower, than that occurring at any other time
during the study. Sleep spindles and K-complexes are absent.
Series of 2 to 5 Hz vertex negative “sawtooth” waves occur, par-
ticularly just before phasic REM activity.
Movement time (or stage M as described by Rechtschaffen
and Kales (11)) is no longer scored. During major body move-
ments that obscure greater than 50% of an epoch, the epoch is
scored as stage W if alpha activity is present at any time in the
epoch. If no alpha activity is identified, but an epoch of stage W
precedes or follows the epoch with a major body movement,
then the epoch is scored as stage W. If neither of these require-
ments can be met, the movement is scored the same as the
epoch that follows it.
HYPNOGRAMS
Normal sleep has a clearly defined architecture that occurs each
night. Hypnograms are a graphic display of the ultradian cycle
within a night’s sleep (34). Sleep stage is depicted in the vertical
axis with the hours of sleep on the horizontal axis (Fig. 41.8).
Sleep onset begins with a transition to stage N1 sleep followed
quickly by stage N2. Stage N3 (formerly NREM stage 3 or slow-
wave sleep) comes next and is particularly sustained in this first
sleep cycle in children and young adults. Sleep then briefly light-
ens to stage N2 and transitions into stage R for the first time,
usually about 90 minutes after sleep onset. This completes the
first sleep cycle, a pattern that repeats itself three to five times
during the typical night’s sleep. With each ensuing 90 minute
sleep cycle, there is decreasing amounts of stage N3 sleep and an
increasing amount of stage R sleep. Predictable changes are
noted in sleep architecture with aging and are illustrated by the
histograms in Figure 41.8. Beginning in middle age, stage 3 sleep
lessens and more wakefulness after sleep onset (WASO) is noted.
The number of arousals and awakening continues to increase
with aging, becoming particularly notable in the elderly (35,36).
AROUSAL SCORING
The original scoring guidelines of Rechtschaffen and Kales (11)
focused primarily on the scoring of the stages of sleep (as dis-
cussed above). The R and K criteria defined that the appearance of
30 seconds or more of waking background (which resulted in the
epoch being labeled as wake) would be labeled an awakening. The
R and K manual made reference to movement arousals, but no
other mention of brief EEG frequency changes was made. There
was little standardization to the scoring of arousals before the
publication of the position paper by the American Sleep Disorders
Association (ASDA) task force in 1992 (37). This consensus paper
carefully defined rules for scoring an arousal. In the recent AASM
scoring manual (12), these rules were distilled to a single rule.
Score arousal during sleep stages N1, N2, N3, or R if there
is an abrupt shift of EEG frequency to alpha, theta, or frequen-
cies greater than 16 Hz (but not spindles) that lasts at least
3 seconds, with at least 10 seconds of stable sleep preceding
the change in EEG frequency. Scoring arousals during REM
requires a concurrent increase in submental EMG lasting at
least 1 second (Figs. 41.9 and 41.10).
The use of the 3-second duration of EEG change is not based
on any judgment of physiologic impact, rather it is an arbitrary
value chosen by the task force partly due to increased interob-
sever reproducibility of arousal scoring (as compared to shorter
duration events). The requirement for an accompanying EMG
change in stage R is due to the routine appearance of faster EEG
frequencies during normal sustained stage R sleep.
 Chapter 41
Periodic Limb Movements in Sleep Scoring
Periodic limb movements in sleep (PLMS) (also called periodic
leg movements of sleep) are repetitive, stereotyped movements
of the legs characterized by tonic extension of the great toe
(with occasional superimposed clonic activity) variably accom-
panied by ankle dorsiflexion and knee flexion (38). These peri-
odic movements can rarely preferentially involve the arm and
thus the use of the term PLMS. These periodic movements may
sometimes be associated with an accompanying EEG arousal.
PLMS are commonly seen on PSGs particularly in the setting of
disrupted sleep or in the elderly. The clinical significance of
PLMS is often unclear, but may sometimes contribute to
insomnia or daytime sleepiness. In 1980, Coleman and co-
workers (38) made a major contribution in understanding
PLMS and various rules of scoring the movements have been
proposed, including that from the ASDA task force in 1993 (39)
and the World Association of Sleep Medicine—International
Restless Legs Syndrome Study Group (IRLSSG) in 2006 (40).
From here on we will focus on the recent guidelines given in the
AASM scoring manual.
■ Polysomnography: Technical and Clinical Aspects 827
Figure 41.7 A: An example of stage R sleep with
low-amplitude mixed frequency EEG activity,
absent axial EMG activity (chin channel repre-
sents submental EMG), and REMs (out-of-phase
activity in channels E1 and E2). (30-second
epoch). B: Another example of stage R sleep. Note
2 to 3 Hz low-amplitude sharp activity (seen in
C4–M1 channel) just prior to REMs. This activity
represents sawtooth waves, a pattern frequently
seen in stage R.
During PSG, the PLMS are quantified as to how many
events occurred and also whether an EEG arousal was associ-
ated with them (Fig. 41.11A,B). A significant leg movement is
defined as an increase in EMG lasting between 0.5 and 10 sec-
onds. The EMG increase has to be at least 8 V above the rest-
ing EMG. The leg movement is labeled as part of a PLM series
if it is one of the four consecutive leg movement events occur-
ring with an interval between 5 and 90 seconds. If each leg
movement event is separated by less than 5 seconds, they are
considered a single leg movement. Arousals are scored as asso-
ciated with the leg movement if there is less than 0.5 second
interval between the end of one event and the onset of the
other regardless of which is first. Limb movements should not
be scored if they occur within 0.5 second of the onset or end
of an apnea or hypopnea.
Similar periodic leg movements can be seen in wakefulness,
particularly in patients suffering from restless legs syndrome
(RLS). RLS is a disorder where the patient suffers from an
uncomfortable sensation in their legs, which is worse at night,
worse while at rest, and relieved with movement. RLS is a
828 Part V ■ Complementary and Special Techniques
Figure 41.8 Histograms representing normal sleep cycles for age. REM
sleep represented by darkened area. Note NREM stages 3 and 4 would
now be combined into stage N3. Horizontal axis is hours of PSG record-
ing. (From Kales A, Kales J. Recent findings in the diagnosis and treat-
ment of disturbed sleep. N Engl J Med. 1974;487–499, with permission.)
Figure 41.9 Example of arousal from NREM
sleep (stage N2) with appearance of fast activity
lasting at least 3 seconds. Note increase in EMG
activity, which commonly accompanies an EEG
arousal pattern but is not required for scoring an
arousal from NREM sleep (30-second epoch).
clinical diagnosis made in the clinic based primarily on the his-
tory and does not require PSG evaluation. PLMS is a diagnosis
made on the PSG as most patients do not have daytime symp-
toms and observer history is not usually helpful. Most patients
with RLS will also have PLMS, but PLMS most commonly
occur in the absence of restless legs symptoms. Periodic limb
movements in wakefulness (PLMW) are not part of the recently
published AASM scoring guidelines (12). However, several
investigators have suggested the utility of PLMW in assessing
patients with RLS. The scoring criteria are exactly the same as
those outlined above for the scoring of PLMS. These investiga-
tors have also proposed a SIT test (suggested immobilization
test) with associated quantification of PLMW as a diagnostic
test for patients with RLS (41).
Scoring of Respiratory Events
With the advent of the recognition of apneic events during
sleep (42), the cessation of identified airflow for 10 seconds has
been the standard definition of an apneic event. With subse-
quent recognition of the significance of partial airflow inter-
ruption (hypopneas), no consistent standard definition was
accepted and technical and scoring standards varied widely. The
recent AASM scoring manual defines both technical recording
requirements as well as scoring rules (12) which hopefully will
lead to an improvement in the standardization of quantifying
respiratory abnormalities during sleep.
The sensor used to score apnea is the oronasal thermal sensor
(thermistor or thermocouple) that has long been used to qualita-
tively assess airflow during PSG recording. The sensor for detec-
tion of a hypopnea is the NPT. The NPT is much more sensitive
to the identification of a decrease in airflow and is appropriate to
use for the identification of a hypopnea. Frequently during a
study, a clear decrease in airflow is identified in the NPT, with no
evidence of change in the thermal sensor. Previously, it was com-
mon to recognize cyclical arousals associated with loud snoring
and oxygen desaturation that were not scorable using the less
sensitive thermal sensors. This led to the recognition of the upper
airway resistance syndrome in which the patient had a clinical
Chapter 41 ■ Polysomnography: Technical and Clinical Aspects 829

Figure 41.10 An arousal from stage R with

appearance of 3 seconds of faster frequencies
accompanied by the appearance of EMG in the
axial (chin) EMG channel (30-second epoch).

Figure 41.11 A: An illustration of periodic limb

movements of sleep. This is a 2-minute epoch
with the appearance of EMG activity in the right
anterior tibialis (RAT) channel occurring approx-
imately every 30 seconds. The patient is in NREM
sleep (which is not able to be confidently deter-
mined in this 2-minute epoch). B: A 30-second
epoch taken from the time period illustrated in A.
Note the appearance of faster frequencies on the
EEG demonstrating an arousal associated with
this periodic movement.
830 Part V ■ Complementary and Special Techniques
syndrome consistent with OSA but without scorable apneic
events. With the use of NPT and the associated identification of
scorable hypopneas, it is much rarer to have a clinical suspicion
of unrecognized respiratory events contributing to a patient’s
symptoms. However, the NPT is frequently overly sensitive in
that it demonstrates no airflow (an apnea) while the thermal sen-
sor continues to demonstrate obvious airflow. As such, the NPT
is not to be used for determination of apneic events, which are to
be scored off of the thermal sensor recording.
An apnea is scored using the thermal sensor when there is a
90% decrease in identified airflow lasting at least 10 seconds.
The apnea is classified based on the accompanying inspiratory
effort identified using the inductive plethysmography belts from
the chest and abdomen. An apnea is labeled as obstructive if there
is continued or increased respiratory effort throughout the entire
period of absent airflow (Fig. 41.12B). The apnea is classified as
central if the event has no associated inspiratory effort through-
out the entire apneic period (Fig. 41.12A). The event is scored as
a mixed apnea if there is initially an absence of inspiratory effort
followed by resumption of inspiratory effort in the later portion
of the event (Fig. 41.12C). The duration of the absent respiratory
effort needed to score a mixed apnea is not defined, but usually
would be at least one complete breath cycle (4 to 6 seconds). It
should be noted that there is no requirement for any accompany-
ing oxygen desaturation or EEG arousal to score an apnea.
Two separate rules for scoring hypopneas were presented in
the AASM scoring manual (12). The “recommended” rule is
that used by Medicare and was chosen in an attempt to be in
concert with Medicare definitions and decision regarding cov-
erage of CPAP and other therapies. However, the “alternative”
set of rules for scoring hypopneas is actually more widely used
and captures a larger number of hypopneas, particularly in
individuals with healthy lungs where oxygen desaturations
occur much less frequently.
The recommended rule for scoring a hypopnea requires the
following:
1. NPT signal decreases by 30% for at least 10 seconds.
2. There is an accompanying oxygen desaturation of 4% or
more.
The alternative rule for scoring a hypopnea requires the
following:
1. NPT signal decreases by 50% for at least 10 seconds.
2. There is an accompanying oxygen desaturation of 3% or
more or the event is accompanied by an EEG arousal within
3 seconds of the event.
Either scoring method is acceptable but the rules used
should be clearly defined in the PSG report. Most labs do not
attempt to classify hypopneic events as obstructive or central
due to the inaccuracy of assessing respiratory effort in the
absence of more invasive monitoring techniques such as
esophageal manometry.
An optional scoring rule outlines the scoring of respiratory
effort–related arousals (RERAs). This may be particularly help-
ful in identifying potential clinically significant respiratory
events if only the recommended rule for hypopnea identifica-
tion is used. A RERA is scored if “there is a sequence of events
lasting at least 10 seconds characterized by increasing respira-
tory effort or flattening of the nasal pressure waveform leading
to an arousal from sleep when the sequence of breaths does not
meet criteria for an apnea or hypopnea.” Flattening of NPT
waveform is thought to identify increased airway pressure
reflecting partial airway occlusion. The flattening of the nasal
waveform is illustrated in Figure 41.13.
The number of scored respiratory events is then divided by
the hours of sleep to yield an AHI (apnea–hypopnea index)
which represents the number of respiratory events per hour of
sleep. This value is most commonly used to categorize the
severity of respiratory abnormality. The consensus statement
from the AASM classifies apnea severity as per the following
criteria: AHI 5: normal; AHI 5 but 15: mild apnea; AHI
15 but 30: moderate apnea; AHI 30: severe apnea (33).
Many observers suggest this categorization of apnea severity
may be too severe, particularly given the lack of normative data.
Several large studies have demonstrated that nonobese healthy
middle-aged adults have an AHI 5 in 20% of individuals, and
an AHI 15 in approximately 6% of individuals. The AHI is
also recognized to increase with age, even in individuals with no
sleep complaint (43–45).
Scoring of Cardiac Events
Before the AASM scoring manual (12), there had been no
attempt to define or standardize the documentation or scoring
of cardiac events during a PSG. The AASM scoring manual
defined the use of a modified ECG lead II (recording electrodes
placed on torso at approximately second rib to right of sternum
and the sixth rib near the apex of the heart on the left chest).
The following ECG scoring rules for adults were defined:
1. Score sinus tachycardia for sustained sinus heart rate 90
beats per minute.
2. Score sinus bradycardia for sustained sinus heart rate 40
beats per minute.
3. Score asystole for cardiac pauses 3 seconds.
4. Score wide complex tachycardia for a rhythm of at least three
consecutive beats at a rate greater than 100 beats per minute
and a QRS duration of 120 msec.
5. Score narrow complex tachycardia for a rhythm of at least
three consecutive beats at a rate 100 beats per minute and
a QRS duration of 120 msec.
6. Score atrial fibrillation if there is an irregularly irregular ven-
tricular rhythm associated with replacement of consistent P
waves with variable rapid oscillations.
7. Other significant arrhythmias (such as heart block) should
be reported if quality of the single lead is sufficient for accu-
rate identification.
SCORING OF CYCLIC
ALTERNATING PATTERN
The cyclic alternating pattern or CAP was first described by
Terzano and colleagues in 1985 (46). CAP has been proposed as
a tool for the comprehensive analysis of sleep microstructure in
both normal and pathologic conditions. Traditional sleep scoring
using the criteria of Rechtschaffen and Kales (11) has been the
Chapter 41 ■ Polysomnography: Technical and Clinical Aspects 831

Figure 41.12 A: Limited PSG montage demon-

strating no evidence of airflow as measured by
nasal pressure transducer (PTAF channel),
oronasal thermocouple (flow channel) accompa-
nied by the absence of ventilatory effort in the
thorax, and abdominal channels (measured by
respiratory inductive plethysmography belts).
This is an example of a central apnea lasting
approximately 16 seconds. B: Limited PSG mon-
tage demonstrating no evidence of airflow as
measured by nasal pressure transducer (PTAF
channel), oronasal thermocouple (flow channel)
accompanied by the evidence of continued venti-
latory effort in the thorax channel (measured by
respiratory inductive plethysmography belts).
This is an example of an obstructive apnea lasting
approximately 18 seconds. Note the appearance
of snore artifact (in snore and PTAF channel) with
resumption of airflow. C: Limited PSG montage
demonstrating an apneic event lasting approxi-
mately 40 seconds (in this 60-second epoch). No
evidence of airflow is noted by the nasal pressure
transducer (PTAF channel) or the oronasal ther-
mocouple (flow channel) throughout the event
identifying it as an apnea. During the first half of
the apneic period, there is no evidence of ventila-
tory effort, but then increasing evidence of effort
in the thorax channel is seen leading up to the ter-
mination of the apneic event. This is an illustra-
tion of a mixed apnea, having central features at
onset but then demonstrating obstructive features
as the event progresses.
832 Part V ■ Complementary and Special Techniques

Figure 41.13 Polysomnographic recording show-

ing an example of upper airway resistance syn-
drome. Note that peak increase in effort (indicated
by the solid arrowhead) is associated with a small
drop in peak flow and tidal volume, triggering a
transient electroencephalographic arousal. ECG,
electrocardiogram; EMGFACIAL, facial muscle elec-
tromyogram; EOG, electro-oculogram (right and
left); FLOWPNEUMOTACH, pnuemotachometer to
quantify airflow; Pes, esophageal manometry to
record esophageal pressure; RESPSUM, respiratory
effort; SaO2, saturation with oxygen. (Reproduced
from Chokroverty S. Sleep Disorders Medicine:
Basic Science, Technical Considerations and Clinical
Aspects. 3rd ed. Philadelphia: Saunders/ Elsevier;
2009, with permission.)

gold standard for looking at sleep macrostructure. CAP presents
another way of looking at NREM sleep within those sleep stages.
NREM sleep has been recognized to have high amplitude EEG
bursts such as K-complexes or delta bursts. These are often seen
as an arousal response to external stimuli but also occur when
sleep disturbance was not evident. An alternative view to this
arousal process is that these phenomena are associated with sleep
instability (due to an internal or external challenge to the sleep
process) and that this type of slow wave activity marks the brain’s
attempt to preserve or sustain the sleep state. If sleep becomes too
unstable, the preservation attempt fails and the high-amplitude
activity is accompanied by a more complete EEG arousal. It is
proposed that the addition of a periodicity dimension to the con-
cept of sleep stability and arousal will provide a valuable perspec-
tive on sleep and its relationship to underlying physiologic and
pathophysiologic mechanisms. CAP represents a slow oscillation
of EEG activity that manifests in the appearance of EEG arousal
patterns with a periodicity of 20 to 40 seconds.
CAP is represented by three characteristics:
1. The recurring high-amplitude EEG activity (phase A of the
period)
2. The intervening background EEG activity (phase B of the
period)
3. The period or cycle that is the sum of phase A and phase B
A CAP sequence is composed of repetitive CAP cycles, with
each cycle composed of a phase A and a phase B. Each phase of
the CAP is 2 to 60 seconds in duration. If there is no phase A
activity for 60 seconds, then that portion of NREM sleep is
scored an non-CAP. Figure 41.14 illustrates an example of the
CAP in stage 2 sleep. Figure 41.15 illustrates a brief period of
CAP surrounded by non-CAP sleep on either side. Phase A can
take on three patterns: A1 shows primarily slow ( 1 Hz) delta
activity with a small degree of autonomic activation; A3
demonstrates an increase in fast rhythms with strong auto-
nomic activation (an EEG arousal as per ASDA rules) (37); A2
represents a middle ground between the A1 and A3 subtypes
(Fig. 41.16). A summary paper outlining the rules for scoring
CAP and including an atlas to illustrate those rules was pub-
lished by Terzano and colleagues in 2001 (47).
In several studies, the CAP A-phase is identified as the “gate”
or “amplifier” that allows the appearance of pathologic events
such as PLMS, bruxism, sleepwalking, and sleep-disordered
breathing to occur. The majority of nocturnal partial seizures
presenting during NREM sleep also occur predominantly in
CAP in association with phase A (in particular in association
with K-complexes and delta bursts) (48). At this point, the clin-
ical significance of CAP remains unclear but further research is
warranted.
Figure 41.14 An example of CAP in stage N2.
The box outlines a CAP cycle (C) composed of
phase A and following phase B. Bipolar EEG
channels for first six channel top to bottom
Fp2–F4, F4–C4, C4–P4, P4–O2, F8–T4, and
T4–T6; bottom eight channels are top to bottom
Fp1–F3, F3–C3, C3–P3, P3–O2, F7–T3, T3–T5,
EOG, and ECG. (From Terzano et al. Atlas, rules,
and recording techniques for the scoring of cyclic
alternating pattern (CAP) in human sleep. Sleep
Med. 2001;2:537–553, with permission.)

Figure 41.15 Consecutive stretches of non-CAP

(top), CAP (middle), and non-CAP (bottom).
The middle section illustrates the minimal
requirements for definition of a CAP sequence (at
least three phase A’s in succession). The CAP
sequence occurs between two black arrows and
the transition between phases is delineated by
the dotted line. EEG derivation is Fp2–F4, F4–C4,
C4–P4, P4–O2, and C4–A1. (From Terzano et al.
Atlas, rules, and recording techniques for the
scoring of cyclic alternating pattern (CAP) in
human sleep. Sleep Med. 2001;2:537–553, with
permission.)

Figure 41.16 The three phase A subtypes. The

dotted line indicates the fast low-amplitude por-
tion of phase A. EEG derivation as in Figure 41.15.
(From Terzano et al. Atlas, rules, and recording
techniques for the scoring of cyclic alternating
pattern (CAP) in human sleep. Sleep Med. 2001;
2:537–553, with permission.)

833
834 Part V ■ Complementary and Special Techniques
INDICATIONS FOR PSG
The indications for PSG have been summarized by the pub-
lished practice parameters of the AASM (49,50). The domi-
nant indication for PSG is for the evaluation of sleep-related
breathing disorders. The increasing awareness of sleep apnea
and its possible impact on long-term cardiovascular health
has lead to a marked increase in evaluation of sleep-
disordered breathing. While the obese individual who pres-
ents with snoring, witnessed apneas, and excessive daytime
sleepiness (EDS) represents the obvious candidate for PSG
evaluation, there are many others who likely need PSG evalu-
ation. Symptoms of concern for possible sleep apnea include
daytime sleepiness, loud snoring, witnessed apneas, and unre-
freshing sleep. Additional medical conditions that warrant
exploring a possible role of OSA include intractable hyperten-
sion, new-onset atrial fibrillation (particularly with onset in
sleep), intractable congestive heart failure, and stroke. Once
the diagnosis of sleep apnea has been documented, patients
usually return for repeat study for CPAP titration. The goal is
to document the minimum CPAP pressure that eliminates
apnea, hypopneas, and arousals with particular attention to
stage R sleep and supine sleep. It is recommended that
patients undergoing upper airway surgery to treat snoring or
apnea have a study to document apnea severity and ensure
appropriate choice of therapy and perioperative management.
Similarly, after upper airway surgery to treat apnea, a follow-
up PSG is warranted to document the residual degree of sleep
apnea and whether additional therapeutic efforts are
warranted.
In addition to the evaluation of sleep-disordered breathing
(SDB), PSGs are also utilized to evaluate other causes of EDS,
sleep disruption, and unusual nocturnal behavior. In an indi-
vidual with EDS where sleep apnea is not a concern, a PSG can
be done to evaluate possible PLMS as a contributor to the
patient complaint. PSG is routinely done the night before an
MSLT (discussed later in this chapter) to document adequate
sleep quantity and quality to allow interpretation of the multi-
ple nap study. In patients with RLS, PSG is not routinely done
as this is a clinical diagnosis and the restrained environment of
the sleep laboratory can be very uncomfortable for the sympto-
matic patient with RLS. PSG may be appropriate if a patient
with RLS does not respond to therapy or has other potential
sleep pathology. Similarly, PSG is not routinely indicated in
patients presenting with insomnia. Both behavioral and phar-
macologic therapies are pursued before considering PSG evalu-
ation in a patient with insomnia.
Most PSG recordings are presently done with video accom-
panying the polygraph recording. Video recording is absolutely
essential if the goal is to evaluate unusual nocturnal behavior.
The diagnosis of a parasomnia can usually be made on a clini-
cal basis and a video PSG is not required in many patients.
However, if there are unusual features or a concern about coex-
istent sleep problems, PSG may be warranted. If the differential
diagnosis include nocturnal seizures, additional EEG channels
are often helpful in more confidently differentiating parasom-
nias from epileptic seizures.
AMBULATORY PSG RECORDING
Ambulatory recording of sleep in the evaluation of sleep disor-
ders has been rarely utilized due to the limitations of the tech-
nique and the lack of third-party reimbursement (51). With the
recent CMS (Centers for Medicare and Medicaid Services)
approval for coverage and the push to control health care costs,
the role of ambulatory or home sleep tests (HSTs) is being re-
evaluated (52). As outlined earlier in this chapter, in-laboratory
PSG is the gold standard for the evaluation of sleep disorders and
is classified as a type 1 study. A full but unattended PSG (seven or
more channels including the recording of EEG) is classified as a
type 2 study. The focus on the HSTs has been with the use of type
3 studies which have four to seven channels and focus on the
evaluation of respiratory parameters, but without recording of
EEG and therefore have no evaluation of sleep stages or arousals.
A type 4 study is a one- or two-channel overnight recording, with
at least one of them being oximetry (53).
In the typical HSTs, the patient presents to the laboratory
during the day and is educated regarding use of the equipment
and placement of electrodes and recording devices. After prepar-
ing for bed, the patient places the recording sensors and pushes
a button indicating that they are retiring for the night. The but-
ton is pressed again upon awakening in the morning and a total
recording time is obtained, which is supplemented by a brief
sleep log from the patient outlining any sustained periods of
wakefulness or other important information. Usual recorded
parameters include nasal/oral airflow, respiratory effort, oxygen
saturation, body position, snoring, and heart rate. The studies
are scored by automated scoring systems, but the data should be
reviewed and edited by a technologist to ensure that artifactual
information is discarded and that the results represent a reason-
able interpretation of the data. From this recording, an estimate
of an AHI and oxygen desaturation index is obtained.
The recent guidelines from the AASM for unattended sleep
studies outline minimal technical expectations for portable
monitoring (53). HSTs should at a minimum record airflow,
respiratory effort, and blood oxygenation and should use the
same sensors that are used in the laboratory studies. Accurate
data collection is paramount to successful use of the techniques,
so the devices should not be used by technologists or physicians
who are inexperienced in sleep disorders and their evaluation.
The devices must also allow review of the raw data so that man-
ual scoring and/or adjustment of the automated scoring algo-
rithms can be made.
The advantages of HSTs are obvious and include lower cost
and greater patient accessibility. Patients who cannot access in-
laboratory studies due to laboratory backlog or physical immo-
bility can be evaluated with HST. Other patients with irregular
or unusual sleep–wake cycles can also have an evaluation that
does not have to fit into the usual laboratory schedule. The pri-
mary disadvantages include unreliable or lost data and the lack
of any assessment of sleep stage or quality.
Because of their obvious limitations, portable or ambulatory
sleep studies are not appropriate in many clinical settings and
this is recognized by the CMS guidelines as well as those of
the AASM. Both bodies agree that the HST may be used as an
 Chapter 41
alternative to in-laboratory PSG for the diagnosis of OSA in
patients with a high pretest probability of moderate to severe
OSA. If positive, the caregiver can move forward with therapy
for the significant respiratory abnormality. Additionally, HSTs
can be helpful after CPAP therapy has been initiated to assess
adequacy of treatment if there is a concern regarding response
to therapy (52–55).-
However, HST is not indicated for use in patients with sig-
nificant comorbidity (COPD, congestive heart failure, neuro-
muscular disease) as the coexistence of these disorders may lead
to a greater degree of inaccuracy. HST has not been evaluated in
asymptomatic but high-risk individuals such as bariatric
surgery patients or commercial truck drivers.
Computerized PSG and Computer-Assisted Scoring
Visual sleep scoring has been the standard for analyzing sleep
from the very beginning and subsequent standardizations of
sleep scoring has emphasized that (11,12). Any attempt to uti-
lize computerized sleep scoring in clinical sleep studies has
attempted to model itself after the visual sleep analysis.
Although it is extremely useful for quantifying easily identified
events (oxygen desaturations), its inability to adjust to hostile
“pathologic states” such as frequent apneas, disrupted sleep,
and movement artifacts has limited its clinical application.
Although many of the commercially available PSG equipment
do come with computerized analysis of sleep stage, the clinical
standard remains visual scoring by a trained technologist or, at
least, the review and over scoring of the computerized sleep
staging by a technologist. Similarly, computerized identification
of respiratory or movement events also requires review and
editing by a technologist.
As part of the efforts to create the new scoring manual, there
was a critical evidence-based look at computerized analysis of
PSG. The individuals in charge of this section suggested that
“computer scoring and quantitative analysis are still in the
formative stage of development” and have not been proved to
be useful in clinical practice (56).
The real strength of computerized analysis of sleep is in its
ability to look at sleep as a continuum rather than the arbitrary
definitions provided for identifying stages N1, N2, N3, and
stage R sleep. Similarly, the microstructure of sleep is likely lost
in the 30-second epoch. Research is still needed to determine
whether this technology will contribute to new methods for
understanding sleep and its disorders (57).
PITFALLS OF PSG
One of the biggest pitfalls of PSG is the recognition of respira-
tory events that may be clinically significant but may not be
consistently recognized by the sensors used for the recording.
The clearest example of this is the upper airway resistance syn-
drome (UARS) (58,59). UARS is a syndrome in which patients
experience repetitive increases in upper airway resistance that
result in brief arousals with associated EDS. Not all of these
individuals snore, and the repetitive arousals can sometimes be
attributed to other causes such as PLMS. The “gold standard” of
documenting this disorder is the use of esophageal manometer
■ Polysomnography: Technical and Clinical Aspects 835
to document large negative intrathoracic pressure swings that
correlate with the EEG arousals. Esophageal manometry is not
commonly used in the sleep laboratory due to its effect on sleep
and patient comfort. In the last several years, the consistent use
of the NPTs to monitor airflow and airway pressure has
markedly increased our sensitivity to this phenomena, with
many of these events now being recognized as hypopneas due
to the superior identification of airflow resistance by NPT
(60–62).- However, the caveat remains that the sleep physician
needs to be sensitive to the possibility of the under recognition
of respiratory events in the sleep laboratory.
ARTIFACTS DURING PSG RECORDING
PSG recording frequently contains extraneous electrical activi-
ties not originating from the brain, eyes, heart, or other regions
of interest interfering with the biologic signals of interest (63).
These include physiologic, environmental and instrumental
artifacts.
Physiologic Artifacts
It is important to recognize physiologic artifacts resulting from
movements of the head, eyes, and other body parts (Figs. 41.17
through 41.19), rhythmic leg movements causing tremor-like
artifacts (Fig. 41.20), movements originating from muscles
causing myogenic artifacts (Fig. 41.21), and movements of the
heart giving rise to electrocardiographic artifacts (Fig. 41.22),
which may interfere with recordings from the regions of inter-
est and may sometimes be mistaken for slow waves in the EEG.
Sweating by causing excessive baseline swaying produces very
slow oscillations lasting for 1 to 3 seconds mainly during non-
REM sleep as a result of action potentials generated by salt con-
tent of the sweat glands (Fig. 41.23).
Environmental Artifacts
The most important environmental artifact is the 60-Hz artifact
(Fig. 41.24) resulting from electromagnetic radiation from AC
current in power lines (in many other countries the main fre-
quency is 50 Hz). Most often this results from high impedance
of the electrodes. The impedance should be kept below 5 K.
Instrumental Artifacts
These artifacts result from faulty electrodes (Fig. 41.25) resem-
bling transient sharp or slow waves limited to one electrode,
wires, switches, and the PSG machine itself due to random
fluctuations of charges causing instrumental noise artifacts.
CLINICAL CONSIDERATIONS
Narcolepsy and Other Hypersomnias
Narcolepsy is a disorder of EDS that is usually accompanied by
other “auxiliary” symptoms such as cataplexy, hypnagogic hal-
lucinations, and sleep paralysis. Additional features commonly
include disrupted night-time sleep and excessive motor activity
in sleep. A more complete review of the clinical features of nar-
colepsy can be found in the literature (64,65). During the initial
diagnostic evaluation of a patient with possible narcolepsy, a
836 Part V ■ Complementary and Special Techniques

Figure 41.17 A 30-second excerpt from

overnight PSG recording shows body movement
artifact following an obstructive apnea as demon-
strated by variable, high-voltage, asymmetrical
and asynchronous slow activity on EEG channels
and increased electromyography activity on chin
and tibialis anterior electrode channels. Top 10
channels: EEG; LT and RT EOG, left and right
electro-oculograms; chin EMG, electromyogra-
phy of chin muscles; Lt. and Rt. Tib EMG, left
and right tibialis anterior electromyography;
P Flow, peak flow; oronasal thermistor; chest
and abdomen effort channels; snore monitor;
EKG, electrocardiography. (Reproduced from
Chokroverty S, Thomas R, Bhatt M, eds. Atlas of
Sleep Medicine. Philadelphia: Elsevier; 2005, with
permission.)

PSG is usually done in tandem with an MSLT. The PSG excludes
possible pathologic entities such as sleep apnea or PLMS that
can contribute to EDS. The PSG is also used to document ade-
quate sleep duration (usually 6 hours) and sleep quality to
allow confident interpretation of the MSLT performed the fol-
lowing day. The focus of the MSLT is to document the patho-
logic hypersomnolence (as a shortened sleep latency)
and evaluate any abnormal REM sleep pressure through the
identification of sleep-onset REM (SOREM) in the brief nap
trials. In addition, it has long been recognized that patients with
narcolepsy seem to have an increased incidence of sleep apnea,
PLMS, RBD, and sleep-related eating disorders.
Idiopathic hypersomnia is another disorder of EDS that also
presents in adolescence or young adulthood. However, it does
not have the accompanying auxiliary symptoms or SOREM
sleep noted in narcolepsy. The EDS of idiopathic hypersomnia

Figure 41.18 A 30-second epoch from overnight

PSG showing repetitive blink artifact best noted
in the electro-oculogram channels and on the
frontal and anterior temporal EEG electrode
recordings. This can be misinterpreted for
abnormal cerebral activity. Top 10 channels:
EEGs; LT and RT EOG, left and right electro-ocu-
lograms; chin EMG, electromyography of chin
muscles; Lt. and Rt. Tib EMG, left and right tib-
ialis anterior electromyography; P. Flow, peak
flow; chest and abdomen effort channels; snore
monitor; EKG, electrocardiography. (Reproduced
from Chokroverty S, Thomas R, Bhatt M, eds.
Atlas of Sleep Medicine. Philadelphia: Elsevier;
2005, with permission.)
 Chapter 41
is usually less severe than that seen in narcolepsy, but it has a
more pervasive all-day character that is not impacted by day-
time naps. Night-time sleep in idiopathic hypersomnia is usu-
ally described as long and uninterrupted (66).
An early study of night-time sleep in narcolepsy and idio-
pathic hypersomnia (67) confirmed most of these clinical
impressions. Patients with idiopathic hypersomnia demon-
strated better sleep efficiency, less stage N1, more stage N3, less
WASO, and longer REM latency. In addition, the coexistence of
sleep-disordered breathing and PLMS was less frequently
■ Polysomnography: Technical and Clinical Aspects 837
Figure 41.19 A 20-second epoch from overnight
PSG showing rapid eye flutter artifact best
recorded in F3 and F4 electrode recordings dur-
ing the latter half of the epoch. A blink artifact is
noted in the initial part of the recording and is
best seen on the electro-oculogram channels.
Well-formed alpha rhythm is noted in the back-
ground on all EEG channels when eyes are
closed. (Reproduced from Chokroverty S,
Thomas R, Bhatt M, eds. Atlas of Sleep Medicine.
Philadelphia: Elsevier; 2005, with permission.)
noted. More recent studies (68,69) have expanded on the obser-
vations of PSG in narcolepsy patients. The night-time sleep of
narcoleptics is characterized by increased awakenings,
decreased stage N1, decreased stage N3, and shortened REM
latency. Total sleep time and total stage R (REM sleep) are not
effected. Evidence of abnormal motor activity during sleep is
prominent in narcolepsy (particularly in patients with accom-
panying cataplexy). PLMS index was greater than 15 in approx-
imately 75% of middle-aged narcolepsy patients, and somewhat
uniquely they also had PLMS during REM sleep, something
Figure 41.20 Overnight PSG from a 50-year-old
man. Nocturnal PSG shows the presence of mild
sleep apnea with an AHI of 12.3. A 30-second
excerpt from nocturnal polysomnography shows
bursts of rhythmic foot and leg movements on
the left/ right tibialis anterior muscle recording
channel during stage 2 NREM sleep (A). The
movement is not associated with respiratory
events, oxygen desaturation, or arousal from
sleep. Top 10 channels: EEGs; LEFT and RT EOG,
left and right electro-oculograms; chin EMG, elec-
tromyography of chin muscles; Lt/ Rt Tib EMG,
left/ right tibialis anterior EMG; oronasal thermis-
tor; chest and abdomen effort channels; snore
monitor; EKG, electrocardiography; SaO2, oxygen
saturation by finger oxymetry. (Reproduced from
Chokroverty S, Thomas R, Bhatt M, eds. Atlas of
Sleep Medicine. Philadelphia: Elsevier; 2005, with
permission.)
838 Part V ■ Complementary and Special Techniques

Figure 41.21 A 30-second excerpt of PSG

recording shows spikelike potentials of varying
amplitude and frequency denoting muscle activ-
ity not only on the chin electromyography chan-
nel but also in all EEG channels, particularly
prominent in T3, T4, A2, and A1 electrode deri-
vations. Top 10 channels: EEGs; LT and RT EOG,
left and right electro-oculograms; chin EMG, elec-
tromyography of chin muscles; Lt. and Rt. Tib
EMG, left and right tibialis anterior electromyog-
raphy; P Flow, peak flow; oronasal thermistor;
chest and abdomen effort channels; snore moni-
tor; EKG, electrocardiography. (Reproduced from
Chokroverty S, Thomas R, Bhatt M, eds. Atlas of
Sleep Medicine. Philadelphia: Elsevier; 2005, with
permission.)

which is distinctly unusual in non-narcolepsy patients. Even
more remarkable is the high incidence of RBD. In a group of
narcolepsy patients off medication (68), 60% provided a his-
tory of RBD and 40% demonstrated RBD during PSG. In addi-
tion, another 14% had evidence of REM sleep without atonia.
It is clear that excessive and pathologic motor activity during
sleep is common in individuals with narcolepsy.
Episodic hypersomnias are also described and include
Kleine–Levin syndrome (KLS) and menstrual-associated hyper-
somnia. KLS is a rare disorder with onset during teenage years
and is predominantly seen in men. Diagnosis is based on the
clinical presentation of recurrent episodes of hypersomnia along
with at least one of the following symptoms during the hyper-
somnolent episodes: binge eating, hypersexuality, irritability,
aggression, other odd behaviors, or cognitive abnormalities. The
pathogenesis is unknown with hypotheses centering on neuro-
transmitter imbalance or autoimmune mechanisms. There is no
consensus on accompanying PSG findings in KLS, likely related

Figure 41.22 This is a 30-second excerpt from

an overnight PSG selected to show electrocardio-
graphy artifact, which is characterized by its
morphology, rhythm, and synchrony with elec-
trocardiographic recording. It is seen throughout
the epoch on electroencephalography, electro-
oculography, and tibialis electromyography chan-
nels. Incidentally, an obstructive apnea is seen in
the airflow and effort channels. Top 10 channels:
EEGs; LEFT and RT EOG, left and right electro-
oculograms; CHIN, electromyography of chin;
Lt/ Rt Tib EMG, left and right tibialis anterior
electromyography; oronasal thermistor; chest
and abdomen effort channels; snore monitor;
EKG, electrocardiography. (Reproduced from
Chokroverty S, Thomas R, Bhatt M, eds. Atlas of
Sleep Medicine. Philadelphia: Elsevier; 2005, with
permission.)
 Chapter 41
to variability in the timing of the studies with reference to the
hypersomnolent period. One recent study (70) documented a
reduction in slow wake sleep early in the hypersomnolent phase,
returning to normal during second half, at which time REM
sleep was decreased. Other studies suggested increased total
sleep time, reduced sleep efficiency, increased WASO, decreased
stage N3, and sometimes shortened REM latency. MSLTs have
consistently demonstrated shortened MSL during the episodic
hypersomnolence (MSL in range of 3 to 5 minutes) and normal
MSL during asymptomatic intervals. SOREMs have been
■ Polysomnography: Technical and Clinical Aspects 839
Figure 41.23 Sweat artifact at 30 seconds.
(Reproduced from Siddiqui F, Osuna E, Walters
AS, et al. Sweat artifact and respiratory artifact
occurring simultaneously in polysomnogram.
Sleep Med. 2006;7(2):197–199, with permission.)
reported but are not consistently noted (71). Menstrual-associ-
ated hypersomnia is a disorder in which a young woman (usu-
ally near the time of menarche) experiences excessive sleep need
and daytime sleepiness in association with her menses. It is usu-
ally improved with oral contraceptive medication and has not
been evaluated extensively in the sleep laboratory.
Parasomnias
NREM parasomnias include confusional arousals, sleep walk-
ing, and sleep terrors (72,73). Typically, arousal parasomnias
Figure 41.24 A 30-second excerpt from an
overnight PSG recording shows a uniform
monorhythmic artifact at 60 Hz on left/ right tib-
ialis electromyography channel, which is further
superimposed by an electrocardiography artifact.
The 60-Hz artifact is due to electrical interference
from power lines and equipment occurring at a
frequency of 60 Hz in North America (but at
50 Hz in many other countries). Maximum inter-
ference is seen in the presence of poor electrode
contact. Incidentally, an obstructive apnea is
noted in the airflow and respiratory effort chan-
nels. Top 10 channels: EEGs; LEFT and RT EOG,
left and right electro-oculograms; chin EMG, elec-
tromyography of chin muscles; Lt/ Rt Tib EMG,
left and right tibialis anterior electromyography;
oronasal thermistor; chest and abdomen effort
channels; snore monitor; EKG, electrocardiogra-
phy; SaO2, oxygen saturation. (Reproduced from
Chokroverty S, Thomas R, Bhatt M, eds. Atlas of
Sleep Medicine. Philadelphia: Elsevier; 2005, with
permission.)
840 Part V ■ Complementary and Special Techniques
Figure 41.25 A 30-second excerpt from an
overnight PSG recording shows the presence of a
C3 electrode pop artifact. Near the middle of the
epoch, it is accentuated by a movement artifact
simultaneously recorded on the chin electromyo-
graphy, left and right tibialis anterior electromyo-
graphy, and snoring channel. Top 10 channels:
EEGs; LEFT and RT EOG, left and right electro-
oculograms; chin EMG, electromyography of chin
muscles; Lt/ Rt Tib EMG, left and right tibialis
anterior electromyography; oronasal thermistor;
chest and abdomen effort channels; snore moni-
tor; EKG, electrocardiography. (Reproduced from
Chokroverty S, Thomas R, Bhatt M, eds. Atlas of
Sleep Medicine. Philadelphia: Elsevier; 2005, with
permission.)
occur in the first third of the night during a transition from
slow wave sleep (stage N3) into lighter stages of sleep. All of
these parasomnias are more common in children (aged 3 to 10
years) but can occur in adults. During confusional arousals the
individual whimpers, moans, moves about slowly, may speak in
short phrases, and resists consoling or attention from others.
These events may last up to 30 minutes. Sleep terrors tend to be
briefer but are associated with screams and autonomic activa-
tion (sweating, tachycardia, flushed face). Again the child resists
the consoling efforts of the parents and returns to sleep without
memory of the event. Sleep walking events are similar to confu-
sional arousals in terms of time of night, duration, and charac-
ter of the behavior, but the individual moves about the house in
a somewhat slow clumsy fashion and without emotional con-
tent. In each of these parasomnias, PSG recordings usually doc-
ument generalized, high-amplitude rhythmic delta or theta
activity during the event. No other abnormalities are usually
identified on PSG. The CAP discussed earlier in this chapter has
been described to be increased in children with sleep terrors.
This makes sense as CAP represents a pattern of “unstable
sleep” and may trigger a variety of events in sleep that can range
from epileptic seizures to periodic movements to bruxism to
confusional arousals. Other sources of sleep disruption, such as
sleep-disordered breathing and PLMS, may also play a role in
precipitating the arousals from slow-wave sleep.
REM sleep–related parasomnias include nightmare disorder
and REM behavior disorder (RBD). The International
Classification of Sleep Disorder, 2nd ed. (ICSD-2) (74) defines
nightmares as “recurrent episodes of awakening from sleep with
recall of intensely disturbing dream mentation, usually involv-
ing fear or anxiety, but also anger, sadness, disgust, or other dys-
phoric emotions.” There is generally full alertness upon
awakening and intact recall of dream experience. PSG is not
routinely indicated for the evaluation of nightmares, but when
such an event is recorded, the patient is usually observed to
abruptly awaken out of REM sleep without pathologic motor
activity or any other PSG-identified abnormality.
RBD is a disorder in which the individual usually demon-
strates aggressive motor activity during dream enactment.
The patient can be awakened and then usually is able to recall
dream content that explains his behavior. The disorder is
more common in men, particularly in the elderly. It usually
comes to medical attention when the aggressive dream enact-
ment has resulted in injury to the patient or bed partner (75).
In up to 50% of patients, it may be a heralding symptom of a
degenerative neurologic condition such as the synucle-
inopathies (Parkinson disease [PD], Lewy body dementia)
(76). It is also seen more frequently in other neurologic disor-
ders with brainstem dysfunction including multiple sclerosis
and narcolepsy (77).
The behavioral events of RBD occur during REM sleep
and as such tend to be more frequent in the early morning
hours. Frequency is variable and as such a single-night PSG
recording will often not record any of the dream enactment
behaviors. Greater attention has been paid to objectively evalu-
ating the degree of muscle atonia present in REM sleep as a way
to identifying patients with increased risk of dream enactment
(78) (Fig. 41.26). The AASM scoring manual has defined crite-
ria for scoring pathologic muscle activity during REM sleep
(12). Sustained muscle activity (tonic activity) in REM sleep is
defined as an epoch of REM in which at least 50% of the epoch
duration has a chin EMG amplitude greater than the minimum
amplitude in NREM sleep. Excessive transient muscle activity
(phasic activity) in REM sleep is defined as when at least 5 of
the 10 three-second “mini-epochs” (in a 30-second epoch) have
bursts of transient muscle activity. In RBD, the excessive transient
 Chapter 41 ■ Polysomnography: Technical and Clinical Aspects 841

Figure 41.26 A 60-second epoch of PSG tracing from a 66-year-old man with idiopathic RBD
showing stage R without muscle atonia and with excessive phasic (transient) EMG bursts in cra-
nially innervated (Masseter, Chin, and Sternomastoid), upper limbs (RT Biceps, and LT Biceps),
lower limbs (RT Quadriceps, LT Quadriceps, RT and LT TIBialis, RT and LT Gastrocnemius),
and trunk (RT upper Rectus and RT Lumbar Paraspinal) muscles. Top four channels: EEGs
(international electrode nomenclature); left and right EOGs (E1–M1; E2–M1); EKG, electrocar-
diogram; respiratory air flow (OroNasal and PFlow), and effort (Chest and Abd); HR, heart rate;
Intercostal 1, intercostal muscle EMG; snore; SAO2, oxygen saturation by finger oximetry; BODY,
body position.

muscle activity bursts are 0.1 to 5.0 seconds in duration and at
least four times the amplitude of background EMG activity.
The PSG characteristics of RBD can include either or both the
tonic and phasic EMG observations. REM sleep without atonia
can occur without accompanying clinical behaviors and as such
RBD remains a clinical diagnosis that can be complemented by
the abnormal observations of muscle activity during REM sleep
as defined by the AASM scoring guidelines (12). Selective sero-
tonin reuptake inhibitors and selective norepinephrine reup-
take inhibitors commonly provoke REM sleep without atonia,
with or without associated clinical RBD episodes. Most investi-
gations have also reported an increased incidence of PLMS in
individuals with RBD. Other abnormalities are not usually
described on PSG.
Sleep-Related Epilepsies
An extensive overview of sleep and epilepsy is beyond the scope
of this chapter. However, sleep-related epilepsies are briefly dis-
cussed in this chapter, with a focus on their differentiation from
other sleep disorders, namely confusional arousals and RBD.
We also briefly review the role that PSG may have in the evalu-
ation of a patient with epilepsy.
An EEG recording during sleep is more sensitive to the iden-
tification of interictal epileptiform abnormality as compared to
an EEG recording only during wakefulness. Given the difficul-
ties in recording a sleep EEG during the daytime (usually
requiring sleep deprivation or sedation), the use of an all-night
sleep recording to identify interictal epileptiform activity may
be useful. No objective data is available to identify how often
this may be helpful. Clinical experience has demonstrated that
the limited three- or four-channel EEG montage routinely used
during nocturnal PSG is of very limited use in identifying
interictal epileptiform discharges. In addition, review of the
recording utilizing 30 or 60 seconds per page further limits the
identification of epileptiform activity. If PSG is used to evaluate
interictal epileptiform activity, a full 16- or 18-channel EEG
recording with review at the rate of 10 seconds per page is rec-
ommended (79,80).
PSG recordings may be utilized to assist in the diagnosis of
unusual nocturnal behavior. In most settings, prolonged video-
EEG recording is used as the primary diagnostic tool if the pri-
mary diagnostic concern is seizures. Overnight PSG may be
used to evaluate potential epileptic seizures, particularly in the
absence of the availability of prolonged video-EEG monitoring.
PSG does more confidently identify sleep stage (and any
accompanying sleep abnormality such as sleep apnea) than
video-EEG recordings. However, the limited EEG montage used
on routine PSG represents a significant disadvantage. Foldvary
and colleagues (79) published a study in 2000 illustrating that a
four-channel EEG recording was inadequate for electrographic
seizure detection during PSG recording. That same group
recently extended their observations to the evaluation of 8- and
18-channel montages (80). In that study at least 25% of seizures
with an EEG correlate were missed on the eight-channel
recording. The authors conclude that abbreviated montages are
inadequate in differentiating seizures from nonepileptic noc-
turnal events during PSG (79,80).
Nocturnal frontal lobe epilepsy (NFLE) is a remarkable dis-
order where the seizures are manifest by bizarre motor behav-
ior that occurs almost exclusively during sleep. The events are
usually stereotyped, but vary in their duration and complexity.
The issue of stereotyped behavior is important as most para-
somnias will not have the same degree of stereotypy common
to epileptic seizures. Recording multiple events may help in the
842 Part V ■ Complementary and Special Techniques
identification of the stereotyped nature of these events.
Interictal EEG recordings are commonly normal and structural
brain abnormalities are unusual in patients with NFLE. So the
recording of the events is frequently required for confident
diagnosis. Even with the recording of the events, unambiguous
ictal EEG correlate may be absent in up to 50% of frontal lobe
seizures, further complicating the diagnosis (81,82).
As previously discussed, the usual EEG pattern seen during
a confusional arousal is diffuse hypersynchronous slowing in
the theta or delta range. During RBD, the EEG is that of REM
sleep. In both cases, superimposed EMG and movement artifact
complicate the interpretation. A wide range of electrographic
patterns may accompany nocturnal frontal lobe seizures, but
bilateral rhythmic sharp waves are most commonly seen. In the
absence of a definite ictal electrographic correlate, final diagno-
sis is often reached clinically. As opposed to parasomnias, the
seizures of NFLE can begin at any age, occur more frequently
(often many times) per night, are briefer (several seconds up to
3 minutes), and have a stereotyped pattern that is absent from
parasomnias. Several reviews in the literature attempt to review
how clinical information may help in differentiating the char-
acter of these nocturnal events (82).
Sleep Apnea–Hypopnea Syndrome
Sleep apnea–hypopnea syndrome is divided broadly into two
types: upper airway OSAS and central sleep apnea syndrome
(CSAS) depending upon pathophysiologic mechanisms. The
most common indication for referral of patients to sleep labo-
ratories for PSG recording is the suspicion of OSAS. Two
groups of neurologists, Gastaut et al. (42) from France and Jung
and Kuhlo (83) from Germany independently discovered in
1965 the site of obstruction in the upper airway. In CSAS, the
Figure 41.27 A 120-second excerpt from stage
N2 showing an initial brief period of obstructive
apnea followed by two sequential hypopneas.
Top 10 channels: EEGs (international 10 to 20
nomenclature); LT and RT EOG: left and right
electro-oculograms; chin EMG, electromyogra-
phy of chin muscles; Lt. and Rt. Tib EMG, elec-
tromyography of left and right tibialis anterior
muscles; P FLOW, peak flow; nasal pressure
recording; oronasal, thermistor for air flow; chest
and abdomen, respiratory effort channels; EKG,
electrocardiogram; SAO2, oxygen saturation by
finger oximetry.
problem lies in the ventilatory control mechanism in the central
nervous system.
Upper Airway Obstructive Sleep Apnea Syndrome
Upper airway OSAS is characterized by repetitive episodes of
apnea (see Fig. 41.12A–C) or hypopnea (Fig. 41.27) during
sleep resulting in arterial oxygen desaturation and repeated
arousals. OSAS causes significant morbidity and mortality and
is often associated with a variety of comorbid conditions. This
often remains undiagnosed or underdiagnosed because of
insufficient knowledge and awareness of serious consequences
resulting from this disorder. Apnea or cessation of breathing
during sleep can be of three types: central, obstructive, and
mixed. In obstructive apnea, the airflow stops while the effort
continues (see Fig. 41.12B), whereas in central apnea, both the
airflow and the upper airways (e.g., pharynx and nose) and the
effort of the diaphragm and other respiratory muscles cease
(see Fig. 41.12A). An initial period of central apnea followed by
a period of obstructive apnea constitutes mixed apnea (see Fig.
41.12C). Upper airway OSAS is the most common cause of
apnea in the adult. The AHI (number of episodes of
apnea–hypopnea per hour of sleep) must be at least 5 to be con-
sidered pathologic. Respiratory disturbance index (RDI) should
not be confused with AHI, as RDI includes, in addition to the
number of apneas and hypopneas, the number of flow limita-
tions accompanied by arousals per hour of sleep (RERAs).
These rules are meant for the adult sleep scoring of breathing
events, but for the pediatric population the rules are slightly dif-
ferent (12).
OSAS is common in middle aged and elderly men. The inci-
dence of OSAS in a previously healthy population is not known.
However, based on a definition of AHI of 5 per hour of sleep
 Chapter 41
accompanied by EDS, the prevalence of OSAS is thought to be
4% in men and 2% in women between the ages 30 and 60 years
(43). In women, the prevalence of OSAS rises after menopause.
The pathogenesis of OSAS includes both local anatomical fac-
tors (e.g., narrowing of the upper airway and excessive relax-
ation of the upper airway muscles during sleep, increased upper
airway resistance) and neurologic factors causing dysfunction
of the respiratory neurons in the brain stem. The symptoms of
OSAS can be divided into nocturnal and daytime symptoms.
Nocturnal symptoms consist of loud snoring (often with a long
history), cessation of breathing as witnessed by bed partners,
choking, abnormal motor activities, and thrashing about in
bed, as well as hyperactivity and nocturnal enuresis, particu-
larly in children, and occasionally excessive sweating (84).
Major daytime symptoms of OSAS include EDS, with patients
falling asleep during the day at inappropriate times and in inap-
propriate places often causing accidents on the road and at
work. The daytime function is impaired complaining some-
times about morning headache and forgetfulness, and men may
report impotence. This pattern of sleepiness of prolonged dura-
tion and nonrefreshing nature is quite different from short-
duration refreshing narcoleptic sleep attacks. On physical
examination, obesity was noted in approximately 70% of the
patients and in many cases local anatomical abnormalities of
the upper airway were found. Repeated hypoxemias associated
with cessation of breathing and arousals during sleep may lead
to hypertension, cardiac arrhythmias, heart failure, and stroke
(84,85). The diagnosis is confirmed by laboratory techniques
(see PSG). PSG findings of OSAS include recurrent episodes of
apnea and hypopneas (see Figs. 41.12A–C and 41.27), which are
mostly obstructive or mixed with a few episodes of central
apneas accompanied by oxygen desaturation followed by
■ Polysomnography: Technical and Clinical Aspects 843
arousals on resumption of breathing with increased arousal
index (the number of arousals per hour of sleep). Percentage of
oxygen desaturation associated with abnormal breathing events
may be classified into mild with desaturation of 80% to 89%,
moderate to moderately severe with desaturation of 70% to
79%, and severe with 69% or below. An arousal index may be
arbitrarily considered to be normal when the index is 10 or
below, whereas an index of 10 to 15 may be borderline and an
index above 15 is abnormal. In terms of sleep architectural
changes, patients with OSAS characteristically show reduction
of slow wave and REM sleep and spend their sleep mostly in
stage N2 (stage 2 NREM sleep). Short sleep latency and
increased time spent awake after sleep onset are the other sleep
architecture findings. In a high percentage of OSAS patients
(may be up to 25%), sleep maintenance insomnia has also been
observed. In addition, patients with OSAS have abnormal sleep
microstructure as evidenced by increased arousal index as well
as increased CAP rate. The best treatment for moderate to
severe OSAS is the application of CPAP titration causing upper
airway pressurization opening up the airway passages acting as
a pneumatic splint, thereby eliminating obstructive apnea, hypox-
emias, snoring, and sleep fragmentation. Figure 41.28 shows
respiratory and sleep hypnogram before and after CPAP titra-
tion. The optimal CPAP pressure must first be determined by
an overnight PSG study and the patient can then purchase a
home unit for use during the night’s sleep. There are also auto-
CPAP machines available, automatically titrating the pressure
according to detected abnormal breathing events. In terms of
patient adherence and clinical outcomes, the roles and advan-
tages of these devices over the standard CPAP machine remain
to be determined (86,87). The role of portable unattended
monitor or home sleep testing for diagnosis and treatment of
Figure 41.28 Sleep and respiratory hypnogram
before and after positive airway pressurization
(Bilevel) in a 56-year-old man with severe upper
airway obstructive sleep apnea syndrome
(OSAS). Note the initial part (left-hand side) of
the hypnogram before pressurization treatment
(None) shows repeated episodes of obstructive
apneas and hypopneas ( Respira tory Event)
accompanied by oxygen desaturation (Desatu -
ration) recorded by finger oximetry and followed
by repeated arousals (Arousal) and snoring
(Snore) during stages N1 and N2. Following
pressurization (right-hand column) treatment
(Treatment: Start), respiratory events and oxygen
desaturation are gradually eliminated. W, wake-
fulness; R: stage REM; N1, sleep stage N1; N2,
stage N2; N3, stage N3; Position: body position;
HR, heart rate; SAO2, oxygen saturation (note
artifacts).
844 Part V ■ Complementary and Special Techniques

OSAS patients is currently evolving (see above under “Ambulatory Tabl e 41 . 3

PSG Recording”) and has not been firmly established yet based on
validated scientific studies. In some mild to moderate cases of
OSAS, dental appliances or uvulopalatopharyngoplasty (UPPP)
may be tried.
Central Sleep Apnea Syndrome
CSAS includes primary or idiopathic type and CSAS due to
Cheyne–Stokes breathing (CSB) pattern (Fig. 41.29), CSAS due
to high altitude and due to central neurologic disorders, partic-
ularly brain stem dysfunction, as well as central sleep apnea due
to drug or alcohol abuse (e.g., use of opiates). Idiopathic or pri-
mary CSAS is rare and the patient may present with years of
frequent awakenings due to repeated episodes of central apneas
followed by arousals and sometimes may present with insom-
nia–hypersomnia syndrome. CSB is noted in patients with con-
gestive cardiac failure, neurologic disorders, and sometimes
with renal failure. The presence of CSB in cardiac failure
increases mortality. In primary CSAS, patients are usually
normocapnic or hypocapnic with PaCO2 of 35 mm Hg or
lower. In addition, the cycle length (apnea plus hyperpnea dura-
tion) is more than 45 seconds in CSB and less than 45 seconds
in primary CSAS (74). Repeated arousals causing sleep frag-
mentation are important factors causing EDS. CSB is a type of
periodic breathing characterized by cyclic changes in breathing
with a crescendo-decrescendo sequence separated by central
apneas or hypopneas (Fig. 41.29).
Circadian Rhythm Sleep Disorders
A mismatch between the body’s internal clock (in the SCN) and
geophysical environment either as a result of malfunction of the
biologic clock or as a shift in the environment causing this to be
out-of-phase results in circadian rhythm sleep disorders. These
patients present with sleep difficulties as a result of desynchro-
nization between their internal circadian rhythm and the exter-
nal time. Jet lag (associated with high-speed air travel across
several time zones) and shift work sleep disorder (in patients
Circadian Rhythm Sleep Disorders
Jet lag
Shift work
Delayed sleep phase type
Advanced sleep phase type
Irregular sleep–wake type
Free-running type (non–24 hour or nonentrained type)
Circadian rhythm sleep disorder comorbid with medical
conditions
Circadian rhythm sleep disorder due to drug or substance
abuse
who work nonstandard shifts) are the two most common
circadian rhythm sleep disorders. Other circadian rhythm sleep
disorders are listed in Table 41.3.
Jet Lag Disorder
Jet lag is experienced as a result of eastward or westward jet
travel after crossing several time zones, disrupting synchroniza-
tion between the body’s inner clock and external cues.
Symptoms do not occur after North/South travel. This desy-
chronization causes sleepiness, tiredness, general malaise,
impairment of judgment and concentration, and sometimes
disorientation. Sleep problems include difficulty maintaining
sleep, frequent arousals, and EDS. These symptoms may also be
associated with gastrointestinal problems and the body’s other
rhythms including those that control endocrine secretions, and
body temperature. These symptoms may last for several days
depending on the direction of travel and the number of time
zones traversed. They also depend on age; older individuals take

Figure 41.29 A 500-second excerpt showing the

classic Cheyne–Stokes breathing (CSB)
crescendo–decrescendo pattern from an
overnight polysomnographic (PSG) recording of a
69-year-old man. The presence of CSB through-
out most of the NREM sleep with marked decre-
ment or absence during REM sleep in this patient
with a history of hypertension and EDS suggests
incipient left ventricular failure. (Reproduced
from Chokroverty S. Sleep and its disorders. In:
Bradley WG, Daroff RB, Fenichel GM, et al., eds.
Neurology in Clinical Practice. 4th ed. 2004:
1993–2054, with permission.)
 Chapter 41
a longer time to readjust than younger individuals. It takes
longer to resynchronize the internal and external clocks when
traveling East than when traveling West. It takes approximately
1½ hours per day to readjust the internal body clock when trav-
eling East as compared with 1 hour per day when traveling
West. It is easier to lengthen a day (for example, when traveling
in a westward direction) than to shorten a day (for example,
when traveling East). Other factors that may contribute to jet
lag disorder include limited mobility when flying, dryness of
the eyes on the plane, headache, fatigue, gastrointestinal distur-
bances, and nasal congestion (jet symptoms).
Shift Work Sleep Disorder
In the United States, up to 2%–5% workers may suffer from
shift work sleep disorder as a result of working irregular shifts,
particularly night shifts. Symptoms include sleep disruption,
fatigue, and gastrointestinal symptoms, increasing chances of
being involved in traffic accidents and making errors on the job.
Adjustment of the work time schedule rarely improves the
symptoms of shift work sleep disorder.
Delayed Sleep Phase State
In delayed sleep phase state (DSPS), the patient’s major sleep
episode is delayed in relation to desired clock time causing
sleep-onset insomnia or difficulty awakening at the desired
time. A typical schedule consists of going to sleep late between
2 AM and 6 AM and waking up late between 10 AM and 2 PM (74).
These patients have difficulty functioning adequately during
daytime hours if they must wake up early in the morning to go
to school or work. They cannot function normally in society
due to the disturbed sleep schedule; their sleep architecture,
however, is generally normal if these individuals are allowed to
follow their own uninterrupted sleep schedule. The onset of
DSPS generally occurs during childhood or adolescence.
Sometimes a history of DSPS in other family members exists
and some patients may complain of depression. An unusually
long intrinsic circadian period due to an abnormality in the
biologic clock in the SCN is the fundamental problem. This
type of sleep problem can be clearly documented by keeping a
sleep log for one to two weeks and by recording sleep–wake
activities with an actigraph (discussed further under ‘Related
Laboratory Procedures’).
Advanced Sleep Phase State
Advanced sleep phase state (ASPS) is the converse of DSPS (74).
These patients go to sleep early in the evening and wake up early
in the morning. They experience sleep disruption and daytime
sleepiness when not going to sleep at an early hour. The condi-
tion is often seen in patients with depression and in normal eld-
erly individuals. Familial ASPS has been ascribed to mutation in
the hPer 2 gene. This circadian rhythm disorder can be easily
documented by monitoring with an actigraph for 1 to 2 weeks.
Free-Running Circadian Rhythm Disorder
A circadian rhythm sleep disorder, free-running type (nonen-
trained type or non–24-hour sleep–wake syndrome), is charac-
terized by a patient’s inability to maintain a regular bedtime
■ Polysomnography: Technical and Clinical Aspects 845
and wake time, and sleep onset occurs at irregular hours. There
is no entrainment or synchronization with the usual time cues
such as sunlight or social activities. This disorder is most often
seen in blind people. In each 24-hour sleep–wake cycle, the
patient displays increasing delay of sleep onset by approxi-
mately 1 hour, causing an eventual progression of sleep onset to
the daytime hours into the evening. This can be documented
clearly in an actigraphic recording (see later).
Irregular Sleep–Wake Circadian Rhythm Disorder
Circadian rhythm sleep disorder, irregular sleep–wake type, is
characterized by lack of clearly defined circadian rhythm of
sleep and wake. Patients are seen to be napping throughout the
24-hour period, but total sleep time is normal. This condition
may be seen in patients with degenerative neurologic disorders
such as dementia, and children with mental retardation.
Actigraphic recordings are very useful in diagnosing this condi-
tion or any of the circadian rhythm sleep disorders (discussed
later).
Sleep-Related Movement Disorders
This new category of sleep-related movement disorder is
included in the ICSD-2 (74). These movements consist of rela-
tively simple stereotyped movements disturbing sleep. This cat-
egory includes RLS, PLMS, sleep-related rhythmic movement
disorder, bruxism, and nocturnal leg cramps.
Restless Legs Syndrome
RLS is a sensory-motor neurologic movement disorder with
lifelong symptoms causing considerable morbidity. RLS is the
most common movement disorder, but it is uncommonly
recognized and treated despite a lucid description of the entity
in the middle of the last century. There is not a single diagnos-
tic laboratory test for RLS, therefore the diagnosis depends
entirely on physician’s critical evaluation of symptoms. The
diagnosis is based on the IRLSSG criteria established first in
1995 (88) and modified slightly in 2003 (89). These criteria
include essential, supportive, and associated features and are
listed in Table 41.4.
RLS is a lifelong disorder that often begins at a very young
age but is often diagnosed in the middle to later years. The
prevalence of RLS increases with age and the overall prevalence
has been estimated at about 10% of the adult population in the
North American continent and Western Europe, but the preva-
lence of most severe cases (symptoms occurring at least twice a
week) is about 2.7% (90). There is an increased incidence (50%
to 60%) in first degree relatives of idiopathic cases of RLS.
Recent genome-wide association and linkage analysis docu-
mented common variations in certain genomic regions con-
firming more than 50% increase in risks for RLS and PLMS
(91,92). Most of the abnormal leg movements are noted in
the evening while the patients are resting in bed. In severe
cases, however, movements are noted in the daytime while
patients are sitting or lying down, hence it is aptly termed as
quiescegenic disease. At least 80% of RLS patients have PLMS
and may also have PLMW. The greatest advance made in
recent years is in the treatment of RLS. Several double-blind
846 Part V ■ Complementary and Special Techniques

Tabl e 41 . 4

Clinical Diagnostic Criteria for Idiopathic Restless Legs Syndrome

Essential criteria
An urge to move the legs usually accompanied or caused by uncomfortable sensations in the legs
The urge to move or unpleasant sensations beginning or worsening during periods of rest or inactivity such as lying or sitting
The urge to move or unpleasant sensations are partially or totally relieved by movements, such as walking or stretching, at
least as long as the activity continues
The urge to move or unpleasant sensations are worse in the evening or night than during the day and only occur in the
evening or night
Supportive features
Dopaminergic responsiveness
Presence of periodic limb movements in sleep or in wakefulness
Positive family history
Associated features
Usually progressive clinical course
Normal neurologic examination in the idiopathic form
Sleep disturbance

placebo-controlled clinical trials have established the value of
dopaminergic treatment in RLS. The condition may have a pro-
found impact on sleep, although sleep disturbance is not listed
as an essential or supportive feature of RLS. In the RLS epi-
demiology, symptoms, and treatment (REST) Primary study
(93), the most troublesome symptom rated by almost half of
the patients included sleep-related symptoms. The same study
noted that about 69% of those with moderately severe to severe
idiopathic RLS subjects took more than 30 minutes to fall
asleep and 60% woke up three or more times at night, thus
causing sleep initiation and sleep maintenance difficulties with
sleep fragmentation. It is notable that despite disturbed sleep
being a prominent complaint in RLS patients, the topic of sleep
disturbance in RLS and its potential consequences (both short
term and long term) have not been adequately addressed.
There are only limited polysomnographic (PSG) studies to
characterize sleep in RLS patients. Hornyak et al. (94) noted
prolonged sleep-onset latency, shorter total sleep time, and
longer REM sleep latency in moderately severe RLS patients (29
women and 16 men) with a mean IRLSSG score of 24 as com-
pared to equal number of age- and sex-matched healthy con-
trols. These patients also had markedly increased PLMS index
and sleep fragmentation. Saletu et al. (95) also found increased
REM latency and decreased REM density but normal sleep
onset latency and total sleep time in 12 RLS patients compared
with healthy controls. Winkelman et al. (96) in a recent study
analyzed sleep data obtained from unattended, in-home PSG
recordings from 2971 subjects included in the sleep heart health
study (SHHS) which is a population and cross-sectional obser-
vational study, and found increased mean sleep latency and
arousal index in RLS patients compared to those without RLS.
They noted no difference in sleep stage percentages between
those with and without RLS. All the studies therefore docu-
mented significant sleep disruption in RLS patients, but there
are no studies adequately addressing the long-term conse-
quences of such sleep disruption. A significant number of RLS
patients are left with chronic insomnia despite improvement or
complete resolution of the RLS symptoms and the mechanism
of such chronic sleep disturbance remains undetermined. In
summary, the typical PSG findings in an RLS patient are pro-
longed sleep-onset latency, increased WASO, increased PLMS,
evidence of sleep-related breathing disorders in some patients,
and transient rise of blood pressure and heart rate associated
with PLMS and arousal.
Periodic Limb Movements in Sleep and Periodic
Limb Movement Disorder
PLMS is a PSG finding (see Fig. 41.11) characterized by recur-
ring stereotyped limb movements, particularly dorsiflexion of
the ankles and sometimes flexion of the knees and hips (see
Section 1). A PLMS index (number of PLMS per hour of sleep)
exceeding 5 in children and 15 in most adult cases is clinically
significant. PLMS appears most commonly in RLS occurring in
at least 80% of cases but may also occur in a large number of
other medical, neurologic, and sleep disorders, and with inges-
tion of medications (e.g., selective serotonin reuptake
inhibitors, tricyclic antidepressants) and even in normal indi-
viduals, particularly in subjects older than 65. In the current
ICSD-2 (74), periodic limb movement disorder (PLMD) is
characterized by PSG findings of increased PLMS without asso-
ciated RLS, causing repeated awakenings and sleep fragmenta-
tion giving rise to insomnia or daytime excessive sleepiness, and
 Chapter 41
is listed as a separate entity. There is, however, some controversy
regarding the diagnosis of PLMD causing sleep fragmentation,
arousals, and EDS. There is a growing body of evidence that
PLMS may simply be a PSG phenomenon and may not have
any specific clinical significance except the presence in the
majority of patients with RLS (97). Many sleep specialists think
PLMD is a separate entity but occurring in a very small number
of patients. Further investigations including outcome studies
are needed to document that PLMS or PLMD may cause sleep
disturbance or EDS.
Sleep-Related Rhythmic Movement Disorder
Sleep-related rhythmic movement disorder is characterized by
repetitive rhythmic stereotyped behavior occurring predomi-
nantly during transition from relaxed wakefulness to sleep
involving large muscle groups (74). The condition is mostly
seen in infants and children. The three most common types of
movements seen in this condition consist of head banging, head
rolling, and body rocking. Other types of sleep-related move-
ments in this disorder include body rolling, leg rolling, and leg
banging. The behavior results in significant complaint as man-
ifested by one of the following: interfering with normal sleep,
significant impairment in daytime function, or self-inflicted
bodily injury. PSG studies show characteristic rhythmic move-
ments occurring in the transition stage or stage N1 or N2 sleep
(Fig. 41.30). Rarely, these movements are noted in stage N3 or
REM sleep. The recent AASM scoring manual (12) established
the rules defining PSG characteristics of rhythmic movement
disorders as follows: the frequency varies from 0.5 to 2.0 Hz
with four consecutive movements forming a cluster of move-
ments showing an EMG activity that is twice the baseline back-
ground activity.
Nocturnal Leg Cramps
These are intensely painful contractions accompanied by mus-
cle tightness occurring during sleep. These spasms usually last
for a few seconds but sometimes persist for several minutes.
Cramps during sleep are generally associated with awakening.
■ Polysomnography: Technical and Clinical Aspects 847
Many normal individuals have nocturnal leg cramps and the
cause remains unknown. PSG recordings during the cramps
show nonperiodic bursts of gastrocnemius EMG activity. The
sleep-related leg cramps are not explained by another sleep dis-
order, medical or neurologic disorder, medications, or sub-
stance use disorder.
Sleep-Related Bruxism
(Nocturnal Tooth Grinding)
Sleep-related bruxism or nocturnal tooth grinding often pres-
ents between the ages of 10 and 20 years, but it may persist
throughout life often leading to secondary problems such as
temporomandibular joint dysfunction. It is also commonly
noted in children with mental retardation or cerebral palsy.
Nocturnal bruxism is noted most prominently during stage N1
or N2 and REM sleep. The episode is characterized by stereo-
typical tooth grinding and is precipitated by anxiety, stress, and
dental disease. Occasionally, familial cases have been described.
The scoring rules in the PSG for sleep bruxism include the fol-
lowing (12): They must consist of brief phasic or tonic eleva-
tion of EMG activity of twice the amplitude of background
EMG; the chin EMG elevations are scored as bruxism if they are
0.25 to 2 seconds in duration and must occur in a regular
sequence of three consecutive elevations; sustained elevations
are scored as bruxism if the elevation is more than 2 seconds
(Fig. 41.31). There must be a period of at least 3 seconds of sta-
ble background chin EMG before a new episode of bruxism can
be scored. Sleep-related bruxism can also be scored reliably by
audio in combination with EMG with minimum of two audi-
ble tooth grinding episodes per night of PSG in the absence of
epilepsy.
Sleep-Related Eating Disorders
Sleep-related eating disorders classified under “Other
Parasomnias” in ICSD-2 (74) are commonly seen in women
between the ages of 20 and 30. This disorder consists of recur-
rent episodes of involuntary eating and drinking during
partial arousals from sleep (74,98). Sometimes the patient
Figure 41.30 A 120-second excerpt from a 5½ -
year-old boy showing four episodes of head
banging (a manifestation of rhythmic movement
disorder) at a rate of about 1.5 Hz during transi-
tioning from wakefulness to sleep stage N2. Note
rhythmic movement artifacts in all channels of
PSG. Top 8 channels: EEGs (international 10 to 20
electrode nomenclature); E1–M1 and E2–M1, left
and right EOG; Chin 1–Chin 2, chin EMG; EKG:
electrocardiogram; LTIB and RTIB, EMGs of left
and right tibialis anterior; LGAST and RGAST,
gastrocnemius muscles; OroNs 1–OroNs 2 and
Pflw1–Pflw2, respiratory air flows; Chest and
ABD, effort.
848 Part V ■ Complementary and Special Techniques

Figure 41.31 A 20-year-old woman referred for

sudden nocturnal awakenings with vocalization
and sleepwalking. PSG shows an arousal from
NREM stage 4 sleep with subsequent rhythmic
masticatory muscle activation and teeth gnashing
(EMG artifacts on EEG channels) typical of sleep
bruxism. (Reproduced from Chokroverty S,
Thomas R, Bhatt M, eds. Atlas of Sleep Medicine.
Philadelphia: Elsevier; 2005, with permission.)

displays strange eating behavior (e.g., consumption of inedi-
ble or toxic substances such as frozen pizza, raw bacon, and cat
food). The episodes cause sleep disruption and weight gain,
and occasionally injury has been reported. The condition can
be idiopathic or can be comorbid with other sleep disorders
(e.g., sleep walking, RLS, PLMS, OSAS, narcolepsy, and irreg-
ular sleep–wake circadian rhythm sleep disorder) and with the
use of medications such as zolpidem, triazolam, and other
psychotropic agents (98,99). The most common PSG findings
are multiple, confusional arousals with or without eating,
arising predominantly from stage N3 sleep, but also from
other stages of NREM sleep and occasionally from REM sleep
(see Fig. 41.32) (74).
Catathrenia (Expiratory Groaning)
This parasomnia is characterized by recurrent episodes of expi-
ratory groaning (high-pitched, loud humming or roaring
sounds) occurring in clusters predominantly during REM, but
it may also occur during NREM sleep (100–102). PSG findings
resemble central apnea with protracted expiratory bradypnea
without oxygen desaturation (Fig. 41.33). Simultaneous audio
recordings will bring out the characteristic groaning sounds.
The clinical relevance and pathophysiology of this condition
remain unknown.
Neuromuscular Disorders
In neuromuscular disorders, SDB is commonly associated with
insidiously developing chronic respiratory failure, especially in
the advanced stages, but is often unrecognized and untreated
(103–106).- The most common SDB in neuromuscular disor-
ders is sleep related, especially REM-related hypoventilation.
Both central and upper airway obstructive apneas also occur.
Nocturnal hypoventilation giving rise to hypoxemia and
hypercapnea during sleep in the initial stage of neuromuscular
disorders causes chronic respiratory failure. In later stages of
the illness, abnormal blood gases may persist even during the

Figure 41.32 A 240-second excerpt of PSG showing awakening (right half of the recording)
out of sleep-stage REM (left half of the tracing) and an episode of nocturnal eating in a
51-year-old man. He woke up with a snore from an episode of REM-related hypopnea. Top four
channels: EEGs; left and right electro-oculographic (E1–M1; E2–M1); chin muscle
(CHIN1–CHIN2) show chewing artifacts. EKG: electrocardiogram; HR: heart rate; LtTib and
RtTib, left and right tibialis anterior EMGs; PFlow1–PFlow2, airflow by thermistor (oronasal)
and by nasal pressure recording; CHEST; ABD: respiratory effort; SaO2, oxygen saturation by
finger oximetry.
 Chapter 41
daytime. As a result of alveolar hypoventilation and ventila-
tion–perfusion mismatching, hypoxemia and hypercapnea
occur, giving rise to chronic respiratory failure even during the
daytime at a later stage of neuromuscular disorders. Nocturnal
hypoventilation and chronic respiratory failure, however, in
neuromuscular disorders may present insidiously and may ini-
tially remain asymptomatic (103–106). For the diagnosis, a
high index of clinical suspicion is needed. Historical clues
include the presence of EDS, daytime fatigue, morning
headache, restless and disturbed nocturnal sleep, and unex-
plained leg edema. Patients with neuromuscular disorders
■ Polysomnography: Technical and Clinical Aspects 849
Figure 41.33 Expiratory groans. A polysomno-
graphic segment of 120-second epoch showing
EEGs (C3–A2, O1–A2, C4–A1, O2–A1), left and
right eye movements (LOC–A2; ROC–A1), chin,
left tibialis (LT. TIBI), and right tibialis (RT. TIBI)
EMGs, snoring (SNORE), EKG, oronasal airflow
(ORONAS), thoracic (THORAX) and abdominal
(ABDM) effort channels, oxygen saturation
(SaO2). Note prolonged expiration in the flow
and effort channels followed by arousals without
oxygen desaturation in stage 2 NREM sleep.
(Reproduced from Siddiqui F, Walters AS,
Chokroverty S. Catathrenia: a rare parasomnia
which may mimic central sleep apnea on
polysomnogram. Sleep Med. 2008;9(4):460–461,
with permission.)
manifesting these symptoms deserve to be investigated to
uncover nocturnal hypoventilation to prevent serious conse-
quences of chronic respiratory failure such as pulmonary
hypertension and congestive cardiac failure. Neuromuscular
disorders causing chronic respiratory failure include primary
muscle disease, neuromuscular junction disorders and motor
neuron diseases or motor neuronopathies, acute and chronic
inflammatory demyelinating polyneuropathies, hereditary
sensory-motor neuropathy, and phrenic neuropathy (107).
The characteristic PSG findings in neuromuscular disorders
consist of central, mixed, and upper airway obstructive apneas
Figure 41.34 Ten- and 30-second excerpts from
a nocturnal PSG showing alpha–delta sleep in a
30-year-old man with history of snoring for many
years. He denied any history of joint or muscle
aches and pains. The alpha frequency is inter-
mixed with and superimposed on underlying
delta activity. Alpha–delta sleep denotes a non-
specific sleep architectural change noted in many
patients with complaints of muscle aches and
fibromyalgia. It is also seen in other conditions
and many normal individuals. Top 10 channels:
EEGs; LEFT EOG and RT EOG, left and right
electro-oculograms; chin EMG, EMG of chin;
Lt/ Rt Tib EMG, left/ right tibialis anterior EMG;
oronasal thermistor; chest and abdomen effort
channels; snore monitor; EKG, electrocardiogra-
phy. (Reproduced from Chokroverty S, Thomas
R, Bhatt M, eds. Atla s of Sleep Medicin e.
Philadelphia: Elsevier; 2005, with permission.)
850 Part V ■ Complementary and Special Techniques
and hypopneas associated with oxygen desaturation, sleep
hypoventilation showing nonapneic oxygen desaturation that
is worse during REM sleep (108). The painful neuromuscular
disorders may also show sleep-onset insomnia and reduced
sleep efficiency.
Neurodegenerative Disorders
Neurodegenerative diseases of the central nervous system
include a group of heterogeneous disorders with insidious
onset and relentlessly progressive course without an identifi-
able cause except for genetic predisposition in some cases
(109). The other characteristics include bilateral symmetry of
the clinical manifestations and increasing incidents with
advancing age and pathologic involvement of system of neu-
rons that are anatomically and physiologically related with
common pathologic features of simple atrophy or apoptosis
without intense cellular reactions but accompanied by mild
gliosis. The concept of neurodegenerative disease has changed
rapidly with the modern scientific understanding of the
molecular neurobiologic mechanisms responsible for nerve
cell degeneration. In contemporary neuroscience, neurodegen-
erative diseases are believed to be proteinopathies due to exces-
sive protein misfoldings and intracellular protein aggregation.
They are mainly classified into tauopathies and synucle-
inopathies. The main tauopathies (Table 41.5) include
Alzheimer disease (AD), progressive supranuclear palsy (PSP),
corticobasal degeneration (CBD), and frontotemporal demen-
tia (FTD). Synucleinopathies are a group of disorders with
abnormal deposition of alpha-synuclein in the cytoplasm of
neurons or glial cells as well as extracellular deposits of amy-
loid, and these include PD, diffuse Lewy body disease (DLBD)
with dementia, and multiple system atrophy (MSA), which is
inclusive of Shy–Drager syndrome, striatonigral degeneration,
and sporadic olivopontocerebellar atrophy (see Table 41.5).
The major sleep complaints in neurodegenerative diseases
include insomnia, hypersomnia, parasomnia, excessive noctur-
nal motor activity, “sundowning,” circadian sleep–wake distur-
bances, and respiratory dysrhythmias.
Sleep fragmentation characterized by increased awakenings,
stage transitions, and arousals is an important sleep dysfunc-
tion in neurodegenerative diseases. Sleep fragmentation results
in reduced slow wave and REM sleep. Hypersomnia includes
EDS and irresistible sleep attacks. The parasomnia most com-
monly associated with neurodegenerative diseases is RBD,
which is characterized by REM sleep associated with intense
motor activity related to dream-enacting behavior and absence
of muscle atonia of REM sleep. It has been increasingly obvious
recently that in the setting of degenerative dementia or
Parkinsonism, RBD is not only a manifestation of an evolving
synucleinopathies but also seen in tauopathies occasionally.
RBD carries the potential for injury to self and others and
therefore its early recognition is very important. Circadian
rhythm sleep–wake disturbances are noted in some conditions
most prominent in AD. It may also present as a cyclic agitation
syndrome popularly known as the “sundowning syndrome.” A
commonly encountered excessive nocturnal motor activity that
may cause sleep disturbance is PLMS.
Tabl e 41 . 5
Molecular Neurobiologic Classification of
Neurodegenerative Diseases
Tauopathies
Alzheimer disease (AD)
Progressive supranuclear palsy (PSP)
Corticobasal degeneration (CBD)
Frontotemporal dementia (FTD)
Synucleinopathies
Parkinson disease (PD) with or without dementia
Diffuse Lewy body disease (DLBD) with dementia
Multiple system atrophy (MSA)
Polyglutamine triplet repeat disorders
Huntington disease (HD)
Spinocerebellar ataxia (SCA)
Hypoprogranulinopathies
FTD with parkinsonism linked to progranulin gene on
chromosome 17 (FTDR-17P)
Miscellaneous diseases
Torsion dystonia
Chorea-acanthocytosis
Amyotrophic lateral sclerosis (ALS)
Reproduced from Chokroverty S. Sleep and neurodegenerative diseases. Semin
Neurol. 2009;29:446–468.
Polysomnographic Findings in
Neurodegenerative Disorders
PSG is the single most important laboratory test for the diagno-
sis of sleep disorders particularly hypersomnia. Simultaneous
video monitoring (Video PSG) additionally helps in the diag-
nosis of parasomnia (109).
PSG findings in AD show reduced total sleep time, increased
nocturnal awakenings, decreased REM and slow-wave sleep,
loss of phasic components (spindles and K-complexes) of
NREM sleep, and sleep–wake rhythm disturbances. Other
abnormalities in PSG of AD patients are sleep apnea and PLMS.
In PD, PSG findings include reduction of total sleep time,
frequent awakenings, decreased slow-wave sleep, decreased sleep
spindles, and K-complexes in stage N2. REM sleep may intrude
into NREM sleep and normal REM/NREM cycling may be
abolished. Repetitive blinking may be noted at sleep onset and
blepharospasm is seen during slow-wave sleep before REM
episodes. Recently, there has been PSG documentation of
abrupt daytime sleep attacks in PD patients characterized by
abrupt occurrence of stage N2 sleep. Absence of muscle atonia
in REM sleep may be noted in patients with RBD (see
Fig. 41.26) or as a preclinical sign of RBD.
 Chapter 41
In DLBD, PSG findings include decreased total sleep time,
frequent awakenings, and absence of muscle atonia in REM
sleep in those with RBD, which may be a preclinical sign of
DLBD. Because of sleep fragmentation and repeated awaken-
ings, the patient may have EDS as reflected in abnormal MSLT.
PSG findings in PSP patients have shown consistent abnor-
malities in sleep architecture with the decreased total sleep
time, decreased sleep efficiency, increased wake time after sleep
onset, increased percentage of N1 sleep, decreased percentage of
REM sleep, and decreased REM duration. Another characteris-
tic is the reduction in abundance and amplitude of sleep spin-
dles. RBD has also been described in PSP patients (110,111).
PSG findings in Huntington disease show decreased sleep effi-
ciency, increase in stage N1, frequent arousals, and reduced
slow-wave sleep. The involuntary movements decrease progres-
sively from wake to stages N1 and N2 to REM sleep. The dysk-
inetic movements in sleep are also of shorter duration, lower
amplitude, and more fragmented as compared to wakefulness.
An increased density of sleep spindles in Huntington disease
patients has also been reported.
PSG findings in torsion dystonia show frequent awakenings,
reduced sleep efficiency, and increased sleep latency. High-
amplitude sleep spindles have been described in some cases but
have been an inconsistent finding in other studies. Segawa and
colleagues (112) recorded movement counts in PSG studies in
patients with hereditary progressive dystonia. They reported a
decrease in gross body movements in stage N1, an increase in
stage N2, and a decrease in REM sleep. Localized twitch move-
ments were depressed in all-sleep stages but followed the nor-
mal relative distribution between stages.
In olivopontocerebellar degeneration (OPCD), PSG study has
been reported to show increased awakenings, reduced slow-wave
sleep, and reduced or absent REM sleep (113,114). In several
cases of OPCD, REM sleep without muscle atonia accompanied
by features of RBD has been described (115). Central, obstruc-
tive, and mixed apneas have been reported by several authors.
In MSA, PSG findings include reduction of total sleep time,
decreased sleep efficiency, decreased sleep efficiency, increased
sleep latency, increased night-time awakenings, decreased slow-
wave sleep and REM sleep, absence of muscle atonia in patients
with RBD, and a variety of respiratory dysrhythmias (109).
Insomnia Including Fatal Familial Insomnia
Insomnia is the most common sleep disorder affecting the pop-
ulation and the most common disease encountered in the prac-
tice of sleep medicine. Patients present with difficulty initiating
and maintaining sleep and nonrestorative sleep occurring three
to four times per week persisting for more than a month and
associated with impairment of daytime function (74). Acute
insomnia is generally associated with an identifiable stressful
situation. Most cases of insomnia are chronic and comorbid
with other conditions including psychiatric, medical, and neu-
rologic disorders and drug and alcohol abuse. If no cause is
found, the condition is labeled idiopathic, primary, or psy-
chophysiological insomnia. There is no indication for PSG in
uncomplicated insomnia unless insomnia is associated with
sleep apnea. Some nonspecific findings have been described in
■ Polysomnography: Technical and Clinical Aspects 851
patients with nonrestorative sleep such as alpha–delta sleep
(Uncommon and atypical PSG patterns). Fatal familial insomnia
(FFI) is a rare prion disease originally described by Lugaresi and
associates (116). Clinical features of FFI include impaired control
of sleep–wake cycle, including circadian rhythms, autonomic and
endocrine dysfunction, as well as somatic, neurologic, cognitive,
and behavioral manifestations (117). Profound sleep distur-
bances and in particular severe insomnia are noted from the very
beginning of the illness. In the PSG study, there is almost total
absence of deep sleep patterns and only short episodes of REM
sleep lasting for a few seconds or minutes without muscle atonia
associated with dream-enacting behavior in the form of complex
gestures and myoclonus. There is progressive reduction of the
total sleep time and reduced sleep cycle with the progression of
the disease. Sleep spindles and K-complexes are totally absent in
the late stage of illness and slow-wave sleep is never recorded. The
neurophysiology hallmark of FFI is severe atrophy of the thala-
mus, particularly anteroventral and dorsomedial thalamic nuclei.
FFI has thus rekindled the role of the thalamus in the sleep–wake
regulatory mechanisms by demonstrating that the thalamus is
essential for generation of slow-wave sleep and for suggesting a
role for prions in sleep regulation.
Uncommon and Atypical PSG Patterns
In certain normal individuals, particularly following sleep dep-
rivation, withdrawal of medications, and alteration of sleep
habits, certain atypical and uncommon sleep patterns may be
observed in the PSG (118). Similar patterns, however, have also
been described in a variety of sleep disorders. Some of these pat-
terns include alpha–delta or alpha–NREM sleep, REM–spindle
sleep, REM intrusion into NREM sleep, REM sleep without
atonia in conditions other than RBD, abnormal REM–NREM
cycling, rhythmic leg movements, bruxism with arousals after
termination of apnea, alternating leg muscle activation (ALMA),
propriospinal myoclonus at sleep onset, and ataxic (biot’s)
breathing in patients on opiates for pain management.
Alpha–Delta Sleep
Hauri and Hawkins (119) originally described the so-called
alpha–delta sleep pattern in patients with depression.
However, since then it has been found in many subjects with
nonrestorative sleep as a result of muscle and joint aches and
pains, and rheumatic disorders including fibromyalgia. A sim-
ilar pattern has also been noted in many normal individuals.
Therefore, it is not specific for fibromyalgia but often noted in
large number of such patients. In this PSG pattern, bursts of
alpha activities are seen during slow-wave sleep or during stage
N2 sleep overriding the slow waves or spindles causing alpha
intrusions into different NREM sleep stages (see Fig. 41.34).
REM–Spindle Sleep
On many occasions bursts of sleep spindles are noted during
REM sleep, giving rise to a pattern called REM–spindle sleep
pattern (Fig. 41.35). This is noted in many normal individuals,
particularly in the beginning of the first REM period. Similar
REM–spindle sleep pattern is noted also in many patients with
hypersomnia and narcolepsy–cataplexy syndrome.
852 Part V ■ Complementary and Special Techniques

Figure 41.35 A 67-year-old man with difficulty

sleeping, loud snoring, and EDS for 5 years.
Medical history is significant for high blood pres-
sure and coronary artery disease. Nocturnal PSG
showed the presence of moderate sleep apnea
with an AHI of 29.5. A 30-second excerpt from
nocturnal PSG showing the presence of frequent
sleep spindles throughout the epoch of REM
sleep. Prominent sawtooth waves of REM sleep
in C3- and C4-derived EEG channels, prominent
phasic eye movements of REM sleep on electro-
oculogram channels, and decreased chin muscle
tone characteristic of REM atonia are seen.
(Reproduced from Chokroverty S, Thomas R, Bhatt
M, eds. Atlas of Sleep Medicine. Philadelphia:
Elsevier; 2005, with permission.)

REM Sleep Without Muscle Atonia
This pattern is of course characteristically seen in patients with
RBD (see Fig. 41.26). However, sometimes similar REM sleep
without atonia may be noted in patients who have been treated
with REM suppressant medications such as tricyclic antidepres-
sant, monoamine oxidase (MAO) inhibitors, selective serotonin
reuptake inhibitors, and phenothiazines, as well as in patients
with PD or narcolepsy in absence of the typical behavioral
manifestation of RBD. Whether this pattern indicates a subclin-
ical or preclinical RBD remains undetermined.
A Mixture of REM Bursts in NREM Sleep
Occasionally, REM-like bursts are noted during NREM sleep in
normal persons. Such pattern, however, is seen particularly in
narcoleptics and patients with depression who have been
treated with REM suppressant medications. This pattern may
also be seen in PD.
Abnormal REM–NREM Cycling
REM–NREM goes through a normal cycling of four to five
cycles throughout the night. The initial cycle is of brief duration
and the longest period of REM cycle is noted toward the end of
the night when it may last up to 1 hour. Such cycling may be
disrupted in patients with narcolepsy–cataplexy and those with
narcolepsy who have been on REM suppressant medications.
Such abnormal cycling may also be seen in patients with head
injury and PD. In addition, some normal individuals may dis-
play an abnormal REM–NREM cycling.
Rhythmic Limb and Body Movements on Arousal
After Termination of Apneic–Hypopneic Events
Rhythmic movement disorder is an abnormal sleep-related
movement disorder characterized by rhythmic oscillations of
the head, trunk, or limbs during wake–sleep transition (74). As
an unusual manifestation, this also has been noted in patients
with upper airway OSAS on arousals after termination of
apneic–hypopneic events (120–122) (Fig. 41.36).
Rhythmic Leg Movements
Rhythmic leg movements in wakefulness (Fig. 41.37) and
NREM (Fig. 41.38) and REM (Fig. 41.37) sleep are sometimes
noted as unusual PSG findings in normal individuals as well as
in patients with sleep pathologies (121). The significance of
such findings remains undetermined.
Sleep Bruxism or Tooth Grinding
As an unusual manifestation, bruxism may rarely be noted on
arousal after termination of apneic events in patients with
OSAS (121) (Fig. 41.39).
Alternating Leg Muscle Activation
ALMA is a recently described unusual polysomnographic
observation characterized by ALMA preceding or following
an arousal or awakening (122,123) (Fig. 41.40). The usual
frequency is 1 to 2 Hz (range of 0.5 to 3Hz), with a duration
of 0.1 to 0.5 second for individual activations, and showing
at least four discrete and alternating muscle activations (12).
The specific significance of ALMA remains undetermined at
present, but it may be a variant of hypnagogic foot tremor.
Propriospinal Myoclonus in Predormitum
Propriospinal myoclonus has recently been described (124,125)
as axial jerks occurring during relaxed wakefulness immediately
before falling asleep (Fig. 41.41), the period termed predormitum
by McDonald Critchley (126). The etiology remains undeter-
mined but may prevent from falling asleep, causing sleep-onset
 Chapter 41 ■ Polysomnography: Technical and Clinical Aspects 853

Figure 41.36 A 90-second excerpt from a portion of an overnight polysomnography during

stage 2 (N2) NREM sleep in a 51-year-old man with upper airway OSA, RLS, and rhythmic
movement disorder showing two episodes of rhythmic EMG bursts (about 0.75 Hz) beginning
in the right biceps (RtBiceps) and right triceps (RtTriceps) and spreading to other muscles.
These bursts were noted following termination of the apneic events and accompanied by
arousals and resumption of normal breathing as well as rhythmic body and head-rolling move-
ments. Top six channels (international nomenclature): EEGs; E1–M1, left electro-oculogram;
E2–M1, right electro-oculogram; EKG, electrocardiogram; OralNasal, oronasal thermistor (air
flow); PFlow, nasal pressure recording for airflow; Chest, thoracic breathing effort; Abd, abdom-
inal breathing effort; SaO2, oxygen saturation by finger oximetry; Snore, snoring recording;
Masseter, right masseter EMG; Chin: submental EMG; RtSterno: right sternocleidomastoideus
EMG; LtBiceps, left biceps EMG; LtTriceps, left triceps EMG; RtBiceps, right biceps EMG;
RtTriceps, right triceps EMG; RtQuad, right quadriceps (biceps femoris) EMG; LtQuad, left
quadriceps EMG; RtTib, right tibialis EMG; RtGast, right gastrocnemius EMG; LtTib, left tibialis
EMG; LtGast, left gastrocnemius EMG. (Reproduced from Gharagozlou P, Seyffert M, Santos R,
et al. Rhythmic movement disorder associated with respiratory arousals and improved by CPAP
titration in a patient with restless legs syndrome and sleep apnea. Sleep Med.
2009;10(4):501–503, with permission.)

Figure 41.37 A 50-year-old man with history of

loud snoring, choking in sleep, and intermittent
leg jerking at night. Nocturnal PSG shows the
presence of mild sleep apnea with an AHI of
12.3. This 30-second excerpt from nocturnal
polysomnography shows bursts of rhythmic foot
and leg movements on the left/ right tibialis ante-
rior muscle recording channel during wakeful-
ness. The movement is not associated with
respiratory events, oxygen desaturation, or
arousal from sleep. EEG, Top 10 channels: EEGs;
LEFT and RT EOG, left and right electro-oculo-
grams; CHIN, EMG of chin; Lt/ Rt Tib EMG,
left/ right tibialis anterior EMG; oronasal thermis-
tor; chest and abdomen effort channels; snore
monitor; EKG, electrocardiography; SaO2, oxygen
saturation by finger oxymetry.
854 Part V ■ Complementary and Special Techniques

Figure 41.38 A 50-year-old man with history of

loud snoring, choking in sleep, and intermittent
leg jerking at night. Nocturnal PSG shows the
presence of mild sleep apnea with an AHI of
12.3. This 30-second excerpt from nocturnal
polysomnography shows bursts of rhythmic foot
and leg movements on the left/ right tibialis ante-
rior muscle recording channel during REM sleep.
The movement is not associated with respiratory
events, oxygen desaturation, or arousal from
sleep. Top 10 channels: EEGs; LEFT and RT EOG,
left and right electro-oculograms; CHIN, EMG of
chin; Lt/ Rt Tib EMG, left/ right tibialis anterior
EMG; oronasal thermistor; chest and abdomen
effort channels; snore monitor; EKG, electrocar-
diography; SaO2, oxygen saturation by finger
oxymetry.

insomnia. The generator is thought to be in the mid-thoracic
spinal cord with propagation of muscle bursts rostrally and cau-
dally along slowly conducting propriospinal pathways.
Ataxic or Biot’s Breathing in Patients on Opiates
Ataxic breathing is a type of periodic breathing with central
apnea without crescendo–decrescendo pattern resulting from
an unstable or destabilized central controller for breathing
(medullary respiratory neurons). This is well known to occur in
medullary lesions, but this has recently been observed in
patients taking large doses of opiates for pain management
(127,128) (Fig. 41.42). Biot’s breathing is a type of ataxic
breathing with two to three normal breaths interspersed with
periodically recurring central apnea.

Figure 41.39 A 120-second excerpt from PSG

recording reveals sleep bruxism in a 70-year-old
woman with a history of insomnia, early morn-
ing awakenings, and EDS. Overnight PSG
revealed the presence of mild to moderate OSA
with an AHI of 21.5. Episodes of bruxism are
recorded repeatedly as part of the arousal
response following respiratory events accompa-
nied by tooth grinding on simultaneous
audio–video recording. Respiratory-related limb
movements are recorded on tibialis anterior
channel in association with the arousal response.
PSG sleep bruxism is characterized by a series of
teeth grinding associated with rhythmic EMG
artifacts with a frequency of approximately 0.5 to
1.0 hertz in stage 1 NREM sleep. Top 10 chan-
nels: EEGs; LEFT and RT EOG, left and right elec-
tro-oculograms; CHIN, EMG of chin; Lt/ Rt Tib
EMG, left/ right tibialis anterior EMG; oronasal
thermistor; chest and abdomen effort channels;
snore monitor EKG, electrocardiography.
(Reproduced from Chokroverty S, Thomas R,
Bhatt M, eds. Atla s of Sleep Medicin e.
Philadelphia: Elsevier; 2005, with permission.)
 Chapter 41
RELATED LABORATORY PROCEDURES
FOR DIAGNOSIS OF SLEEP DISORDERS
PSG and video-PSG recordings remain the most important lab-
oratory tests for diagnosis of sleep disorders. Three other tech-
nical procedures, however, have found an important place in
the assessment of sleep disorders: MSLT, MWT, and actigraphy.
Additionally, there are two other procedures (129,130) that also
have some place in assessment of sleep disorders and these
include peripheral arterial tonometry (PAT) and pulse transit
time (PTT), but these are not described in this chapter.
Multiple Sleep Latency Test
The most common indication for MSLT is EDS. The initial step
in assessment of a patient with EDS is a detailed sleep history
covering not only sleep at night but also daytime function as
■ Polysomnography: Technical and Clinical Aspects 855
Figure 41.40 A 20-second except of PSG show-
ing alternating leg muscle activation (ALMA) in
the left and right tibialis anterior muscles (LTIB
and RTIB) at a state of approximately 1.5 Hz in
an adult man during stage N2 sleep. Top 8 chan-
nels: EEGs; E1–M1 and E2–M1, left and right
electro-oculograms; Chin 2–Chin 3, chin muscle
EMG; EKG, electrocardiogram; HR, heat rate;
OroNs1–OroNs2 and Pflw1-Pflw2, respiratory
airflow; Chest and ABD, respiratory Effort; SaO2,
oxygen saturation by finger oximetry; EtCO2,
expiratory carbon dioxide.
well as other history and physical examination. The first step
before ordering the laboratory tests is an assessment of EDS.
There are certain clinical clues that may suggest EDS (131):
falling asleep in inappropriate places and inappropriate circum-
stances, for example, sleepiness while sitting and relaxing on a
couch, watching television, reading, watching movies, sitting in
classes or conferences, listening to a lecture and in very severe
cases even talking on the telephone and giving the history to the
physician; dosing off while driving; poor attention span and
difficulty coping with work and school work; driving accidents
or accidents at work; frequent naps during the day and nonre-
freshing sleep; and feeling sleepy and tired upon waking up first
thing in the morning. For assessment of persistent sleepiness,
Epworth Sleepiness Scale (ESS) (132) is often used to assess a
general level of sleepiness. This is a subjective propensity to
sleepiness assessed by the patient under eight situations on a
Figure 41.41 Propriospinal myoclonus at the
wake–sleep transition. A 40-year-old man pre-
sented with a 4-year history of axial jerks during
relaxed wakefulness impending falling asleep.
PSG recording shows repetitive myoclonic axial
jerks. The EMG activity originates in the left rec-
tus abdominis muscle, thereafter propagating to
rostal (sternocleidomastoid, masseter, mylohy-
oid) and caudal (biceps femoris) muscles. The
left panel shows jerks at low speed; the right
panel shows one of the jerks at a high speed.
R, right; L, left; Mylohyoid, mylohyoideus;
S.C.M., sternocleidomastoideus; I.T.C., inter-
costalis; Rectus abd, abdominis; Rectus fem., rec-
tus femoris; Biceps fem, biceps femoris.
(Reproduced from Chokroverty S, Thomas R,
Bhatt M, eds. Atla s of Sleep Medicin e.
Philadelphia: Elsevier; 2005, with permission.)
856 Part V ■ Complementary and Special Techniques
Figure 41.42 A 240-second excerpt of PSG dur-
ing sleep stage N2 showing periodic breathing
(ataxic or Biot’s breathing with two to three nor-
mal breaths interspersed with periodically recur-
ring central apneas) in a 68-year-old woman on
large doses of OxyContin (opiate) for pain man-
agement. Her abnormal breathing improved
somewhat on bilevel positive upper airway pres-
surization but could not eliminate the events
most of the time. Top four channels, EEGs;
E1–M1 and E2–M1, left and right EOGs;
CHIN1–CHIN2, chin EMG; EKG, electrocardio-
gram; HR, heart rate; LtTib and RtTib, left and
right tibialis anterior EMG (); CFLOW, air flow
during positive pressure titration; CHEST and
ABD, respiratory effort; CPRES; positive pressure
in centimeter; SaO2, oxygen saturation.
scale of 0 to 3 with three indicating the chances of dosing off are
at the highest. The maximum score is 24 and a score of 10 or
above suggests the presence of EDS. The test has been weakly
correlated with the MSLT score. The ESS and MSLT, however,
test different types of sleepiness. MSLT tests the propensity to
sleepiness objectively and ESS tests the general feeling of sleepi-
ness or subjective propensity to sleepiness. The Stanford
Sleepiness Scale (SSS) (133) is a 7-point scale to measure sub-
jective sleepiness at the time of testing, but it does not measure
persistent sleepiness. There is a somewhat similar scale called
Karolinska Sleepiness Scale (KSS) (134), which is used mostly
in Europe using 9-point scale but again does not measure per-
sistent sleepiness. Visual Analog Scale is the other scale used to
assess alertness and well-being in which subjects indicate their
feelings of alertness at an arbitrary point on a line of 0 to 100
mm scale with 100 being the maximum sleepiness and 0 being
the most alert.
Technique of MSLT
The MSLT has been standardized and includes several general
and specific procedures (135,136). The general procedures
before the actual recording include keeping a sleep diary for 1
to 2 weeks before the test, which records the information about
bedtime, time of rising, napping, and any drug use. The patient
should discontinue central nervous system active drugs before
the test. The test is preceded by an overnight PSG study and is
scheduled about 2 to 3 hours after the conclusion of the PSG
study. The actual test consists of four to five nap opportunities
at 2-hour intervals and each recording session lasts for 20 min-
utes. Between sessions, test subjects must remain awake. The
subject must not smoke for 30 minutes before lights are turned
out. Physiologic calibrations (e.g., grit teeth, blink your eyes,
look up, look down, look to the right, look to the left, open your
eyes, and close your eyes) are then performed and the patients
are instructed to relax and fall asleep, and the lights are turned
out. The test must be conducted in a quiet, dark room. The
recordings may include three to six channels of EEG, submen-
tal EMG, and EOG recordings.
The measurements include average sleep onset latency and
the presence of SOREMs, defined as the occurrence of REMs
within the first 15 minutes of sleep onset. Sleep onset is defined
as the time from lights out to the first epoch of any stage of
sleep including stage N1 sleep. The first epoch of 15 seconds of
sleep or greater in a 30-second epoch is considered sleep onset.
If no sleep occurs, the test is concluded 20 minutes after lights
out. The test is terminated 15 minutes after the first 30-second
epoch in any stage of sleep. If the finding is indefinite, it is bet-
ter to continue the test than to end it prematurely. Mean sleep
latency is calculated as the average of the latencies to sleep onset
for each of the four to five naps. A sleep latency of less than 8
minutes is consistent with pathologic sleepiness. A mean sleep
latency of 8 to 10 minutes is consistent with mild sleepiness and
a mean sleep latency of 10 to 15 minutes is considered normal.
Repeat MSLT is indicated if the patient is strongly suspected to
have narcolepsy but did not show the characteristic findings of
pathologic sleepiness, with sleep-onset latency of less than 8
minutes and the presence of two or more SOREMs during four
to five tests as may be seen in certain percentage of narcolepsy
patients after the first MSLT. MSLT may not be diagnostic in the
initial test and the yield increases after the second test. The
other situation for repeating MSLT is when the finding is
ambiguous and the sleep onset or REM sleep cannot be ade-
quately interpreted. Finally, if the MSLT guidelines have not
been followed, the test results may be technically invalid.
Indications for MSLT
The AASM developed indications and practice parameters for
clinical use of the MSLT (136). The single most important indi-
cation for performing MSLT is for the evaluation of patients
with suspected narcolepsy. The MSLT may also be useful in the
 Chapter 41
evaluation of patients with suspected idiopathic hypersomnia
in which condition the patient develops pathologic sleepiness
or sleep onset of less than 8 minutes with one or no SOREM
during 4 to 5 nap tests. The MSLT is not routinely indicated in
the initial evaluation and diagnosis of OSAS nor for the moni-
toring of the treatment effects after nasal CPAP therapy. The
MSLT is also not routinely indicated for patients complaining
of EDS due to medical or neurologic disorders except nar-
colepsy, insomnia, or circadian rhythm sleep disorders.
Reliability, Validity, and Limitations of the MSLT
The sensitivity and specificity of the MSLT in detecting sleepi-
ness have not been clearly determined (137). The test–retest
reliability of the MSLT, however, has been documented in both
normal subjects and patients with narcolepsy. In subjects with
sleepiness caused by circadian rhythm sleep disorders, sleep
deprivation, and ingestion of alcohol, pathologic sleepiness has
been validated by MSLT. There is, however, poor correlation
between the MSLT and ESS. The patient’s psychological and
behavioral states also interfere with the MSLT results. The
MSLT objectively measures tendency to sleep rather than the
likelihood of falling asleep. If the patient suffers from severe
anxiety or psychological disturbances causing behavioral stim-
ulation, the MSLT may not show sleepiness even in a patient
complaining of EDS.
The Maintenance of Wakefulness Test
The MWT is a variant of the MSLT to measure a patient’s abil-
ity to stay awake (138,139). The MWT is performed similar to
MSLT at 2-hour intervals in a quiet dark room beginning about
1
12 to 3 hours after waking up. Four to five such tests are per-
formed and each one lasts for 40 minutes. A polysomnogram
and sleep logs are not required before the MWT. The patient
should be seated in bed with the back and head supported by a
bed rest, ensuring that the neck is comfortable. The patient
should not smoke at least 30 minutes before each test and
should not have caffeinated beverages on the day of the test.
The conventional recording montage for the MWT is similar to
that used for the MSLT. At the start of the test, after lights out,
the patient is instructed to sit still and remain awake for as long
as possible. The patient is not allowed to use extraordinary
measures to stay awake. Unequivocal sleep is defined as three
consecutive epochs of stage N1 sleep or one epoch of any other
stage of sleep. Trials are ended after 40 minutes if no sleep
occurs and also after sleep onset as defined above.
Clinical Indications, Normative Data,
Validity, and Reliability
The MWT 40-minute protocol may be indicated in assessment
of individual’s ability to remain awake when his/her inability to
remain awake constitutes a public or personal safety issue
(136). The MWT is also indicated in patients with excessive
sleepiness to assess response to treatment (e.g., response to
stimulants in narcolepsy and idiopathic hypersomnia and
response to CPAP titration in OSA patients). The MWT is less
useful and less sensitive than the MSLT as a diagnostic test for
narcolepsy. It should, however, be noted that the MSLT
■ Polysomnography: Technical and Clinical Aspects 857
and MWT do have separate functions. The MSLT unmasks
physiologic sleepiness, which depends on both the circadian
and homeostatic factors. In contrast, the MWT is a reflection of
the individual’s capability to resist sleep and is influenced by
physiologic sleepiness. The validity, reliability, and specificity of
MWT need to be studied using large samples and different pop-
ulations of sleep disorders. No universally accepted normative
data for the MWT are available. Using standard deviation crite-
ria, lower limit of normative data for sleep onset latency (mean
2 SD) has been defined by Mitler et al. (140) as 19.4 minutes.
In contrast, the AASM practice parameters (136) advocate the
use of a percentile cut-off score. Accordingly a mean sleep
latency of less than 8 minutes on the 40-minute protocol is con-
sidered abnormal and values greater than this but less than 40
minutes are of uncertain significance. It should be noted that
there is little evidence linking mean sleep latency on the MWT
with the risk of accidents in real-world circumstances, therefore
the sleep clinician should not rely exclusively on mean sleep
latency for determining the risk of accidents but should take
into consideration clinical history, physical findings, and com-
pliance with treatment. An important study by Philip et al.
(141) correlated subjective and objective measures of sleepiness
and driving performance in patients suffering from EDS. The
study included 38 untreated sleep apnea patients and 14 healthy
controls. Based on the number of inappropriate line-crossings
during a 90-minute real-life driving session, they determined
the MWT sleep-onset score as follows: very sleepy, 0 to 19 min-
utes; sleepy, 20 to 33 minutes; and alert, 34 to 40 minutes.
Actigraphy
Colburn et al. (142) first documented use of accelerometers to
record body movements. Tyron (143) in 1991 gave a compre-
hensive account of activity monitoring in psychology and med-
icine. Actigraphy is defined as the technique of recording and
quantifying movements. Monitoring of body movements and
other activities can be performed continuously for days, weeks,
or even months by using an actigraph, also known as actometer
or actimeter (144,145). This can be worn on the wrist or alter-
natively on the ankle for recording arm, leg, and body move-
ments. Actigraph uses piezoelectric sensors which generate
electric currents in direct proportion to the amount of acceler-
ation. The activity usually covers frequency range of 0.5 to 15
Hz, which is generally the dominant range for human move-
ments. The mechanical movements are converted into electrical
signals that are then sampled every tenth second over a prede-
termined time or epoch, which is generally one minute. These
samples are stored in the memory of the actigraph and then
retrieved and analyzed in a computer. The principle of analysis
is based on the fact that increased movements (as indicated by
black bars in the actigraphy) are seen during wakefulness in
contrast to markedly decreased movements or no movements
(as indicated by the white bars interrupting the black bars dur-
ing sleep), although normal physiologic body and limb move-
ments and postural shifts during sleep will cause interruptions
(small black bars) of the white background (Fig. 41.43). Several
actigraph models are in the development stage to carefully reg-
ulate the sampling frequencies, durations, filters, sensitivities,
858 Part V ■ Complementary and Special Techniques
Figure 41.43 Normal sleep–wake schedule.
This shows a wrist actigraph recording from a
healthy 60-year-old man with a regular
sleep–wake schedule. The subject goes to bed
around 10:00 PM and wakes up at 6:00 AM.
Physiologic body shifts and movements are indi-
cated by a few black bars in the white (sleep
period) areas. The waking period is indicated by
black bars reflecting activities.
and the dynamic range in order to detect and quantify PLMS,
but no generally accepted standardized technique of quantify-
ing and identifying PLMS discriminating from other move-
ments (e.g., those resulting from parasomnia, nocturnal
seizures, and other dyskinesias) is currently available. The role
of actigraphy in detecting, quantifying, and differentiating
abnormal motor activities currently remains controversial, but
there is immense potential for such applications with the devel-
opment of sophisticated models and techniques. More recently,
in some models, sophisticated systems for counting movements
have been developed and validated (146,147). This system
therefore may be useful for therapeutic monitoring and for
diagnostic ability in patients with PLMS and RLS (148,149).
One of the available monitors (PAM-RL, Respironics,
Pittsburgh, PA) through a fine-grained analysis with 40-Hz
sampling and storage at 10 Hz provides a description closely
matching the EMG recordings for these movements. The data
provide an excellent agreement with the EMG recordings of
the legs (Fig. 41.44) and with the nocturnal PSG for number
of leg movements (150). Studies have been conducted using
actigraphy and PSG recordings simultaneously to validate the
ability of the actigraph to distinguish sleep from wakefulness.
Assessment of sleep–wakefulness for over a prolonged period
of time (days to weeks) has been the main utility of actigraph.
Total sleep time and sleep-onset time, however, have shown
variable results with discrepancy from PSG results. Although
computer algorithms are available for automatic sleep–wake
scoring, visual inspection of the raw data is necessary.
Rreliable and valid data are obtained using a specific actigraph
model, but no universally validated data are available.
Actigraphs can differentiate sleep from wakefulness but can-
not differentiate REM from NREM sleep and cannot identify
different NREM sleep stages. Actigraphs and sleep logs are
complementary.
Recommendations for Actigraph
The AASM’s Standards of Practice Committee suggests the fol-
lowing recommendations for actigraphy. Actigraphy may be a
useful adjunct to history and physical examination and sleep
logs in patients with insomnia including sleep state mispercep-
tion (paradoxical insomnia) and inadequate sleep hygiene
(Figs. 41.45 and 41.46), circadian rhythm sleep disorders (151)
(see Figs. 41.47 and 41.48), and excessive daytime somnolence.
Actigraphy may be a useful adjunct to detect the rest time and
activity patterns during modified portable sleep apnea testing.
Actigraphy may be useful in assessing EDS in situations where
MSLT is not practical.
Actigraphy is useful to detect rest time and activity patterns
over days and weeks when sleep log or other methods cannot
provide such data.
Actigraphy may be useful in monitoring circadian rhythm
disturbances in special populations (e.g., the elderly and nurs-
ing home patients); newborns, infants, children, and adoles-
cents; hypertensive individuals (to monitor circadian patterns
of blood pressure and circadian therapeutic effectiveness of
medications); depressed or schizophrenic patients; and individ-
uals during space flight.
Actigraphy may be useful to measure outcomes in clinical
trials. Actigraphs are not indicated for the diagnosis, assess-
ment, or severity of any sleep disorder including insomnia
and OSAS.
Advantages of Actigraph over PSG
These include the following: easy accessibility; inexpensive
recording over extended periods (e.g., days, weeks, or even
months); recording of 24-hour activities at all sites: home, work,
or laboratories; usefulness in uncooperative and demented
patients when laboratory PSG study is not possible; ability to
conduct longitudinal studies to monitor disease progression,
 Chapter 41 ■ Polysomnography: Technical and Clinical Aspects 859

Figure 41.44 Example of the real-time output of

a high-precision activity monitor worn on the
ankle (bottom line) compared to anterior tibialis
EMG activity (top line). The middle line shows
the real-time automatic detection of a significant
leg movement made by the activity meter. The
decision rules for the real-time leg movement
detector create a 7-second delay in the detection.
(Reproduced from Chokroverty S. Sleep Disorders
Medicine: Basic Science, Technical Considerations
a nd Clinica l Aspects. 3rd ed. Philadelphia:
Saunders/ Elsevier; 2009, with permission.)

Figure 41.45 Actigraphy in insomnia (sleep state misperception): a 59-year-old man complain-
ing of insomnia since the age of 12 years. He was diagnosed to have DSM IV Axis personality
disorder (dependent personality) and panic attacks in the past, treated with benzodiazepines
(clordemetildiazepam 3 mg, flurazepam 30 mg) and zolpidem 10 mg. He denies any symptoms
of RLS, EDS, or daytime sleep attacks. Subjective sleep duration is 3 to 4 hours per night. In the
past he had numerous drugs for sleep amelioration but no clear and stable subjective improve-
ment was noted. An actigraphic monitoring (during drug reduction: clordemetildiazepam 2 mg,
flurazepam 15 mg, and no zolpidem) shows a clear misperception of sleep duration and qual-
ity. The recording shows normal nocturnal motor activity and sleep efficiency and duration.
Note sleep period during the afternoon. He complained of sleeping not more than 3 hours each
night. PSG, polysomnography; TST, total sleep time; SE, sleep efficiency; WASO, wakefulness
after sleep onset; SWS, slow wake sleep; PLMS, periodic limb movements in sleep. (Reproduced
from Chokroverty S, Thomas R, Bhatt M, eds. Atlas of Sleep Medicine. Philadelphia: Elsevier;
2005, with permission.)
860 Part V ■ Complementary and Special Techniques

Figure 41.46 This wrist actigraphic recording is

taken from a 50-year-old woman with chronic
insomnia due to inadequate sleep hygiene. The
actigraph worn for 11 days shows highly irregular
bedtime and wake-up time. The waking period is
indicated by black bars and the sleep period by
gray bars. Note excessive body movements (indi-
cated by black bars) during sleep period and brief
periods of sleepiness (white areas) during wake-
fulness (black bars). Note the subject did not
wear the actigraph on Monday, 5/ 16/ 05 from
7:00 PM to 6:30 AM (fourth from the top showing
a continuous, thick, black line).

Figure 41.47 Primary delayed sleep phase syn-

drome. This wrist actigraphic recording is taken
from a 29-year-old man with a lifelong history of
delayed sleep onset and delayed wake-up time.
The actigram shows his typical sleep period from
3:00 AM to 4:00 AM to 9:00 AM to noon (white
areas). If he has to wake up early in the morning,
he feels exhausted and sleepy all day. He feels
fine if he is allowed to follow his own schedule.
Melatonin at night did not help him. Morning
bright light therapy was suggested but the patient
declined. (Reproduced from Chokroverty S,
Thomas R, Bhatt M, eds. Atlas of Sleep Medicine.
Philadelphia: Elsevier; 2005, with permission.)

Figure 41.48 Sleep–awake schedule disorder in

a patient with acquired immunodeficiency syn-
drome (AIDS). This wrist actigram is taken from
a 46-year-old man with AIDS. The patient pre-
sented with sleep difficulty due to inability to fall
asleep and wake at desired bedtime and wake-up
time. This 10-day recording shows disorganiza-
tion of the sleep–wake schedule. There is a sug-
gestion of non–24-hour (hypernychthemeral)
syndrome with progressive delay of sleep onset
(arrows) from day 1 to day 6, and again delayed
sleep onset (arrows) from day 7 to day 10.
(Reproduced from Chokroverty S, Thomas R,
Bhatt M, eds. Atla s of Sleep Medicin e.
Philadelphia: Elsevier; 2005, with permission.)
 Chapter 41
remission, and therapeutic responses; ability to overcome the
problem of night-time variability noted in many sleep disorders
and some nocturnal movement disorders; ability to discriminate
real movements from clinically insignificant EMG potentials;
usefulness in sleep state misperception; and ability to document
delayed or advanced sleep phase syndrome or non–24-hour cir-
cadian rhythm disorders although sleep logs may suggest such
diagnosis. Finally, actigraphs are small, lightweight, easily main-
tained, and inexpensive in contrast to bulky, labor intensive, and
expensive, as well as inconvenient PSG recordings.
Disadvantages and Limitations of Actigraph Recordings
These include the inability to diagnose sleep apnea and to clarify
the etiology of insomnia, overestimation of sleep when some
insomniacs may lie down in bed for prolonged periods without
moving; inability to identify subjects with feigning sleep prob-
lems and to discriminate types of movements such as PLMS from
other body movements and provide any information about other
physiologic characteristics (e.g., EEG, EOG, and respiration); and
inability to record specific data as the actigraph records all kinds
of movements. Lack of standardization of placement of the acti-
graph may also be seen as a pitfall. Most commonly, the actigraph
is placed on the nondominant side. Activity is somewhat differ-
ent between the two sides with more activities being recorded
from the dominant than from the nondominant limbs. The over-
all agreement, however, for sleep period between the two sites is
not frequently different when compared with the PSG data.
In conclusion, the actigraph is an inexpensive, useful
method for longitudinal assessment of sleep–wake pattern, can
differentiate normal sleep patterns from those of disturbed
sleep due to insomnia and SDB, and can differentiate normal
sleep from sleep–wake schedule disturbances by recording over
long periods and longitudinal monitoring.
REFERENCES
1. Berger H. Uber Das elektroencephalogramm Des Menschen. Arch
Psychiatr Nervenkr. 1929;87:527–570.
2. Aserinsky E, Kleitman N. Regularly occurring periods of eye motil-
ity and concomitant phenomenon during sleep. Science. 1953;
119:273.
3. Holland JV, Dement WC, Raynal DM. Polysomnography: a response
for a need for improved communication. Paper presented at the
14th annual meeting of the Association of the Psychophysiological
Study of Sleep, Association of the Psychophysiological Study of
Sleep, Jackson Hole, WY, 1974:121.
4. Caton R. The electric currents of the brain. BMJ. 1975;2:278.
5. Loomis AL, Harvey EN, Hobart GA. Cerebral states during sleep,
as studied by human brain potentials. J Exp Physiol. 1937;21:127.
6. Kohlschutter E. Messeungen Dar Festigkei Des Schlafes. Zeitschrift
Fur Rationelle Medicin. 1863;17(3):209–253.
7. Basner M. Arousal threshold determination in 1862: Kohlschutter’s
measurements on the firmness of sleep. Sleep Med. 2010;11:417–422.
8. Dement WC, Kleitman N. Cyclic variations of EEG during sleep
and their relation to eye movements, body motility and dreaming.
Electroencephalogr Clin Neurophysiol. 1957;9:673.
9. Jouvet M, Michel F. Correlations electromyographiques Du Sommeil
Chez Le Chat Decortique et Mesencephalique Chroniue. Comp Rend
Soc Biol (Paris). 1959;153:422–425.
■ Polysomnography: Technical and Clinical Aspects 861
10. Berger RJ. Tonus of extrinsic laryngeal muscles during sleep and
dreaming. Science. 1961;134:840.
11. Rechtschaffen K, Kales A. A Manual of Standardized Terminology,
Techniques and Scoring System for Sleep Stages of Human Subjects.
Washington DC: US Government Printing Office; 1968.
12. American Academy of Sleep Medicine. The AASM Manual 2007 for
the Scoring of Sleep and Associated Events: Rules, Terminology and
Technical Specifications. Westchester, IL: American Academy of Sleep
Medicine; 2007.
13. Griesinger W. Berliner medizinisch-psychologische Gesclleschaft.
Arch Psychiatr Nervenkr. 1868;1:200–204.
14. Freud S. Project from a scientific psychology. In: Bonaparte M,
Freud A, Kris, E eds. The Origins of Psychoanalysis: Letters to
Wilhelm Fliess, Drafts and Notes, 1895 to 1902. New York: Basic
Books; 1954:400.
15. Wolpert S. A New History of India. New York: Oxford University
Press; 1982:48.
16. Jones B. Neurology of NREM sleep. In: Montagna P, Chokroverty S,
eds. Sleep Disorders, Handbook of Clinical Neurology. Amsterdam:
Elsevier 2011 (In Press).
17. Saper CB, Scammell TE, Lu J. Hypothalamic regulation of sleep
and circadian rhythms. Nature. 2005;437:1257–1263.
18. Peyron C, Tighe DK, Van den Pol AN, et al. Neurons containing
hypocretin (orexin) project to multiple neuronal systems. J Neurosci.
1998;18:9996.
19. Siegel JM. Mechanisms of sleep control. J Clin Neurphysiol.
1990;7:49.
20. McCarley RW. Neurobiology of REM and NREM sleep. Sleep Med.
2007;8:302–330.
21. Lu J, Sherman D, Devor M, et al. A putative flip-flop switch for
control of REM sleep. Nature. 2006;441:589.
22. Brooks PL, Peever JH. Glycinergic and GABAA-mediated inhibi-
tion of somatic motoneurons does not mediate rapid eye move-
ment sleep motor tonia. J Neurosci. 2008;28:3535.
23. Saper CB, Chou T, Scammell TE. The sleep switch: hypothalamic
control of sleep and wakefulness. Trends Neurosci. 2001;24:726.
24. Porkka-Hieskanen T, Strecker RE, Thakkar M, et al. Adenosine: a
mediator of the sleep-inducing effects of prolonged wakefulness.
Science. 1997;276:1265.
25. Chokroverty S. Physiological changes in sleep. In: Chokroverty S,
ed. Sleep Disorders Medicine: Basic Science, Technical Considerations
and Clinical Aspects. 3rd ed. Philadelphia: Elsevier; 2009:80–104.
26. Moore RY. Circadian timing and sleep-wake regulation. In:
Chokroverty S, ed. Sleep Disorders Medicine: Basic Science,
Technical Considerations and Clinical Aspects. 3rd ed. Philadelphia:
Elsevier; 2009:105–111.
27. Daan S, Beersma DGM, Borbely AA. The timing of human sleep:
A recovery process gated by a circadian pacemaker. Am J Physiol.
1984;246:R161–R178.
28. Dinges DF, Broughton RJ. Sleep and Alertness: Chronobiological,
Behavioral and Medical Aspects of Napping. Raven Press; 1989.
29. Dijk DJ, Czeisler CA. Contribution of the circadian pacemaker and
the sleep homeostat to sleep propensity, sleep structure, electroen-
cephalographic slow waves and sleep spindle activity in humans.
J Neurosci. 1995;15:3525–3538.
30. Jasper HH. The ten-twenty electrode system of the International
Federation. Electroencephalogr Clin Neurophysiol. 1958;10:371.
31. Mouret J, Delorme F, Jouvet M. Activité des muscles de la face au
cours du sommeil paradoxal chez l’homme. Compt Rend Soc Biol
Paris. 1965;159:391–394.
32. Bliwise D, Coleman R, Bergmann B, et al. Facial muscle tonus dur-
ing REM and NREM sleep. Psychophysiology. 1974;11:497–508.
862 Part V ■ Complementary and Special Techniques
33. Report of AASM Task Force. Sleep-related breathing disorders in
adults: recommendations for syndrome definition and measure-
ment techniques in clinical research. Sleep. 1999;22(5):667–689.
34. Kales A, Kales J. Recent findings in the diagnosis and treatment of
disturbed sleep. N Engl J Med. 1974;209:487–499.
35. Williams RI, Karacan I, Hursch CJ. EEG of Human Sleep: Clinical
Applications. New York: Wiley; 1974.
36. Bonnet MH, Arand DL. EEG arousal norms by age. J Clin Sleep
Med. 2007;3(3):271–274.
37. Bonnet M, Carley D, Carskadon M, et al. EEG arousals: scoring
rules and examples: a preliminary report of the Sleep Disorders
Atlas Task Force of the ASDA. Sleep. 1992;15:173–184.
38. Coleman RM, Pollak C, Weitzman ED. Periodic movements in
sleep (nocturnal myoclonus): relation to sleep-wake disorders. Ann
Neurol. 1980;8:416–421.
39. Bonnet M, Carley D, Carskadon M, et al. Recording and scoring
leg movements. Sleep. 1993;16:748–759.
40. Zucconi M, Ferri R, Allen R, et al. The official World Association
of Sleep Medicine standards for recording and scoring PLMS and
PLMW developed in collaboration with a task force from the
International RLS Study Group. Sleep Med. 2006;7:175–183.
41. Montplaisir J, Boucher S, Lesperance P, et al. Immobilization tests
and periodic leg movements in sleep for the diagnosis of restless
legs syndrome. Mov Disord. 1998;13(2):324–329.
42. Gastaut H, Tassinari C, Duron B. Etude polygraphique des manifes-
tations episodiques (hypniques et respiratoires) du syndrome de
Pickwick. Rev Neurol. 1965;112:568–579.
43. Young T, Palta M, Dempsey J, et al. The occurrence of sleep disor-
dered breathing among middle-ages adults. N Engl J Med. 1993;
328:1230–1235.
44. Bixler EO, Vgontas AN, Ten Have T, et al. Effects of age on sleep
apnea in men: prevalence and severity. Am J Resp Cricial Care Med.
1998;157:144–148.
45. Pavlova MK, Duffy JF, Shea SA. PSG respiratory abnormalities in
aymptomatic individuals. Sleep. 2008;31:241–248.
46. Terzano MG, Mancia D, Salati MR, et al. The cyclic alternating
pattern as a physiologic component of normal NREM sleep. Sleep.
1985;8:137–145.
47. Terzano MG, Parrino L, Sherieri A, et al. Atlas, rules, and record-
ing techniques for the scoring of cyclic alternating pattern (CAP)
in human sleep. Sleep Med. 2001;2:537–553.
48. Parrino L, Halasz P, Tassinari CA, et al. CAP, epilepsy, and motor
events during sleep: the unifying role of arousal. Sleep Med Rev.
2006;10:267–285.
49. Indications for PSG Task Force. Practice parameters for the
indications for PSG and related procedures. Sleep. 1997;20:
406–422.
50. Kushida CA, Littner MR, Morgenthaler T, et al. Practice parame-
ters for the indications for PSG and related procedures: an update
for 2005. Sleep. 2005;28:499–521.
51. Chesson AL Jr, Berry RB, Pack A. Practice parameters for the use
of portable monitoring devices in the investigation of suspected
obstructive sleep apnea in adults. Sleep. 2003;26:907.
52. Department of Health and Human Services, Center for Medicare
and Medicaid Services. Decision Memo for Continuous Positive
Airway Pressure (CPAP) Therapy for Obstructive Sleep Apnea (OSA).
CAG#0093R. March 13, 2008. Available at: http://www.cms.hhs.gov/
mcd/viewdecisionmemo.asp?from2 =viewdecisionmemo.asp&id=
204&. Accessed August 7, 2010.
53. Collop NA, Anderson WM, Boehlecke B, et al. Clinical guidelines
for the use of unattended portable monitors in the diagnosis of
obstructive sleep apnea in adult patients. Portable Monitoring
Task Force of the American Academy of Sleep Medicine. J Clin
Sleep Med. 2007;3:737–747.
54. White DP, Gibb TJ, Wall JM, et al. Assessment of accuracy and
analysis time of a novel device to monitor sleep and breathing in
the home. Sleep. 1995;18:115.
55. Tonelli de Oliveira AC, Martinez D, Vasconcelos LF, et al. Diagnosis
of obstructive sleep apnea syndrome and its outcomes with home
portable monitoring. Chest. 2009;135:330.
56. Penzel T, Hirshkowitz M, Harsh J, et al. Digital analysis and tech-
nical specifications. J Clin Sleep Med. 2007;3(2):109–120.
57. Schulz H. Rethinking sleep analysis: comment on the AASM
Manual for Scoring Sleep and Associated Events. J Clin Sleep Med.
2008;4:99–103.
58. Guilleminault C, Stoohs R, Clerk A, et al. A cause of daytime
sleepiness: the upper airway resistance syndrome. Chest. 1993;104:
781–787.
59. Exar E, Collop N. The upper airway resistance syndrome. Chest.
1999;115:1127–1139.
60. Norman RG, Ahmed MM, Walsleben JA, et al. Detection of respi-
ratory events during NPSG: nasal cannula/pressure sensor versus
thermistor. Sleep. 1997;20:1175–1184.
61. Epstein MD, Chicoine SA, Hanumara RC. Detection of UARS using
a nasal cannula/pressure transducer. Chest. 2000;117:1073–1077.
62. Budhiraja R, Goodwin JL, Parthasarathy S, et al. Comparisons of
NPT and thermistor for detection of respiratory events during
PSG in children. Sleep. 2005;28(9):1117–1121.
63. Chokroverty S, Bhatt M, Goldhammer T. Polysomnographic record-
ing technique. In: Chokroverty S, Thomas RJ, Bhatt M, eds. Atlas
of Sleep Medicine. Philadelphia: Elsevier/Butterworth-Heinemann;
2005:1–28.
64. Dauvilliers Y, Arnulf I, Mignot E. Narcolepsy with cataplexy. Lancet.
2007;369:499–511.
65. Nishino S. Clinical and neurobiological aspects of narcolepsy. Sleep
Med. 2007;8:373–399.
66. Frenette E, Kushida CA. Primary hypersomnia of central origin.
Semin Neurol. 2009;29:354–367.
67. Baker TL, Guilleminault C, Nino-Murcia G, et al. Comparative
PSG study of narcolepsy and idiopathic CNS hypersomnia. Sleep.
1986;9:232–242.
68. Mattarozzi K, Bellucci C, Campi C, et al. Clinical, behavioral and
PSG correlates of cataplexy in patients with narcolepsy/cataplexy.
Sleep Med. 2008;9:425–433.
69. Harsh J, Peszka J, Hartwig G, et al. Night-time sleep and EDS in
narcolepsy. J Sleep Res. 2000;9:309–316.
70. Huang YS, Lin YH, Guilleminault C. PSG in Kleine–Levine syn-
drome. Neurology. 2008;70(10):795–801.
71. Rosenow F, Kotagel P, Cohen BH, et al. MSLT and PSG in diagnos-
ing Kleine–Levin syndrome and periodic hypersomnia. J Clin
Neurophysiol. 2000;17(4):519–522.
72. Kales A, Soldatos CR, Kales JD. Sleep disorders: insomnia, sleep-
walking, night terrors, nightmares, and enuresis. Ann Intern Med.
1987;106:582–592.
73. Ohayon M, Guilleminault C, Priest R. Night terrors, sleepwalking,
and confusional arousals in the general population: their frequency
and relationship toother sleep and mental disorders. J Clin
Psychiatry. 1999;60:268–276.
74. The International Classification of Sleep Disorders. Diagnostic and
Coding Manual. 2nd ed. Westchester, IL: American Academy of
Sleep Medicine; 2005.
75. Schenck CH, Bundie SR, Ettinger MG, et al. Chronic behavioral
disorders of human REM sleep. A new category of parasomnia.
Sleep. 1986;9:293–308.
 Chapter 41
76. Boeve B, Silber M, Parisi J, et al. Synucleinopathy pathology and
REM sleep behavior disorder plus dementia or parkinsonism.
Neurology. 2003;61:40–45.
77. Schenck C, Mahowald M. REM sleep behavior disorder: clinical
developmental, and neuroscience perspectives 16 years after for-
mal identification. Sleep. 2002;25:120–138.
78. Lapierre O, Montplaisir J. PSG features of REM sleep behavior
disorder: development of a scoring method. Neurology. 1992;42:
1371–1374.
79. Foldvary N, Caruso AC, Maschaa E, et al. Identifying montages
that best detect EEG seizure activity during polysomnography.
Sleep. 2000;23:221–229.
80. Foldvary-Schaefer N, DeOcamp J, Mascha E, et al. Accuracy of
seizure detection using abbreviated EEG during polysomnogra-
phy. J Clin Neurophysiol. 2006;23:68–71.
81. Walczak TS, Radtke RA, Lewis DV. Accuracy and interobserver
reliability of scalp ictal EEG. Neurology. 1992;42:2279–2285.
82. Derry CP, Davey M, Johns M, et al. Distinguishing sleep disorders
from seizures: diagnosing bumps in night. Arch Neurol. 2006;63:
705–709.
83. Jung R, Kuhlo W. Neurophysiological studies of abnormal night
sleep and the Pickwickian syndrome. Prog Brain Res. 1965;18:140.
84. Banno K, Kryger MH. Sleep apnea: clinical investigations in humans.
Sleep Med. 2007;8:400.
85. Yaggi HK, Concato J, Kernan WN, et al. Obstructive sleep apnea as
a risk factor for stroke and death. N Engl J Med. 2005; 353:2034.
86. Sanders MH, Strollo PD, Jr, Atwood CW, Jr., et al. Positive airway
pressure in the treatment of sleep related breathing disorders. In:
Chokroverty S, ed. Sleep Disorders Medicine: Basic Science, Technical
Considerations and Clinical Aspects. 3rd ed. Philadelphia: Saunders-
Elsevier; 2009:340.
87. Berry RB, Hill G, Thompson L, et al. Portable monitoring and auto
titration vs. polysomnography for the diagnosis and treatment of
sleep apnea. Sleep. 2008;31:1423.
88. Walters AS. Toward a better definition of the restless legs syn-
drome. The International Restless Legs Syndrome Study Group.
Mov Disord. 1995;10:634–642.
89. Allen RP, Picchietti D, Hening WA, et al. Restless legs syndrome:
diagnostic criteria, special considerations, and epidemiology. A
report from the restless legs syndrome diagnosis and epidemiology
workshop at the National Institute of Health. Sleep Med. 2003;
4:101–119.
90. Hening WA, Allen RP, Chokroverty S, et al., eds. Restless Legs
Syndrome. Philadelphia: Saunders/Elsevier; 2009.
91. Winkelmann J, Schormair B, Lichtner P, et al. Genome-wide asso-
ciation study in restless legs syndrome identifies common variants
in three genomic regions. Nat Gen. 2007;39:1000–1006.
92. Stefansson H, Rye DB, Hicks A, et al. A genetic risk factor for peri-
odic limb movements in sleep. N Engl J Med. 2007;357:703–705.
93. Hening W, Walter AS, Allen RP, et al. Impact, diagnosis and treat-
ment of restless legs syndrome (RLS) in a primary care study. Sleep
Med. 2004;5(3):237–246.
94. Hornyak M, Feig B, Voderholzer U, et al. Polysomnography find-
ings in patients with restless legs syndrome and in healthy con-
trols: a comparative observational study. Sleep. 2007;30:861–865.
95. Saletu B, Anderer P, Saletu M, et al. EEG mapping, psychometric,
and polysomnographic studies in restless legs syndrome (RLS)
and periodic limb movement disorder (PLMD) patients as com-
pared with normal controls. Sleep Med. 2002;3(suppl):S35–S42.
96. Winkelman JW, Redline S, Baldwin CM, et al. Polysomnographic
and health related quality of life correlates of restless legs syn-
drome in the sleep heart health study. Sleep. 2009;32:772–778.
■ Polysomnography: Technical and Clinical Aspects 863
97. Mahowald M. Hope for the PLMS quagmire. Sleep Med. 2002;3:
463–464.
98. Howell MJ, Schenck CH, Crow SJ. A review of nighttime eating
disorders. Sleep Med Rev. 2009;13(1):23–34.
99. Siddiqui F, Osuna E, Chokroverty S. Writing emails as part of sleep-
walking after increase in Zolpidem. Sleep Med. 2009;10(2):262–264.
100. Pevernagie D, Boon P, Mariman A, et al. Vocalization during
episodes of prolonged expiration: a parasomnia related to REM
sleep. Sleep Med. 2001;2:19–30.
101. Vetrugno R, Provini F, Plazzi G, et al. Catathrenia (nocturnal
groaning): a new type of parasomnia. Neurology. 2001;56:681–683.
102. Siddiqui F, Walters AS, Chokroverty S. Catathrenia: A rare para-
somnia which may mimic central sleep apnea on polysomno-
gram. Sleep Med. 2008;9:460–461.
103. Chokroverty S. Sleep disordered breathing in neuromuscular dis-
orders: a condition in search of recognition. Muscle and Nerve.
2001;24:451–455.
104. Chokroverty S. Neuromuscular disorders and sleep dysfunction.
In: Montagna P, Chokroverty S, eds. Sleep Disorders, Handbook of
Clinical Neurology. Amsterdam: Elsevier 2011 (In Press).
105. Labanowski M, Schmidt-Nowara W, Gulleminault C. Sleep and
neuromuscular disease: frequency of sleep disordered breathing
in a neuromuscular disease clinic population. Neurology. 1996;17:
1173–1180.
106. Langevan B, Petijean T, Philip F, et al. Nocturnal hypoventilation in
chronic respiratory failure (CRF) due to neuromuscular disease.
Sleep. 2000;23(suppl 4):S204–S208.
107. Martin TJ, Sanders MH. Chronic alveolar hypoventilation: a
review for the clinician. Sleep. 1995;18:617–634.
108. Chokroverty S. Polysomnography in movement disorders. In:
Hallett M, ed. Handbook of Clinical Neurophysiology: Movement
Disorders. Amsterdam: Elsevier; 2002:139.
109. Chokroverty S. Sleep and neurodegenerative diseases. Semin
Neurol. 2009;29:446–468.
110. Arnulf I, Merino-Andreu M, Blocs F, et al. REM sleep behavior
disorder and REM sleep without atonia in patients with progres-
sive supranuclear palsy. Sleep. 2005;28:349–354.
111. Sixel-Doring F, Scweitzer M, Mollenhauer B, et al. Polysomno-
graphic findings, a video-based sleep analysis and sleep percep-
tion in progressive supranuclear palsy. Sleep Med. 2009;10:
407–415.
112. Segawa M, Hysaka A, Miwakawa F, et al. Hereditary progressive
dystonia with marked diurnal fluctuation. Adv Neurol. 1976;
14:215–233.
113. Salazar-Grueso EF, Rosenberg RS, Roos RP. Sleep apnea in olivo-
pontocerebellar degeneration: treatment with trazodone. Ann
Neurol. 1988;23:399–401.
114. Neil JF, Holzer BC, Spiker DG, et al. EEG sleep alterations in
olivopontocerebellar degeneration. Neurology. 1980;30:660–662.
115. Salva MA, Gulleminault C. Olivopontocerebellar degeneration,
abnormal sleep and REM sleep without atonia. Neurology. 1986;36:
576–577.
116. Lugaresi E, Medori R, Montagna P. Fatal familial insomnia and
dysautonomia with selective degeneration of thalamic nuclei. N
Eng J Med. 1986;315:997.
117. Montagna P. Fatal familial insomnia and the role of the thalamus
in sleep regulation. In: Montagna P, Chokroverty S, eds. Handbook
of Clinical Neurology. Amsterdam: Elsevier 2011 (In Press).
118. Broughton RJ, Mullington JM. Polysomnography; principles and
applications in sleep and arousal disorders. In: Niedermeyer E,
Lopes Da Silva F, eds. Electroencephalography. 5th ed. Philadelphia:
Lippincott Williams & Wilkins; 2005:899–936.
864 Part V ■ Complementary and Special Techniques
119. Hauri P, Hawkins DR. Alpha-delta sleep. Electroencephalogr Clin
Nuerophysiol. 1973;34:233.
120. Gharagozlou P, Seyffert M, Santos R, et al. Rhythmic movement
disorder associated with respiratory arousals and improved by
CPAP titration in a patient with restless legs syndrome and sleep
apnea. Sleep Med. 2009;10:501–503.
121. Chokroverty S, Thomas RJ, Bhatt M. Atlas of Sleep Medicine.
Philadelphia: Butterworth-Heinemann; 2005.
122. Cosentino FL, Lero I, Tripodi M, et al. the neurophysiology of the
alternating leg muscle activation (ALMA) during sleep: study of
one patient before and after treatment with pramipexole. Sleep
Med. 2006;7:63–71.
123. Chervin RD, Consens FB, Kutluay E. Alternating leg muscle acti-
vation during sleep and arousal: a new sleep-related motor phe-
nomenon? Mov Disord. 2003;18:551–559.
124. Montagna P, Provini F, Plazzi G, et al. Propriospinal myoclonus
upon relaxation and drowsiness: a cause of severe insomnia. Mov
Disord. 1997;12:66–72.
125. Vetrugno R, Provini F, Meletti S, et al. Propriospinal myoclonus
at the sleep-wake transition: a new type of parasomnia. Sleep.
2001;24:835–843.
126. Critchley M. The Predormitum. Rev Neurol (Paris). 1955;93:101.
127. Alattar MA, Scharf SM. Opioid-associated central sleep apnea: a
case series. Sleep Breath. 2009;13:201–206.
128. Farney RJ, Walker JM, Boyle KM, et al. Adaptive servoventilation
(ASV) in patients with sleep disordered breathing associated with
chronic opioid medication for non-malignant pain. J Clin Sleep
Med. 2008;4:311–319.
129. Pepin JL, Dale D, Argod J, et al. Recommendations for practical use
of pulse transit time as a tool for respiratory effort measurements
and microarousal recognition. In: Chokroverty S, Thomas R, Bhatt
M, eds. Atlas of Sleep Medicine. Philadelphia: Elsevier/Butterworth-
Heinemann; 2005:262–271.
130. Pittman SD, Thomas RJ. Peripheral arterial tonometry. In:
Chokroverty S, Thomas R, Bhatt M, eds. Atlas of Sleep Medicine.
Philadelphia: Elsevier/Butterworth-Heinemann; 2005:285–288.
131. Chokroverty S. Sleep disorders. In: Bradley W, Daroff R, Fenichel
G, et al., eds. Neurology in Clinical Practice. Philadelphia: Elsevier/
Butterworth-Hinemann; 2008:1947.
132. Johns MW. A new method for measuring daytime sleepiness: the
Epworth Sleepiness Scale. Sleep. 1991;14:540.
133. Keenan SA. Polysomnography technique: an overview. In:
Chokroverty S, ed. Sleep Disorder Medicine: Basic Science, Clinical
Consideration and Technical Aspects. Philadelphia: Saunders/
Elsevier; 2009:137–156.
134. Ackersted T, Gilberg M. Subjective and objective sleepiness in the
active individual. Int J Neurosci. 1990;52:29–37.
135. Carskadon MA, Dement WC, Mitler M, et al. Guidelines for the
multiple sleep latency test (MSLT: a standard measure of sleepi-
ness). Sleep. 1986;9:519.
136. Littner MR, Kushaida C, Wise M, et al. Practice parameters for
clinical use of the multiple sleep latency test and the maintenance
of wakefulness test. Sleep. 2005;28:113–121.
137. Cherbin R. Assessment of sleepiness. In: Chokroverty S, Hening
WA, Walters AS, eds. Sleep and Movement Disorders. Boston:
Butterworth-Heinemann; 2002.
138. Doghramji K, Mitler M, Sangal RB, et al. A normative study of
the maintenance of wakefulness test (MWT). Electroencephalogr
Clin Neurophysiol. 1997;103:554.
139. Doghramji K. The maintenance of wakefulness test. In:
Chokroverty S, ed. Sleep Disorders Medicine: Basic Science,
Technical Considerations and Clinical Aspects. Philadelphia:
Saunders/Elsevier; 2009:224–228.
140. Mitler MM, Miller JC, Lipsitz JJ, et al. The sleep of long-haul
truck drivers. N Engl J Med. 1997;337:755–761.
141. Philip P, Sagaspe P, Taillard G, et al. Maintenance of wakefulness
test, obstructive sleep apnea syndrome, and driving risk. Ann
Neurol. 2008;64:410–416.
142. Colburn TR, Smith BM, Guarini JJ, et al. An ambulatory activity
monitor with solid state memory. ISA Transaction. 1976;15:
114–154.
143. Tyron WW. Activity Measurement in Psychology and Medicine.
New York: Plenum Press; 1991.
144. Sadeh A, Hauri PJ, Kripke DF, et al. The role of actigraphy in the
evaluation of sleep disorders. Sleep. 1995;18:288–302.
145. Littner M, Kushida CA, Anderson WM, et al. Practice parameters
for the role of actigraphy in the study of sleep and circadian
rhythms: An update for 2002. Sleep. 2003;26:337–341.
146. Sforza E, Johannes M, Claudio B. The PAM-RL ambulatory device
for detection of periodic leg movements: a validation study. Sleep
Med. 2005;6:407.
147. Kemlink D, Pretal M, Sounka K, et al. A comparison of
polysomnographic and actigraphic evaluation of periodic limb
movements in sleep. Neurol Res. 2008;30:234.
148. Allen RP. Improving RLS diagnosis and severity assessment:
polysomnography, actigraphy and sleep log. Sleep Med. 2007;8:S13.
149. Kohnen R, Allen RP, Benes H, et al. Assessment of restless legs
syndrome—methodological approaches for use in practice and
clinical trials. Move Disord. 2007;22:S485.
150. Allen RP. Activity monitoring to diagnose and evaluate motor
abnormalities of sleep. In: Henning W, Chokroverty S, eds. Topics
in Movement Disorders in Sleep, Course Syllabus: ASDA Annual
Meeting. San Francisco: American Sleep Disorders Association;
1997.
151. Morganthaler TI, Lee-Chiung T, Friedman AG, et al. Practice
parameters for the clinical evaluation of circadian rhythm sleep
disorders; an American Academy of Sleep Medicine report. Sleep.
2007;30:1445–1459.

Magnetoencephalography:
Methods and Applications
R COMPARISON OF EEG AND MEG
ecording of weak magnetic fields outside the head by
means of magnetoencephalography (MEG) emerged
in the late 1960s, 40 years after the invention of
the human EEG. The first instrument was an induction coil
magnetometer with 2 million turns of wire. It was used to
detect the magnetic alpha rhythm by means of signal averaging,
with the electric alpha as the time reference (1). The MEG
method became more practical with the introduction
of SQUID (superconducting quantum interference device)
magnetometers in the early 1970s; since then it has been possi-
ble to record both spontaneous and evoked magnetic signals of
the human brain without any EEG reference. Rapid develop-
ment of the technology has taken place during the last few
years: sensor arrays with whole-scalp coverage are now com-
mercially available, and signal analysis methods have progressed
quickly. At present, tens of laboratories worldwide utilize MEG
for exploration of normal and abnormal functions of the
human brain, and MEG results continue to influence interpre-
tation of EEG data.
In this chapter, the basic principles of MEG are reviewed
briefly, with examples mainly from our own laboratory. Both
spontaneous and evoked activities in normal subjects and in
some neurologic patients are discussed to the extent which is
supposed to be of immediate interest to a clinical neurophysiol-
ogist. Several review articles are available for consulting MEG
findings directly related to basic brain research and relevant
methodology (2–18). The main advantages of MEG are its good
spatial resolution in locating cortical events and its selectivity to
activity of the fissural cortex. The sites of active brain areas can
be located with respect to external landmarks on the head,
brain structures, or functional brain regions (identified by
means of sources of evoked responses).
One goal of MEG recordings is to obtain information about
the neural generators of various signals. The EEG research has
traditionally focused on temporal waveforms, and a whole
branch of electrophenomenology has arisen around EEG
“graphoelements,” which have been correlated with different
tasks and stimulation parameters, clinical states, and even with
personality factors. It is clear that a better understanding of the
generators underlying the EEG signals would permit a more
precise and physiologic interpretation of abnormalities. MEG
has considerably clarified these issues during the last decade,
both at conceptual and practical level, and may have much to
offer in the future.
CHAPTER
RIITTA HARI
42
MEG is closely related to EEG. Despite different sensitivities of
EEG and MEG to sources of different orientations and locations,
the primary currents causing the signals are the same. Similarities
between the MEG and EEG waveforms are therefore to be
expected. The advantage of MEG over EEG in source identifica-
tion results mainly from the transparency of the skull and other
extracerebral tissues to the magnetic field, in contrast to the sub-
stantial distortion and smearing of the electric potentials. Thus
the MEG pattern outside the head is less distorted than the EEG
distribution on the scalp. The magnetic recording is also refer-
ence-free, whereas the electric brain maps depend on the location
of the reference electrode. As a result it is often difficult to make
a reasonable guess of the source locations by visual inspection of
the EEG data. Reference-free EEG presentations can be obtained
by calculating the “surface Laplacians” (19,20). However, even
then one has to proceed to neural sources for proper interpreta-
tion, and problems arise for several simultaneous source. The
Laplacians cannot be calculated for the outermost electrodes,
which reduces the brain areas that can be characterized.
Figure 42.1 shows a current dipole in a spherical volume
conductor consisting of four layers of different conductivities
that simulate the brain, the cerebrospinal fluid, the skull, and
the scalp. The resulting distributions of electric potential and
magnetic field are “dipolar,” that is, display two extrema of
opposite polarities, but rotated by 90 with respect to each
other. The isocontour lines are relatively more tight in the mag-
netic than in the electric pattern because concentric electric
inhomogeneities smear only the electric potential, while the
magnetic pattern is not influenced. In an ideal sphere, a single
superficial tangential dipole can be found one third more accu-
rately on the basis of magnetic than electric recordings (21). For
interpretation of MEG signals recorded from the real brain, it is
sufficient to use a realistically shaped model consisting of the
brain only, while accurate EEG calculations require a full mul-
ticompartment model with known conductivities and shapes
for the brain, skull, cerebrospinal fluid, and scalp.
MEG’s selectivity to tangential currents in the presence of
several simultaneous sources is an important advantage in
practical work, as is illustrated, for example, by the early differ-
entiation between multiple cortical areas activated by
somatosensory stimuli (12,22–26). Moreover, it is often more
straightforward to interpret MEG than EEG data.
865
866 Part V ■ Complementary and Special Techniques
Figure 42.1 Magnetoencephalographic (MEG) and electroencephalo-
graphic (EEG) field patterns over a concentric four-layer sphere when a
tangential current dipole (shown by the arrow) is active in an area
approximating the second somatosensory cortex in the upper lip of the
Sylvian fissure. The shadowed areas indicate the magnetic flux out of
the head (MEG) and positive potential (EEG). The lower part of the
figure illustrates schematically, with shadowed ellipsoids, the inaccu-
racy region of the dipole location when computed “backwards” from
the signals on the surface. The dipole is assumed to be in a wall of a
cortical fissure.
Since both electric and magnetic signals are generated by
the same primary currents that flow in the brain and the
nearby tissues, one should pay attention to both MEG and
EEG data when drawing conclusions on brain function from
either type of recordings. For example, a dipolar potential dis-
tribution could be explained equally well with two radial cur-
rent dipoles or with one tangential dipole. However, the
existence of a clear magnetic pattern, rotated 90 with respect
to the electric one, implies that the latter explanation should
be favored (27). The minimum requirement for sound data
interpretation is that the conclusions based on EEG and MEG
do not contradict.
Some current sources (very deep and radial) are more reli-
ably picked up by EEG than MEG, but MEG counterparts of
auditory brainstem responses can be detected with MEG as
well and they can give important information about the
source configuration (28). However, the contribution of even
deep activity to the MEG signals can be probed by inserting
sources to interesting brain structures (such as thalamus or
hippocampus) and then calculating the best fit of these
sources, as a function of time, to the measured signals (29,30).
In theory, current loops are electrically silent but magnetically
visible. In practice, however, ideal current loops have not been
observed, and all cerebral currents that give rise to MEG sig-
nals are also expected to generate electric potentials on the
scalp. For maximum information, the MEG and EEG tech-
niques should be combined, and the simultaneously recorded
data should be interpreted with methods that take advantage
of the complementarity of the records (see also the section
“The contribution of neuronal models: in vitro and computa-
tional models” in Chapter 5).
MEG INSTRUMENTATION
The magnetic field generated by cerebral currents is two orders
of magnitude weaker than that produced by the heart and only
a tiny part of the steady magnetic field of the earth (Table 42.1).
To avoid external magnetic artifacts caused by moving vehicles,
power lines, radio transmission, and so on, it is common to
carry out the recordings within a magnetically shielded room,
usually made of several layers of aluminum and mu-metal.
Biomagnetic measurements can also be performed without
magnetic shielding when special compensation techniques are
available. In general, it is however better to prevent than to
compensate for artifacts. Adequate magnetically shielded rooms
are commercially available.
Figure 42.2A gives a schematic illustration of, now already
old-fashioned, arrangement during the MEG recording. The
subject lies in a magnetically shielded room and the neuromag-
netometer, containing SQUID sensors immersed in liquid
helium (at –269 C), is positioned close to the head. To replace
the evaporating helium, the Dewar container has to be refilled
regularly, in modern devices about once a week. During record-
ings with the present-day devices that cover the whole scalp in
a helmet-shaped array, such as the 306-channel neuromagne-
tometer in Figure 42.3, the subject is typically sitting during the
measurement.
The cerebral magnetic field is coupled into the SQUID
sensors through superconducting flux transformers. The
transformer configuration is important for the device’s sensi-
tivity to different source current configurations and to arti-
facts. A magnetometer, containing a single pick-up loop in
the flux transformer, is most sensitive to signals but also to
artifacts (Fig. 42.2C). A more elaborate axial first-order gra-
diometer contains a compensation coil, wound in the direc-
tion opposite to the pickup coil (Fig. 42.2B,C); this
Tabl e 4 2 . 1
Orders of Magnitude of Magnetic Fields
(in femtotesla, fT = 10–15 T)
Magnetic resonance 1,000,000,000,000,000 (=1 T)
imaging
Steady magnetic field 100,000,000,000
of the earth
Magnetocardiogram 100,000
Cerebral -rhythm 1000
Cerebral evoked response 100
Sensitivity of magnetometers 10
Noise within a shielded room 1
Note that the sensitivity of the present-day magnetometers is below 5 fT/ 1 Hz.
 Chapter 42
configuration decreases the influence of distant disturbances
that link the same magnetic flux into both coils. Therefore,
the output of the axial first-order gradiometer is essentially
determined by the signal of the nearby neuronal source itself.
Furthermore, since the distance between the pickup and
compensation coils of a first-order axial gradiometer is sev-
eral centimeters, the measured signal approximates the
amplitude of the radial field component Br rather than its
axial derivative. Higher-order gradiometers are necessary for
measurements performed without a magnetic shield; in these
systems the sensitivity is reduced to the distant artifacts but
to some extent also to the nearby brain currents.
The contour plots in Figure 42.2B illustrate the pattern of
Br , the magnetic field radial to the surface, generated by a tan-
gential current dipole in a sphere. The signal strengths meas-
ured with an axial gradiometer form a spatial pattern similar
to the field itself, with extrema of opposite polarities at the
two sides of the dipole. In contrast, the planar gradiometer
yields the maximum signal when centered just over the dipole
at the location of the steepest field gradient. The planar gra-
diometer is able to detect the location of the source even at the
edge of the sensor array and the essential information from
the field pattern can thus be obtained from a rather small
measurement area. On the other hand, information about the
depth of the source is more accurate with axial than planar
sensors since the gradient decreases relatively more rapidly as
the function of source depth than does the field itself (Fig.
42.2C). However, whole-scalp sensor coverage largely coun-
terbalances this drawback. Although the patterns measured by
both axial and planar gradiometers are easily interpreted, in
practice it is often useful that the maximum signal detected by
the planar gradiometers suggests the approximate source loca-
■ Magnetoencephalography: Methods and Applications 867
Figure 42.2 (A) Magnetoencephalogram
recording. The subject is lying in a magnet-
ically shielded room with his head sup-
ported by a vacuum cast. The Dewar
containing the SQUID sensors is brought as
close to the head as possible, without direct
contact, and the magnetic field (or its gradi-
ent) is picked up outside the head at several
locations simultaneously. (B) Two flux
transformer configurations. The first-order
axial gradiometer (left) measures essentially
Br and detects the maximum signals at both
sides of the dipole. The planar gradiometer
(right) measures the tangential derivative
e Br/ x or Br/ y; the maximum signal is
e e e
detected just above the dipole. (C)
Dependence of signal strength (in arbitrary
units) on the depth of a current dipole when
measured by (i) a magnetometer, (ii) a first-
order axial gradiometer, (iii) a second-order
axial gradiometer, and (iv) a planar figure-
of-eight gradiometer. Simulations in a
sphere of 10-cm radius.
tion. Methods have been developed to present data measured
with different coil configurations in a standard format (31).
On the basis of axial gradiometers, some users compute vir-
tual planar gradiometer signals that work well as indicators of
the most likely source areas but have ( 2 times) more noise
than the original signals.
An essential part of the measurement system is the head-posi-
tion indicator, which gives the exact measurement sites and the
orientations of the sensors with respect to the head. We obtain
this information by placing three to four small wire loops on
known sites on the scalp. The field pattern produced by currents
led through the loops is then measured with the multichannel
magnetometer. In another commonly used head-position indica-
tor system a transmitter is connected to magnetic sensors, which
are fixed on the dewar, and three receivers are placed on the sub-
ject’s head. The head-position indicator devices allow the posi-
tion of the magnetometer to be determined with respect to
external landmarks on the head with 2- to 3-mm accuracy. The
coordinate system is preferably global, typically fixed on the basis
of the nasion, inion, and the preauricular points so that it can be
easily transported to, for example, Talairach space commonly
used in functional magnetic resonance imaging (MRI). The
important landmarks on the head and the shape of the skull can
be determined with a three-dimensional digitizer.
To determine the current distribution within the head, the
magnetic field must be sampled at several locations, preferably
simultaneously. When only single-channel magnetometers were
in use, it sometimes took several days to complete a field map.
For example, the MEG recordings of the first epileptic patients
lasted for 16 hours (32)! Changes in the subject’s attentive state
and vigilance were thus unavoidable. With the present-day
helmet-shaped neuromagnetometers that cover the whole scalp
868 Part V ■ Complementary and Special Techniques

Figure 42.3 A modern 306-SQUID neuromagnetometer used in our laboratory. Each of the 102 three-channel sensor
elements comprises two orthogonal planar gradiometers and one magnetometer; the helmet-shaped arrangement of the
elements is shown on the right (courtesy of Mika Seppä). The planar flux transformers measure the tangential derivatives
Br/ x and Br/ y of the radial field component Br.
e e e e

(Fig. 42.3), the whole field pattern can be measured without
moving the instrument. This progress of instruments has finally
made MEG recordings feasible for comprehensive studies of
integrative brain functions in health and disease.
The optimal spacing of the sensors is determined by the spa-
tial frequency of the signal distribution and, therefore, only
marginal benefit can be obtained by reducing the sensor spac-
ing below the distance between the source and the sensor, that
is, about 3 cm (33,34). Thus about 150 sensors would suffice to
cover the whole cortex, and the newest instruments with 200 to
300 sensors provide dense-enough spatial sampling for MEG
recordings. The diameter of the sensor coil is always a matter of
compromise: the larger the coil, the more sensitive it is, but also
the more it averages the magnetic field, thereby leading to loss
of information.
Currenlty (in 2010) about 140 whole-scalp neuromagnetome-
ters are in use in clinical and research settings all over the world.
The devices of different manufacturers vary in the number of
channels, in the pickup coil configuration (e.g., first-, second-, or
third-order axial gradiometer, planar gradiometer, or magne-
tometer), as well as in the area covered by the sensor array.
However, the different devices provide practically very similar
information about the brain’s neuronal currents.
With the present technology, the intrinsic noise of the
SQUID is no longer a problem, and the main “noise” arises
from the brain itself. Taking into account the distance of the
coils from the source, pickup coils with diameters around 2 cm
are reasonably sensitive and do not lose significant information.
Decreasing the distance to the source improves the resolution of
the system. Thus it would be important to develop flux trans-
former arrays that can be put close to the scalp of each subject,
independent of the head shape. Such systems could be made
with high-Tc superconductor arrays that require less insulation.
However, despite the promising recordings of MEG signals with
a single-channel high-Tc neuromagnetometer (35), the tech-
nique has not developed during the last years. The main reasons
are the higher (thermal) noise of the sensors and the difficulty
to easily manufacture multichannel high-Tc sensor arrays.
Interestingly, totally new sensor principles have been
recently introduced: atomic magnetometer (36) and magne-
toresistive sensors operating up to 77 K (37), in contrast to the
SQUIDs that are operated at 4 K.
 Chapter 42 ■ Magnetoencephalography: Methods and Applications 869

SOURCE IMAGING
Since the aim of MEG studies is to obtain information about
brain function, the field pattern should be interpreted in terms
of cerebral currents. Yet, due to the nonuniqueness of the
inverse problem—namely that several current distributions
can, in principle, produce identical magnetic field patterns out-
side the head—the interpretation usually requires the use of
specific source and volume conductor models. Thus the situa-
tion is more complicated than in MRI, functional magnetic res-
onance imaging (fMRI), or in positron emission tomography
(PET), where the inverse problem can be solved uniquely.
However, physiologically meaningful solutions of MEG pat-
terns can be found by utilizing constraints based on the known
anatomy and physiology of the brain.

CURRENT DIPOLE
The most commonly used source model in MEG studies is a
current dipole within a sphere. The dipole can be characterized
by means of five parameters, three for its three-dimensional
location, one for its orientation (only the plane parallel to the
surface of the sphere is relevant), and one for its strength. In a
sphere, only primary currents tangential to the surface will
produce a magnetic field. In reality, signals may also be
detected from the convexial cortex where the currents are con-
Figure 42.4 The dependence of the radial component of the magnetic
siderably closer to the detector than are the currents within the
field on the distance between the source and the detector. Two current
wall of a fissure. Deviation of the convexial source by only 10
dipoles are situated in a homogeneous sphere (radius 10 cm), 1 cm
to 20 degrees from the radial orientation is enough to give a
beneath the surface, 3 cm from each other, and symmetrically with
signal as large as that produced by a tangential dipole of the
respect to the origin. The field was calculated on an arc perpendicular
same size but 2 cm deeper in the fissural cortex. It therefore
to the orientation of the dipoles (x-axis). The pattern is complicated on
seems probable that under realistic conditions nearly radial
the surface, whereas 2 cm outside the sphere the higher spatial frequen-
sources may contribute significantly to the MEG signals. This
cies have faded away and the dipolar term dominates. The amplitudes
notion is supported by simulations that took into account the
have been normalized according to the maximum value; the maximum
real geometry of the cerebral cortex (38). A practical example
field would be about seven times stronger on the surface than 2 cm
of the possibility to detect tilted currents is the identification of
above it. (Adapted from Hari R. Interpretation of cerebral magnetic
magnetic counterparts of the P22 and P25 somatosensory
fields elicited by somatosensory stimuli. In: Basar E, ed. Springer Series
responses, which are believed to arise in the convexial cortex
of Brain Dynamics. Berlin/ Heidelberg: Springer-Verlag; 1988:305–310.)
(39–41).
In dipole modeling, the location of the “equivalent current
dipole” (ECD) is found by a least-squares fit to the data, typically
at a time point of a clearly dipolar field pattern. In progressing poor. It is also often useful to study the residual field, that is, the
toward a multidipole model, one may extract the field patterns difference between the measured field and that predicted by the
produced by the already identified sources to facilitate further model, or to compare the waveforms predicted by the model with
analysis (42–44). The success of dipole models in MEG interpre- the measured waveforms. If the residual field shows systematic
tation derives, in part, from the difficulty to discern the details of features that cannot be explained by noise, a different source con-
the brain activation pattern from the typical measurement dis- figuration must be considered.
tance of at least 3 cm from the source (Fig. 42.4). Consequently, The inaccuracy of MEG localization is smallest in the direc-
a single current dipole is a reasonably accurate description of a tion transverse to dipole orientation (cf. Fig. 42.1) and largest
local active cortical area of less than 2 to 3 cm in diameter. (about double) in the direction of depth. The locating accuracy
The adequacy of the model can be evaluated by calculating the depends drastically on the signal-to-noise ratio, which is
goodness-of-fit (g) value (5,45), that is the squared correlation improved by signal averaging for evoked responses but cannot
coefficient between the measured signals and those predicted by be affected much—except by filtering—in the case of sponta-
the model. The g-value, however, depends on several factors such neous activity. For comparison, note that the best localization
as the number and distribution of the measurement locations accuracy for EEG is in the direction along the dipole (Fig. 42.1).
(46). A low g-value means that either the brain source signifi- Since the dipoles mainly reflect synchronous activation of the
cantly deviates from the model or that the signal-to-noise ratio is cortical pyramidal cells and thus are perpendicular to the
870 Part V ■ Complementary and Special Techniques
Figure 42.5 Schematic illustration of equivalent current dipole (ECD)
locations when the source is (A) a layer of less than 2 cm in diameter,
(B) a layer with angular extension, and (C) a layer extended along the
radius of the sphere. The activity is assumed to occur in the wall of a
cortical fissure. This behavior depends slightly on the coil configuration
used. (Adapted from Hari R. On brain’s magnetic responses to sensory
stimuli. J Clin Neurophysiol. 1991;8:157–169.)
cortical surface, changes in the focus of activation along the cor-
tex are more easily seen with MEG than EEG. On the other
hand, EEG may be more accurate in indicating which wall of a
fissure has been activated, since here the distinction must be
made along the direction of the dipole.
When a single current dipole is used to model an extended
cortical area consisting of a layer of dipoles, the estimates of
dipole depth and, consequently, of dipole strengths may be
erroneous (Fig. 42.5). Better source models are thus needed for
extended areas. When multiple brain regions are simultane-
ously active, the relative contribution of each area to the meas-
ured signal depends on its site, strength, and synchrony.
Multidipole models with time-varying source strengths are use-
ful in interpreting the resulting complex field patterns (47).
NEURAL CURRENTS UNDERLYING
THE CURRENT DIPOLE
To understand the cellular events underlying the ECD, it is rea-
sonable to divide the currents associated with postsynaptic
potentials (PSPs) to transmembrane currents at the active
synaptic area, intracellular currents within the neuron, and
extracellular “volume” currents. Transmembrane currents are
the driving force for both intra- and extracellular current flow.
ECD calculated from the measured MEG signals reflects the
direction of the net intracellular current flow. A PSP at the end
of one dendrite produces an “elementary” dipole moment Q =
I (~10–14 nAm), where I is the intracellular current driven by
the synapse and is the length constant of the cell membrane
(48). The strength of the observed dipole moment can be
strongly affected by the simultaneous calcium currents.
Since it is not known how many PSPs occur synchronously
in each pyramidal cell and how much cancellation takes place
within the cortex, it is not possible to derive accurate estimates
for the source area of a typical evoked response on the basis of
cell/synapse density and the size of the elementary dipole.
Okada et al. (49) note that many animal species across a wide
phylogenetic scale (turtle, guinea pig, and swine) show an
apparent invariance in the maximum current dipole moment
density of 1 to 2 nAm/mm 2. Considerably smaller dipole
moment densities have been estimated on the basis of available
information on intracortical current densities and values of
cortical (50). However, intracortical current densities depend
strictly on the degree of neuronal synchrony, and thus on the
specimen studied, and may vary according to background
activity (51). Since is proportional to the square root of the
fiber diameter, the dipole moments weigh the larger neurons.
Glial cells may affect the MEG and EEG signals: they change
the return paths of volume currents and have activation latencies
that fit with the occurrence of long-latency-evoked responses.
Several important electrical events probably remain beyond the
reach of both MEG and EEG measurements. For example, intra-
cortical short-distance interactions are—when viewed from the
top of the cortex—radially symmetric; only their net effect along
the dendrites of the pyramidal cells gives rise to an MEG signal.
MINIMUM-NORM ESTIMATES
An example of a different approach to the neuromagnetic
inverse problem is the minimum-norm estimate (MNE) pres-
entation of the source currents (52). MNE gives the most prob-
able current distribution, in the sense of the minimum norm,
giving rise to the recorded magnetic field. Calculation of MNE
does not require specific assumptions about the source config-
uration (one dipole, multiple dipoles, quadrupoles, etc.), which
is an advantage when there is no basis for explicit models. On
the other hand, the unconstrained MNE solution favors super-
ficial currents and does not give a reliable estimate of the source
depth. Therefore MNE solutions have been weighted with con-
straints based on the known brain anatomy and physiology,
especially source depth (53,54). Minimum-current estimates
(MCEs) (55), a subclass of MNE solutions that give more local
solutions, have been applied for visualization and analysis of
single-subject and group-level MEG data. One example of MCE
visualization is given in Figure 42.6.
The distributed MNE-based source activations often look
“more physiologic” than the point-like current dipoles, and
therefore current dipoles have been considered old-fashioned
and less appropriate than the distributed models to describe the
complex brain activity. However, a word of caution has its place
 Chapter 42 ■ Magnetoencephalography: Methods and Applications 871

Figure 42.6 Minimum current estimates of the 20-Hz band activity projected to the brain surface of a single subject before
(left) and after (middle) oral benzodiazepine administration; the brain is viewed from right. Coregistration of the sources
on the subject’s MRI on the right indicates activation of the primary motor cortex, with virtually identical source areas
before and after application of benzodiazepines. (Adapted from Jensen O, Goel P, Kopell N, et al. On the human motor-
cortex beta rhythm: sources and modelling. Neuroimage. 2005;26:347–355.)

here because the appearance of the result depends strongly on
the method used: the MNE approach will give a distributed
solution and the dipole approach a local solution, whatever the
real current distribution is (see Ref. (16))! In general, both
dipole and MCE solutions give rather similar results, although
the temporal accuracy may be better with dipole modeling and
the group data may be easier to visualize with MCE; however,
even well-informed users tend to report more false positives
with MCE than with current dipole modeling (56).
COMBINATION OF MEG WITH
MRI/ fMRI/ PET DATA
Functional information from MEG is at present routinely com-
bined with structural information from MRI. Moreover, all
source estimates could be improved by constraints based on the
known anatomy of the brain, derived from MRI data, and forc-
ing the dipoles/currents to the cortical tissue. In preoperative
evaluation, reconstruction of the three-dimensional outer
surface of the brain from MRI scans helps the surgeon to recog-
nize the landscape after opening the skull when the brain tissue
retracts and the relationship between the brain and the land-
marks on the skull changes (57). Adding surface vessels to the
reconstruction further facilitates the orientation during surgery
(58) (Fig. 42.30).
Compared with PET and fMRI (59), the advantage of MEG is
its good temporal resolution that allows monitoring of cortical
dynamics in millisecond scale. In combined use of multiple
methods, active brain regions have been first determined with
PET/fMRI and then used as source constrains in the inverse
solution of the MEG data to reveal the corresponding temporal
behavior. However, not all changes in the synchronization of a
neuronal population, reflected in the MEG/EEG signals, induce
significant changes in the mean neuronal firing rates or the
blood flow and metabolism, and vice versa. Therefore MEG and
fMRI results may significantly differ in certain conditions (60),
and the MEG/EEG and fMRI/PET methods remain comple-
mentary in studies of human brain function. New approaches to
combine fMRI and MEG data have been introduced, and the
theoretical basis of the fusion is improving (see, e.g., Ref. (61)).
VOLUME CONDUCTOR MODELS
The sphere model works well in most areas of the head when
the radius of the sphere is fitted to the local radius of curvature
of the measurement area (62), likely because the realistic noise
largely masks the errors caused by the different conductor mod-
els (63). Realistic head models that take more computing time
may be needed for proper modeling of the temporobasal and
frontobasal brain areas (64). Fortunately, it is sufficient for
MEG to model the intracranial space since only a relatively
small proportion of the currents flow in the poorly conducting
skull. In contrast, proper modeling of the EEG signals necessi-
tates a multicompartment model consisting of 3 to 4 concentric
layers with known resistivities.
Some nonspherical electric inhomogeneities may also affect
MEG distributions (65). For example, the falx cerebri may
change volume current paths and lead to slight mislocalizations
of current dipoles in the mesial wall of the hemisphere. There
has been discussion about the significance of holes in the skull
to the distribution of volume currents. In the intact human
skull, the main holes are in the base of the skull and in the
orbits. The absence of electro-oculographic artifacts in
intracranial recordings from the frontal lobe (66) suggest a
rather good isolation between the outer and the inner sides of
the skull. Therefore the effects of normal holes of the skull seem
negligible for the interpretation of MEG distributions, as is also
supported by calculations showing that radial anisotropy added
to a spherically symmetric conductor does not affect the exter-
nal magnetic field (67).
872 Part V ■ Complementary and Special Techniques
PRACTICAL ISSUES
Successful clinical MEG recordings need sophisticated instru-
mentation and close interdisciplinary collaboration. The exper-
imental situation introduces some restrictions. Since the head
has to be immobile, recordings cannot be performed during
major motor seizures, and problems are encountered in studies
of uncooperative subjects who either cannot keep still during
the recording or who are not willing or able to perform the
tasks. However, continuous monitoring of the head position is
currently possible with some neuromagnetometers. Moreover,
multichannel MEG mappings are quick to perform and the rel-
ative locations of the sensors, required for accurate source
analysis, are exactly known without extra effort.
Figure 42.7 illustrates some common MEG artifacts. Since
significant contamination can be caused by eyeblinks and move-
ments (68), the electro-oculogram should be used to reject, or at
Figure 42.7 Different artifacts in magnetoencephalogram (MEG)
recordings of spontaneous brain activity over the temporal area; the
measurements were made with a planar gradiometer. Blinking and ver-
tical eye movements produce signals of about equal size; for horizontal
eye movements the signals are similar in waveform but have different
spatial distributions. Muscular tension refers to biting the teeth
together. Respiration artifacts were due to a metallic piece over the
chest of the subject. The magnetic ballistocardiogram was due to move-
ment of magnetized metal on the bed, on which the subject was lying
on his left side. Note the timing differences between the maximum sig-
nals in the electrocardiogram (ECG) and the ballistocardiogram. The
lowest artifact was due to a digital watch 70 cm from the Dewar.
least monitor, contaminated signals. Muscular artifacts cause
problems less frequently. Magnetic lung contamination or
magnetic material in clothing can cause respiration-related slow
shifts. Cardiac artifacts can be either magnetocardiographic sig-
nals (69,70), picked up at distance, or ballistocardiographic fluc-
tuations, caused by the movement of magnetized material at the
rhythm of the cardiac cycle. For recognition of both artifacts,
simultaneously recorded ECG is essential: the magnetocardio-
gram coincides with the ECG, whereas the main peak of the
magnetic ballistocardiogram is broader and lags the QRS com-
plex by several hundred milliseconds.
All magnetic materials must be avoided in the clothing of the
subject. Although intraoral metallic devices for orthodontics or
magnetized material used in staples and sutures during brain
surgery may contaminate the MEG measurement, recent devel-
opment of signal-space separation (SSS) method and its tem-
poral extension (tSSS) help to remove artifacts in the
postprocessing phase (71–73). It is even possible to clean the
MEG data from artifacts elicited by deep brain stimulator (74),
as is illustrated in Figure 42.8.
Raw data
VERTEX FRONTAL
EOG OCCIPITAL
After tSSS-cleaning
VERTEX FRONTAL
EOG OCCIPITAL
Figure 42.8 The effect of tSSS artifact suppression on spontaneous
magnetoencephalogram (MEG) of a person with deep brain stimulator.
Top: Spontaneous MEG signals from the left hemisphere when the stim-
ulator was on. The approximate locations of the planar gradiometers are
indicated; total duration 30 seconds. Bottom: After tSSS when even the
normal posterior alpha activity is visible. The amplitude scales are the
same in both panels. Courtesy of Jyrki Mäkelä, Helsinki University
Central Hospital. (See color insert)
 Chapter 42
50 fT/cm
optic fiber switch
bundle
receiver
transmitter
reflectance emittance
skin trigger
–100 0 100
Figure 42.9 Tactile stimulator made of a bundle of optical fibers and the somatosensory responses obtained by stimulating
the left hand and the left lip. The corresponding source areas in the right-hemisphere primary somatosensory cortex
and the current directions are indicated on the right (hand up, lip below). (Adapted from Jousmäki V, Nishitani N,
Hari R. Brush stimulator for functional brain imaging. Clin Neurophysiol. 2007;118:2620–2624.) (See color insert)
Special attention must also be paid to the proper design of
stimulators. High-quality sounds can be produced with elec-
troacoustic transformers placed outside the shielded room and
connected through plastic tubes to the subject; the system may
need sophisticated equalizing to guarantee flat frequency trans-
fer. Excessive artifacts from electric somatosensory stimuli can
be avoided by twisting the stimulator wires tightly together.
Multichannel devices with balloon diaphragms driven by com-
pressed air (75) provide a nice and artifact-free method for tac-
tile stimulation. A hand-held bundle of optical fibers, half
emitting red light and the other half detecting the reflected light
from the skin can be used to evoke reliable somatosensory
responses to taps to the skin (76); see Figure 42.9.
Noxious laser heat stimuli for pain studies can be brought to
the shielded room via optical cables. Visual stimuli can be trans-
mitted through mirrors, a data projector, a bundle of optic
fibers, or the subject can view a monitor through a hole in the
wall. EEG can be measured with nonmagnetic electrodes and
wires without causing problems to the MEG recordings. Eye
tracking can be performed during the MEG recordings with an
infrared camera.
■ Magnetoencephalography: Methods and Applications 873
52
Bandwidth 0.1–145 Hz
right hemisphere
31
Left Hand
41
28
46
25
Left Lip
32
18
200 300 ms
Figure 42.10 draws attention to dangers of filtering. A too
high high-pass filter setting will cause well-known distortions
in the signal waveforms but—what may be less obvious—also
transfer the dipolar field patterns, with opposite polarities, to
latencies distant from the real signal and thereby lead to erro-
neous interpretations of brain activation.
ANALYSIS OF SPONTANEOUS ACTIVITY
Very useful information about spontaneous activity (for a
review, see Ref. (11)) can be obtained by calculating frequency
spectra of the signals (Fig. 42.11) and by mapping the abun-
dance of different frequencies at various sensor sites. The
sources can then be identified either in time or in frequency
domain (77–79). In multichannel recordings, cross-spectra
between channels suggest which signals arise from the same
source, and time lags may tell about the sequence of activation.
One may quantify the level of different brain rhythms by
using time–frequency representations, or more simply using the
temporal spectral evolution (TSE) method (80) by focusing on
one frequency band at a time. The TSE method resembles the
Figure 42.10 Dangers of filtering,
illustrated with simulated data. A cur-
rent dipole with a monophasic activa-
tion curve (gray line on the left, “no
filter”) and with 30 nAm dipole
moment was inserted to a site corre-
sponding to the right auditory cortex.
The field patterns above show a dipolar
field pattern at peak activation at 100
msec. After high-pass filtering at 6 Hz
the waveform changes (black curve on
the left, “HP 6 Hz”) and ghost sources
appear at 60 and 150 msec.
874 Part V ■ Complementary and Special Techniques
Figure 42.11 Amplitude spectra from a 1-minute period of spontaneous
activity when the subject was resting with eyes closed. Two orthogonal
derivatives of the radial magnetic field were measured at each location,
along the longitude and latitude. The insets show reactivity of the - and
-rhythms to opening of the eyes and to movements of the left and right
hand. (Adapted from Salmelin R, Hari R. Characterization of sponta-
neous magnetoencephalogram (MEG) rhythms in healthy adults.
Electroenceph Clin Neurophysiol. 1994;91:237–248.)
“event-related desynchronization” technique (81) used to study
task-dependent changes in the human scalp EEG. However, TSE
preserves the original units and the signal levels can thus be
directly compared with the sizes of evoked responses. In TSE,
the brain signals are first bandpass filtered, then rectified (tak-
ing absolute values of the signals), and thereafter averaged with
respect to the triggering event. Thus, all signals that are time-
locked but not exactly phase-locked to the event will be
detected.
Sometimes one may be interested in just seeing the signal
waveform and the temporal changes in the field pattern. This
type of analysis, which resembles the classical use of EEG
recordings, can be helpful in screening candidates for surgical
treatment: clear changes of the field pattern from one irritative
phenomenon to the next discourage the assumption of a local
onset of the discharge, although they do not rule out a deeper
common trigger.
Due to noise problems and the use of single-channel instru-
ments, the first studies of epileptic MEG activity employed
averaging, with the simultaneously recorded EEG signal as a
trigger (32,82). The drawback of such a procedure is that the
choice of the trigger channel largely determines the observed
activity, and the method works only if well-discernible spikes or
sharp waves are present. With multichannel low-noise instru-
ments, averaging is necessary only for the detection of prespikes,
which might be accurate indicators of the onset area of the
paroxysmal activity (83). With template matching, signals can
be automatically classified and the source locations determined
separately for each class.
SPONTANEOUS ACTIVITY OF
AWAKE NORMAL SUBJECTS
Alpha Rhythm from Parieto-Occipital Areas
The typical parieto-occipital alpha rhythm, with a peak fre-
quency around 10 Hz, is damped by opening the eyes
(Fig. 42.11). In the first study of the human magnetic alpha
rhythm (1), a phase reversal was observed between signals meas-
ured from the right and left hemispheres, and the currents were
therefore suggested to be parallel to the longitudinal fissure.
Later, the sources of the alpha rhythm have been suggested
to cluster mainly in the region of the calcarine sulcus (84) and
around the parieto-occipital sulcus (POS) (85–88). Figure
42.12, based on whole-scalp MEG recordings, shows sources in
both of these locations, typically the POS region is the far most
dominant source of the MEG alpha rhythm.
Vvedensky et al. (85) suggested that alpha spindles often
have the same generators for about 1 second, whereas the
sources differ for successive spindles. Later, a multitude of MEG
alpha sources were found with the help of “signal-space analy-
sis” (89). Different spindles seemed to have different sources
but the source configuration of one spindle was rather stable.
Later recordings, analyzed with time-varying fixed-location
dipole models, however, suggest that a typical alpha spindle
cannot be explained by a single source (88).
The parieto-occipital alpha rhythm is strongly suppressed
during visual stimuli, visual memory tasks, and visual imagery
(90–93). When the subject has to differentiate between visually
presented images of objects vs. nonobjects, the poststimulus
alpha level is consistently higher after nonobjects (94). This
Figure 42.12 Sources of alpha oscillations superimposed on two sagit-
tal MRI slices demonstrating a calcarine (left) and a parieto-occipital
(right) source cluster. Source current orientations are also indicated.
(Adapted from Hari R, Salmelin R, Mäkelä JP, et al. Magnetoence -
phalographic cortical rhythms. Int J Psychophysiol. 1997;26:51–62.)
 Chapter 42 ■ Magnetoencephalography: Methods and Applications 875

modulation may reflect interaction between the dorsal and

ventral visual pathways during an attention-demanding task.
The level of parietal alpha activity has been related to memory
load (95).
Early visual deprivation may lead to the abolishment of
parieto-occipital spontaneous rhythms: from the five early
blind subjects in Figure 42.13, four do not display alpha activ-
ity and opening/closing the eyes did not have any effect on
their brain rhythms. Only one subject (B5) has rhythmic
alpha-like activity; she had lost her sight earlier than the
other four subjects. The reason for the observed alpha
absence in blinds, also known from EEG recordings (96),
is unknown, but it certainly speaks against a purely idling
nature of the normal alpha rhythm; otherwise one might
expect strong alpha during the continuous lack of
visual input.

Mu Rhythm from Sensorimotor Areas

The electroencephalographic -rhythm has a magnetic coun-
terpart that starts to dampen 1 to 2 seconds before a movement
(Fig. 42.14). In association with unilateral movements, the sup-
pression is bilateral, although contralaterally dominant (97).
The -rhythm consists of two main frequency components,
one around 10 Hz and the other around 20 Hz (see inset in
Fig. 42.11); however, these two frequencies are usually not exact
harmonics and show independent temporal behaviors (98).
Additional evidence for functionally separate frequency compo-
nents of the -rhythm derives from the source locations, which
center on average 5 mm more anterior for the 20 Hz than the
10 Hz frequencies (80), suggesting the existence of separate pre-
central (20 Hz) and a postcentral (10 Hz) rhythms (Fig. 42.15).
In addition to movements (99), the magnetic -rhythm is
modified by electric stimulation of peripheral nerves, with a
Figure 42.13 Traces of parieto-occipital alpha activity from five early
more clear increase, “rebound” of the 20-Hz than the 10-Hz
blind subjects and seven control subjects. The blind subjects were 25 to
component around 400 msec after the stimulus (80,97). The
32 years in age. Subjects B2 and B3 were able to see some light; subject
poststimulus rebound is abolished when the subject moves the
B5 is blind due to retinal cancer and has seen light only during her first
fingers of the same hand. Interestingly, a similar, but weaker,
year. M. Huotilainen and T. Kujala participated in data collection.

Figure 42.14 Spontaneous magnetic activity over the rolandic area, shown on six gradiometer channels, when the subject
has his eyes open and clenches his contralateral fist (see the electromyogram [EMG] channel). (Unpublished data from
F. Lado and R. Hari.)
876 Part V ■ Complementary and Special Techniques
Figure 42.15 Top: A 5-second trace of magnetic -rhythm from the
rolandic region. Bottom: Isodensity clusters of sources of 10- and 20-Hz
oscillations over the rolandic region of one subject, based on thousands of
equivalent current dipole (ECD) locations. The black ovals show the area
activated by median nerve stimulation at the wrist, and the line shows the
estimated course of the rolandic fissure. (Adapted from Salmelin R, Hari
R. Spatiotemporal characteristics of rhythmic neuromagnetic activity
related to thumb movement. Neuroscience. 1994;60:537–550.)
suppression is seen when the subject just imagines making the
movements, indicating involvement of the primary motor cortex
in motor imagery (100). Suppression of the rebound, as a sign of
motor–cortex activation, also occurs when the subject just views
another person to make certain movements (Fig. 42.16) (101).
Figure 42.16 Level of the 20-Hz rolandic activity after stimulation of
the contralateral median nerve when the subject either rested (solid
lines), viewed another person to manipulate a small object with the
right hand fingers (dashed line; “viewing”), or manipulated the same
object herself (“acting”). The sources of these 20-Hz signals were in the
precentral primary motor cortex. (Adapted from Hari R, Forss N,
Avikainen S, et al. Activation of human primary motor cortex during
action observation: a neuromagnetic study. Proc Natl Acad Sci USA.
1998;95:15061–15065.)
The 20-Hz rebounds do—whereas the 10-Hz rebounds do
not—follow the moved body part in a somatotopic manner:
they appear at lateral rolandic areas after mouth movements,
more medially after finger movements, and close to midline
after foot movements (102).
Recent MEG recordings after oral administration of benzo-
diazepine demonstrated a strong increase in the power of
approximately 20-Hz oscillations in both hemispheres, with
sources in the primary motor cortex close to the hand area
(103) (Fig. 42.6). These data suggest that the motor cortex is
an important effector site of benzodiazepine, and they also
agree with the proposed generation of the rolandic 20-Hz
oscillations in the motor cortex. The frontal-lobe dominance
of benzodiazepine-related scalp-EEG beta activity can be
explained by tangential current dipoles in the wall of the
central sulcus.
Oscillatory Cortex–Muscle Coupling
The human rolandic MEG activity has been observed to have
a close temporal relationship to peripheral muscular activity
(54,104–106). For example, during isometric contraction of
different muscles, the muscular and cortical signals are coher-
ent at frequencies varying between 15 and 33 Hz in individ-
ual subjects (106). The sites of maximum coherence in the
motor cortex show gross somatotopic organization, with
activations during foot muscle contraction closer to the head
midline than during hand muscle contractions (Fig. 42.17).
Figure 42.17 Top: Surface electromyogram (EMG) from isometrically
contracted left hallucis brevis muscle and simultaneously recorded
magnetoencephalogram (MEG) signal over the parietal midline (3 to
100 Hz). Middle: Coherence spectra between MEG and EMG during iso-
metric contraction of small hand and foot muscles (left- and right-sided
contractions superimposed). Bottom: Spatial distributions of the
strongest peaks of the coherence spectra. (Adapted from Salenius S,
Portin K, Kajola M, et al. Cortical control of human motoneuron firing
during isometric contraction. J Neurophysiol. 1997;77:3401–3405.)
 Chapter 42 ■ Magnetoencephalography: Methods and Applications 877

The MEG signals lead the EMG signals in time, with increas-
ing time lags with increasing brain–muscle distance. Such
data strongly suggest that the 20-Hz rolandic rhythm reflects,
at the population level, the common central drive to spinal
motoneurons (for reviews, see Refs. (107,108)). The
corresponding cortex–muscle coherence can also be shown
with EEG recordings (109).
Studies with different gripping tasks have led to the proposal
that the cortical oscillations have a role in recalibrating the control
system after a change in the cortex–muscle relationship (110,111).
At strong contractions, the frequency of the coherence jumps
from 20 to 40 Hz (“Piper rhythm”; (112,113)). The cortex–muscle
coherence provides a complementary tool to identify the primary
motor cortex as a part of preoperative functional mapping (58).
Gross et al. (114) demonstrated, using a sophisticated
dynamic imaging of coherent sources (DICS) method, developed
Figure 42.18 Locations of equivalent current dipoles for the tau oscil-
to characterize synchronously firing neural networks in different
lations (6.5 to 9.5 Hz range) in two subjects. The sources (clusters of
parts of the brain, that 6- to 9-Hz velocity changes of slow finger
white dots) are projected to the surface of the brain and shown on the
movements are directly correlated to oscillatory activity of the
subject’s own MRI. The black dot indicates the source of the auditory
primary motor cortex. Moreover, the coherence patterns sug-
N100m response. In the coronal sections in the middle, the dipole dis-
gested that the pulsatile velocity changes were sustained by a cere-
tributions are presented as contour plots, with highest dipole densities
bellar drive through thalamus and premotor cortex.
in white and lowest in black. (Adapted from Lehtelä L, Salmelin R, Hari
Tau Rhythm from Auditory Areas R. Evidence for reactive magnetic 10-Hz rhythm in the human auditory
cortex. Neurosci Lett. 1997;222:111–114.)
A third MEG rhythm, which I have started to call -rhythm
( referring to temporal lobe), was described by Tiihonen et al.
(115). This 8- to 10-Hz oscillatory activity is best seen in the
planar gradiometer recordings just over the auditory cortex. the occipital alpha and from the temporal tau generators.
Occasionally the rhythm is reduced by sound stimuli, such as Epidural electrode recordings (120,121) show that also the tem-
bursts of noise, but it is not dampened by opening the eyes—a poral-lobe 6- to 11-Hz activity arising from the convexial cortex
feature clearly distinct from the occipital alpha. The sources of is strikingly resilient to decreased vigilance.
single-tau oscillations cluster to the supratemporal auditory
cortex close to the generations site of auditory-evoked fields, A Note on Nomenclature
with right-hemisphere dominance (Fig. 42.18) (116). The existence of local cortical rhythms explains in part the
On the basis of the observed current orientations, the corre- confusion in the literature concerning effects of various tasks
sponding electric -rhythm should be seen mostly in the fron- on the spontaneous EEG. Some people consider alpha to refer
tocentral midline where the tau sources, with dipole moments to the occipital “Berger rhythm,” independently of its shape
of 40 nAm, would lead to potentials of about 10 to 20 V. In (i.e., frequency composition), whereas others refer to alpha as
fact, such an EEG rhythm may appear during drowsiness: all signals falling in the 8- to 13-Hz frequency range,
Decreased vigilance is often considered to be associated with a independently of the brain area where they are generated.
“spread” of occipital EEG alpha toward more anterior regions, To avoid confusion, it would be necessary to characterize
with simultaneously slightly decreasing frequency (117,118). A each rhythm both by its site of origin and by its frequency
real spread of the alpha would be in clear contrast to the fixed, range.
although distributed, generators of alpha activity, and a more
plausible explanation is that the anterior “alpha” in fact reflects Other Brain Rhythms
“tau” generated in the supratemporal auditory cortex. A 7- to 9-Hz “sigma” rhythm has been demonstrated in MEG
The combined MEG and EEG data of Lu et al. (119) would recordings in the second somatosensory cortex (122). Single
agree with such a hypothesis: During the awake state (the first median-nerve stimuli elicited some sigma oscillations, and the
panel of Fig. 42.19) the occipital EEG displays rhythmic 10 to level of sigma was enhanced during rhythmic stimulation at
11 Hz alpha activity, and the simultaneous MEG in the tempo- the rhythm’s dominant frequency.
ral region is of low amplitude, with little rhythmic activity. Theta-range MEG activity has been scarcely studied. Sasaki
However, when the occipital electric alpha becomes discontin- et al. (123) recorded 5- to 7-Hz MEG signals from the frontal
uous during light drowsiness (Stage 1a) and seems to spread cortex during mental calculation and intensive thinking. Tesche
more anteriorly, with slightly slowed down frequency, rhythmic (30) estimated the temporal waveforms of sources (computa-
activity of the same frequency appears in the MEG sensors over tionally) inserted to the hippocampus and found in some sub-
the temporal lobe. jects around 5-Hz rhythmic activity during mental calculation.
Therefore, the apparent spread of the EEG “alpha” during Since 1980s, much attention has been paid to the 40-Hz fre-
drowsiness might reflect changed relative contributions from quency band, often supposed to have a role in perceptual binding
878 Part V ■ Complementary and Special Techniques

Figure 42.19 Different vigilance stages in one subject. The recording locations (24-channel planar gradiometer) are indi-
cated on the schematic head; the x-gradients at 12 locations are plotted above and the y-gradients below. In the electric
channels EOGv and EOGh refer to vertical and horizontal oculograms, respectively. The electric amplitude scale refers to
all other electric recordings except the electromyogram (EMG) for which only relative amplitudes are of importance. Note
the increase of rhythmic magnetoencephalographic activity during Stage 1a, compared with the awake stage. A typical V-
wave occurs during Stage 1b. During Stage 2, light sleep, K-complexes occur both in MEG and in electroencephalogram
(EEG). In deep sleep, Stages 3 and 4, slow activity is seen in the whole measurement area. Note the eye movements and
the decreased EMG activity during the REM stage. (Adapted from Lu ST, Kajola M, Joutsiniemi SL, et al. Generator sites of
spontaneous MEG activity during sleep. Electroenceph Clin Neurophysiol. 1992;82:182–196.)

(124,125). Both stimulus-related and later “induced” 40-Hz
activity has been reported, with differences between visual EEG
and MEG 40-Hz signals (53,126). Even higher MEG frequencies
(from 30 to 150 Hz) have been recently used to demonstrate dis-
sociation between visual awareness and spatial attention (127).
Although the 40-Hz activity, according to animal experiments,
results from thalamocortical interaction, it is at present unknown
how strongly the subcortical structures contribute to the MEG
signals that are biased toward cortical currents.
Spontaneous Activity during Sleep
In the first MEG studies of sleep, performed with single-chan-
nel MEG devices, it was difficult to draw conclusions about the
 Chapter 42
relationship between the electric and the magnetic spontaneous
signals (128–130).
Figure 42.19 illustrates the first multichannel MEG record-
ing of sleep where it was possible to classify the different stages
of vigilance on the basis of both EEG and MEG (119). The
activity during awake stage and the appearance of -rhythm
during Stage 1a were discussed above. During deep drowsiness
and light sleep (Stages 1b and 2), vertex waves of 150 to
250 msec duration appeared in the frontocentral EEG leads and
on several MEG channels. The electric and magnetic V-waves
did not always coincide.
Magnetic spindles of 11 to 15 Hz in frequency and 0.3 to
2 seconds in duration appeared during Stage 2. A later study
has shown that the sources of the spindles differ from the
sources of wake-state oscillatory activity (131). The spindles
were occasionally superimposed on high-amplitude tran-
sients, thereby resembling the typical electric K-complexes.
Often the successive magnetic V-waves and K-complexes dif-
fered in waveform and distribution. For example, the midline
distributions of two electric K-complexes of Figure 42.20
resemble each other, whereas the simultaneous magnetic dis-
tributions differ. A typical magnetic K-complex lasted 0.8 to
1.2 seconds.
During slow-wave sleep (Stages 3 and 4; Fig. 42.19), 0.5 to
2 Hz polymorphic MEG activity was widely spread over the
measurement area. During the REM stage, defined on the basis
of rapid eye movements and decreased submental muscular
tone, MEG activity was lower in amplitude and faster in
frequency than during the other sleep stages. Sharp magnetic
transients, resembling the V-waves, were frequently seen.
In the first whole-scalp MEG study of sleep (132), the
sources of K-complexes in the two hemispheres were active
independently. Only in one subject out of six were the K-
complex distributions satisfactorily explained by two current
dipoles, one in each inferior parietal lobe, that is, in line with
the model suggested by Lu et al. (119). In five other subjects the
distributions were more complex. Therefore the K-complexes
■ Magnetoencephalography: Methods and Applications 879
Figure 42.20 Two K-complexes from
the same subject, recorded over the
same location. In the magnetic sig-
nals, the upper traces of each pair
indicate the vertical gradient and the
lower trace the horizontal gradient of
the magnetic field; the recording
locations are shown on the inset
head. Simultaneous electric signals
from the midline are shown in the
insets; their amplitude calibration is
100 V. (Adapted from Lu ST, Kajola
M, Joutsiniemi SL, et al. Generator
sites of spontaneous MEG activity
during sleep. Electroenceph Clin
Neurophysiol. 1992;82:182–196.)
do not appear to be stereotyped responses of the cortex to exter-
nal or internal stimuli, comparable to evoked responses, but
they rather seem to reflect a diffuse and variable reaction
involving large cortical areas.
Up to now, no focal activation spots have been detected
during slow-wave sleep. The sources underlying the spin-
dles also seem very complex. In principal component analy-
sis, four to six separate components were often needed to
explain at least 90% of the field variance during 1 second of
the spindle activity (119). Complexity of spindle generation,
with a multitude of sources and frequencies, has also been
suggested in animal experiments (133), in human depth elec-
trode and scalp topography studies (134,135), as well as in
recent MEG recordings applying synthetic aperture magne-
tometry (136).
Llinas and Ribary (137) point out that while the spontaneous
activity is rather similar during awake and REM sleep stages,
responses to repetitive 40-Hz stimulation are dampened during
REM sleep compared with wakefulness, possibly reflecting the
brain’s isolation from the external world during the REM stage.
Anesthesia
Burst-suppression patterns have been recorded from a dog dur-
ing enflurane and propofol anesthesia (138). The signal pattern
during the bursts was complex and the signal did not show any
organized structure (Fig. 42.21). A practical message from these
recordings is that an essentially artifact-free MEG measurement
can be obtained during respirator-assisted anesthesia when the
respirator is kept outside the magnetically shielded room and
when special care is taken in choosing nonmagnetic tubings
and valves.
MEG IN EPILEPSY
The first MEG recording of an epileptic individual was reported
by Cohen (139): slow EEG and MEG activity increased during
hyperventilation in a patient with psychomotor epilepsy.
880 Part V ■ Complementary and Special Techniques

Figure 42.21 Four magnetoencephalogram (MEG) channels and one

electroencephalogram (EEG) channel from a dog during burst-suppres-
sion phase induced by enflurane anesthesia. (Adapted from Jäntti V,
Baer G, Yli-Hankala A, et al. MEG burst suppression in an anaesthetized
dog. Acta Anaesth Scand. 1995;39:126–128.)

Starting from the early 1980s, a number of epileptic subjects

have been studied in several laboratories (4,32,82,140–157). In
November 2003, the PubMed database already contained 400
publications with keywords of “MEG/magnetoencephalogra-
phy/neuromagnetic and epilepsy,” and thereafter more than one
study per week has appeared in this field. Therefore, it is impos-
sible to cover here the progress in this rapidly expanding field.
The earliest MEG recordings were used to locate both single
and multiple irritative brain areas (32,82,140). In one study,
both scalp and sphenoidal EEG spikes were used as triggers for
averaging the magnetic spikes. The latter had different latencies
for the two triggers, consistent with the propagation of the
epileptic activity (142). The sources of the MEG spikes were
deeper for sphenoidal than for scalp triggers in all five subjects
studied, suggesting propagation of the discharge between dif-
ferent parts of the temporal lobe.
Differentiation between ictal and interictal spikes is as diffi-
cult in MEG as it is in EEG. Since MEG recordings can only
infrequently be performed during motor seizures, the relevance
of the interictal MEG focus has been questioned. Sometimes
the activity just preceding the seizure, assumed to have the best
localizing value, can be detected. One possibility to reveal the
significance of the observed focus is to tailor an electrode array
on the basis of the interictal MEG distribution, by calculating
the forward potential solution from the identified sources, and
then to monitor the ictal activity during telemetric EEG record-
ings to ascertain whether the interictal focus is responsible for
triggering the seizure.
In preoperative evaluation of epileptic patients it is important
to know whether the epileptic discharges are focal, whether there
are multiple foci, and whether these foci show any systematic
time lags. One should also find how close the foci are to function-
ally irretrievable brain regions, such as the sensorimotor and
Figure 42.22 Top: Spatial distribution of one multispike displayed on
the 122-channel sensor array viewed from top. Bottom left: Field pat-
tern during the spike (asterisk above the curves); the sensor array is
viewed from left, top, and right. The shadowed areas indicate magnetic
flux emerging from the head; the isocontours are separated by 400
fT/ cm. The arrows indicate the sites and orientations of the two ECDs
required to account for the field pattern. Bottom right: Dipole moments
as a function of time in both hemispheres. Each trace corresponds to
one unaveraged spike, whose distribution was explained by the two-
dipole model. (Adapted from Hari R, Ahonen A, Forss N, et al. Parietal
epileptic mirror focus detected with a whole-head neuromagnetometer.
Neuroreport. 1993;5:45–48.)
speech areas, and whether the source configuration is stable over
time. MEG seems especially suitable for identifying epileptic foci
in the convexial neocortex. MEG results have been also used to
guide placement of intracranial electrodes and to suggest re-eval-
uation of structural MRI data.
Two early examples of epileptic MEG discharges are presented
below. Figure 42.22 shows spikes, field patterns, and dipole
moments of an 18-year-old female who suffered from complex
partial epilepsy (158). During night her 20- to 30-second
seizures included awakening and bilateral increase of muscular
 Chapter 42
tone, as well as tonic jerks in left extremities. During daytime she
had epochs of paresthesia and loosening of the grip in the left
hand. The abundant spikes, recorded with a 122-channel
whole-scalp magnetometer, could be adequately modeled with
two current dipoles, one in each posterior parietal cortex. The
peak activity occurred consistently 20 msec later in the left than
the right hemisphere, suggesting callosal conduction of the sig-
nals from right to left. Such time lags are useful in differentiating
the primary and the secondary (mirror) foci.
Figure 42.23 illustrates an ictal MEG recording of a man
who suffered from left-sided hemifacial convulsions that were
often triggered by touching the left mouth region (159).
Computerized tomography and PET revealed no abnormali-
ties. The MEG recording displays abundant epileptic spikes
that become increasingly polyphasic during the 14-second
seizure. In the beginning, spikes appear only in the right hemi-
sphere, with a source in the motor face area (defined on the
basis of evoked responses and anatomical structures), whereas
later during the seizure spikes also appear in the homolog area
in the left hemisphere, again (as in Fig. 42.24) with about 20-
msec time lag compared with the other side.
In patients with Landau–Kleffner syndrome (LKS), stereo-
typic 3 c/s spike-and-wave complexes have been shown to be
generated in the auditory cortex (160–163), and the continuous
epileptic discharges in the auditory cortices of LKS children
prevent the proper analysis of sounds, including speech.
Many groups are currently using MEG in preoperative
evaluation of epileptic patients, but the discussion about the
benefits of the procedure continues in quite a heated manner
(164–166).
■ Magnetoencephalography: Methods and Applications 881
Figure 42.23 Epileptic discharges after
voluntary triggering. The top trace illustrates
magnetoencephalographic activity from the
right hemisphere during the whole 14-second
seizure. In the middle and lower parts,
selected periods (A–C) are expanded and sig-
nals from the corresponding area in the left
hemisphere are shown for comparison.
(Adapted from Forss N, Mäkelä JP, Keränen T,
et al. Trigeminally triggered epileptic facial
convulsions. Neuroreport. 1995;6:918–920.)
SPONTANEOUS ACTIVITY
IN OTHER PATIENTS
Ischemic Lesions
Ischemic hemispheric lesions are associated with slow magnetic
shifts and with considerable slowing of cortical activity (167).
Gallen et al. (168) showed that 2- to 3-Hz MEG activity was
mainly concentrated in the damaged but viable tissue at the
periphery of the infarcted area. It is not known, though, to
which extent the changed tissue conductivities at the lesion area
contribute to the measured signals.
Local unilateral lesions in the left anteromedial thalamus
caused extensive and surprisingly symmetric changes in the
parieto-occipital spontaneous activity: the frequency spectra
were broadened in the parieto-occipital area, but the reactivity
to eye opening was preserved and the generation sites of the
activity were similar as in normal subjects (169,170).
Infantile Neuronal Ceroid Lipofuscinosis
As expected, patients with infantile neuronal ceroid lipofusci-
nosis (INCL) have flat MEG, pointing toward a total absence of
cortical activity (171). Brains of INCL patients show almost total
neuronal loss and strong gliosis. Therefore, the flat MEG from
these patients also indicates that glial cells have negligible contri-
bution to the MEG signals in the absence of neuronal activity.
Migraine
Spreading depression (SD) has been suggested as the neuro-
physiologic basis of migraine aura. In turtle cerebellum,
882 Part V ■ Complementary and Special Techniques
Figure 42.24 Effects of electroconvulsive ther-
apy on rolandic magnetoencephalographic sig-
nals and their frequency spectra. (A) Before
treatment, (B) after four treatments, (C) after
eight treatments, and (D) 1 month after the last
treatment. On the right, the before-treatment (A)
spectrum is shown for comparison in all other
conditions ( dashed lin es). (Adapted from
Salmelin R, Mäkelä JP, Heikman P, et al. Human
brain rhythms and electroconvulsive therapy. Soc
Neurosci Abstr. 1996;22(pt 1):342.)
induced SD is associated with strong magnetic fields (172).
Consequently, it has been proposed that SD propagating in the
human brain during the prodromal phase of migraine could
be detected by MEG and such shifts have been reported during
the migraine aura (173). The propagation of SD, assumed to be
about 3 mm/min on the basis of changes in the scotoma in the
visual field, has been elegantly demonstrated in fMRI record-
ings (174).
Effects of Electroconvulsive Therapy
Electroconvulsive therapy (ECT), efficient in treatment of
severe depression, is known to increase slow EEG rhythms.
Heikman et al. (175) tracked the temporal and spatial modula-
tion of spontaneous oscillations in five ECT-treated depressive
patients. ECT resulted in a prominent increase of frequencies
below 8 Hz, which largely disappeared within 1 month after the
last treatment. In three patients, relief of depression was accom-
panied by an increase of the 4- to 8-Hz oscillations by at least
50% after the first four treatments and up to 500% after eight
treatments, particularly in the temporal and rolandic areas.
Encephalitis
In herpes simplex encephalitis, sources of abnormal MEG
waveforms have to be identified in the temporal lobes (176).
Further similar studies might help in the evaluation of the sites
and extent of the functionally disturbed brain areas in posten-
cephalitic patients.
Parkinsonian Tremor
Volkmann et al. (54) found strong coherence, at the frequency
of the 3- to 6-Hz resting parkinsonian tremor, between oscilla-
tory MEG activity in premotor and somatomotor cortices and
muscle activity (54). Previously, Mäkelä et al. (177) had noted
that in patients with hemiparkinsonism the onset of tremor
induces a similar dampening of the 10- and 20-Hz components
of the -rhythm as do normal movements. It was thus sug-
gested that Parkinsonian tremor activates cortical mechanisms
that normally produce rapid alternating movements.
Salenius et al. (178) observed suppression of the normal
15- to 30-Hz cortex–muscle coherence in Parkinsonian patients
who were withdrawn from medication from the preceding
evening; the coherence was, to a large part, restored with the
patients’ normal levodopa treatment. Abolishment of the nor-
mal 15- to 30-Hz synchronized oscillatory activity has been
proposed to lead in bradykinesia (179).
In a more extensive analysis with the DICS method,
Timmermann et al. (180) demonstrated coherence between 4
and 6-Hz Parkinsonian tremor and activity of several cerebral
areas. Coherence was seen both at the tremor frequency and,
even more strongly, at its double. The analysis of partial coher-
ence and phase shifts was especially interesting: The interaction
between M1 and EMG agreed with previous studies, indicating
that cortex leads the muscle, whereas the interaction between
EMG and diencephalic areas was bidirectional. Moreover, a
direct afferent coupling was evident from EMG to second
somatosensory cortex and to posterior parietal cortex.
Cerebellum and premotor areas seemed to be connected with
the periphery via other cerebral areas. Similar widespread cor-
tical circuitries seem to support postural tremor accentuated in
hepatic encephalopathy (181).
Auditory-Evoked Fields
The first demonstrations of the generation of the main
auditory-evoked field (AEF) deflections in the superior surface
of the temporal lobe (182,183), deduced from the measured
field patterns, considerably enlightened the origin of the coin-
ciding EEG evoked potentials that had been previously believed
to be “nonspecific” due to their maximum amplitude at the ver-
tex. Consequently, various long-latency EEG and MEG
responses are used nowadays widely as tools to study functions
of the supratemporal auditory cortex.
Transient Responses
Figure 42.25 shows examples of middle-latency and long-
latency auditory-evoked magnetic fields, obtained using differ-
ent stimulus repetition rates, different filter settings, different
analysis periods, and different number of averaged responses.
After the earliest cortical auditory responses peaking around 19
to 20 msec, or even around 11 msec according to some reports
(184), the interindividually variable but otherwise reliable
 Chapter 42 ■ Magnetoencephalography: Methods and Applications 883

Figure 42.25 Example of middle-latency (left) and long-latency (right) auditory-evoked magnetic fields recorded from
one subject with a 122-channel device. (Adapted from McEvoy L, Mäkelä JP, Hämäläinen M, et al. Effect of interaural time
differences on middle-latency and late auditory evoked magnetic fields. Hear Res. 1994;78:249–257.)

middle-latency responses peak around 30 msec (185–188) and
are followed by longer-latency responses around 50, 100, and
200 msec (P50m, N100m, and P200m, respectively). A sus-
tained field is seen during long sounds. All these responses have
cortical origin and their generation sites differ slightly, indicat-
ing contribution from several cytoarchitectonic areas in the
supratemporal auditory cortex. Also hippocampal activity may
be seen during an auditory detection task (189). The clear
dependence of the auditory N100m response on the interstim-
ulus interval (ISI), with saturation at ISIs of 4 to 8 seconds
(190,191), has been related to the duration of sensory memory
(192–194).
Different ECD locations for N100m depending on the fre-
quency and amplitude of the sound have been suggested to
reflect tonotopic and amplitopic organization of the auditory
cortex ((195,196); see, however, the critical evaluation by
Lütkenhöner et al. (197)). The auditory cortex is very sensitive
to amplitude and frequency modulations (198) as well as to var-
ious changes within the stimuli (199,200). Direction of atten-
tion can modulate activity of at least two areas in the
supratemporal cortex (201–204). The auditory cortex also
reacts to small changes in the direction of binaural stimuli,
tested for example by changes in the interaural time difference
of binaural click trains (205,206).
The reactivity of the auditory cortex to physical changes in the
stimuli in an otherwise monotonous sequence has been studied
extensively (for reviews, see Refs. (6,50,207–210)). Infrequent
deviations in a monotonous sound sequence evoke “mismatch
fields” (MMFs) in the close vicinity of the generation site of the
auditory N100m, often with right-hemisphere dominance. MMF,
related to automatic sound processing, can be used in testing of
auditory cortical functions, including the duration of echoic
memory (194). A total omission of a tone from a monotonous
sequence activates several brain areas, including the supratempo-
ral auditory cortex, the superior temporal sulcus and the frontal
lobe; all with right-hemisphere dominance (211).
The AEFs with known sources are good tools to study brain
function. For example, Sams et al. (212) showed, applying the
so-called McGurk illusion related to speech-sound perception,
that articulation movements seen from a videotape have an
effect on AEFs evoked by acoustical syllables. It was thus
demonstrated that visual information from lip movements can
modify activity of the auditory cortex, a finding supported by
later fMRI recordings (213).
884 Part V ■ Complementary and Special Techniques
Steady-State Responses
Tonotopic organization of the human auditory cortex was first
shown by steady-state recordings (214). The amplitude of the
auditory steady-state response reaches its maximum at repeti-
tion rates around 40 Hz. This so-called 40-Hz response (215)
has a cortical origin (216,217). (To avoid confusion, it is impor-
tant to emphasize that the notion “cortical origin” refers to the
major contributor to the measured signal. It does not, by any
means, deny the possibility of a subcortical trigger or simulta-
neous activity in some subcortical areas. However, in many
cases the adequate explanation of the signal patterns by cortical
sources indicates negligible contribution from subcortical areas
to the measured MEG/EEG signals.)
The amplitude enhancement of the steady-state response of
around 40 Hz has been adequately explained by the summation
of subsequent single responses, each with a 40-Hz content,
thereby suggesting linearity of the responses, at least at stimula-
tion rates from 10 to 40 Hz (218,219). In other words, the ampli-
tude enhancement does not indicate stronger reactivity or
“resonance” of the auditory cortex at stimulation rates of 40 Hz.
However, PET data, based on quantifying regional cerebral
blood flow during repetitive auditory stimulation, suggest that
the 40-Hz response enhancement reflects increased cortical
synaptic activity (220); further studies are required because the
behavior of response strengths as a function of stimulus repeti-
tion rate differed between EEG and PET recordings.
Stimulus-related 40-Hz oscillations, elicited by paired clicks,
have been related to temporal binding (221).
During normal binaural hearing, sounds from each ear reach
the auditory cortices of both hemispheres. Therefore the binau-
ral cortical responses, even when recorded from one hemisphere
only, are unknown mixtures of inputs from both ears. One solu-
tion to this problem is to “frequency-tag” the inputs from both
ears (222): Continuous tones, presented either monaurally to
left or right ear, or binaurally, are amplitude modulated with dif-
ferent frequencies in the two ears, and the MEG signals of each
hemisphere are analyzed either in time or in frequency domain.
Such analysis demonstrates significantly stronger suppression of
the ipsilateral than contralateral input during binaural hearing.
The effect is robust despite interaural intensity differences (223).
AEFs in Patients
Ischemic lesions of the temporal lobe often cause marked defects
in speech perception and production. Details of symptomatol-
ogy cannot, however, be determined from the anatomical lesion
alone and additional methods are welcome. In patients with dif-
ferent types of ischemic brain lesions AEFs disappeared only
when the lesion extended to the deep region of the temporal
lobe (224,225). Figure 42.26 illustrates the extent of the lesion in
one such patient (226). The responses are normal in the healthy
left hemisphere but totally abolished in the lesioned left hemi-
sphere. The finding is of interest also from the source analysis
point of view: although MEG source modeling is hampered by
the nonunique inverse problem, this kind of result strongly sup-
ports the correctness of dipole modeling of normal AEFs, that is,
sources in the auditory cortices of both hemispheres.
In deaf patients with cochlear prosthesis, AEFs have been used
to show that the auditory cortex reacts to very artificial input
(227–230). AEFs also provide a tool to study the reorganization
of central auditory pathways after various peripheral lesions or
congenital disorders (231–233).
MMFs are very sensitive indicators of different sensory and
brain disorders at group level (for a review, see Ref. (209)) but,
unfortunately, the specificity of the responses has not yet been
rigorously addressed and thus the clinical usefulness of MMFs
at single-subject level remains to be shown. For example, MMFs
are altered in patients with thalamic infarction, but without
clear relationship to the degree of memory deficit (170).
Patients with Alzheimer disease have dampened cortical
40-Hz responses and delayed ipsilateral N100m responses, prob-
ably as signs of deteriorated cortical processing (53,234). The
N100m responses were delayed and dampened during auditory
hallucinations in schizophrenic patients, most likely indicating
top-down activation of the auditory projection cortices (235).
Patients with tinnitus may have a significantly decreased
amplitude ratio between P200m and N100m (236). In tinnitus
due to acute noise trauma, the P200m/N100m amplitude ratio
can normalize during the period of recovery from tinnitus
(237). However, many contradictory findings have been pre-
sented as well (238,239), most likely due to different types of
patients studied and due to variations in the experimental
setups and filter settings. In support of the early tinnitus stud-
ies, Shiomi et al. (240) showed that lidocaine-injection-induced
tinnitus release is associated with narrowing of the N100m
deflection; such a change could reflect tinnitus-related modula-
tion of neuronal mechanisms underlying the P200m response.
Tinnitus, similarly as exposure to external sounds, can be
related to plastic reorganization of the auditory cortex (241).
AEFs have been frequently studied in dyslexic individuals,
both as indicators of problems of phonologic processing and as
signs of a more general disorder in processing of rapidly pre-
sented auditory stimuli. For example, Nagarajan et al. (242)
observed that N100m to the second sound of a pair is, at short
stimulus onset asynchronies, smaller in dyslexic than normal-
reading adults. The result was interpreted to indicate abnormal
neural representation of brief and rapidly successive sensory
input, and thereby be a probable correlate of poor reading abil-
ity. Renvall and Hari (243) noted decreased responses to acoustic
frequency changes in the left auditory cortex of dyslexic adults.
Gootjes et al. (244) developed a simple paradigm to predict
language lateralization in right-handed subjects by presenting
vowels, tones, and piano notes in pairs of two stimuli.
Responses recorded when the subjects had to indicate when the
two stimuli were the same showed, in all the 11 subjects studied,
left-hemisphere dominance for responses to vowels.
SOMATOSENSORY-EVOKED FIELDS
Responses from SI
The spatial distribution of sources of somatosensory-evoked
fields (SEFs) represents well the known somatotopic organiza-
tion of the primary somatosensory cortex SI (22,23,41,245–
 Chapter 42
Figure 42.26 Auditory responses from a patient with an extended
lesion of the right temporal lobe; the shadowed areas in the brain slices
show the extent of the lesion. Responses are shown for a representative
left- and right-hemisphere channels. The stimuli were noise/ square-wave
sequences (shown on the time scale) that typically elicite one onset
response (N100m) and one transition-related response N100m. (Adapted
from Mäkelä JP, Hari R. Neuromagnetic auditory evoked responses after
a stroke in the right temporal lobe. Neuroreport. 1992;3:94–96.)
247). Since MEG mainly picks signals from tangential currents,
the convexial cortex is either silent or poorly represented in the
SEF distributions. At SI, the hand-area signals should therefore
arise mainly from cytoarchitectonic area 3b, whereas in the foot
SI all cytoarchitectonic areas should produce currents tangen-
tial to the skull and thus their changed relative contributions
should be seen as rotation of the field patterns. Evoked field
data to lower-limb stimulation suggest that such rotations really
occur as signs of successive activation of several cytoarchitec-
tonic areas (248–252).
Figure 42.27 shows one subject’s responses to left tibial-
nerve stimuli, with largest signals at the top of the head (251).
The field patterns rotate as a function of time, as predicted
above. A two-dipole model, with one dipole in area 3b and the
other in area 5, both in the mesial wall of the hemisphere,
explained the data during the first 100 msec, with a consistent
3-msec time difference in their initial peak latencies. Other
source models were rejected on the basis of their inconsisten-
cies with experimental data and/or anatomical information. A
slightly different two-source model for tibial-nerve SEFs was
suggested by Hashimoto et al. (252) on the basis of beam-
former analysis.
SEFs from the SI cortex are typically strictly contralateral.
However, tactile interference applied on the palm suppresses
■ Magnetoencephalography: Methods and Applications 885
the median-nerve 30-msec response at the contralateral SI but
enhances it in the ipsilateral SI (253), implying that tactile input
from one hand has access to the SI cortices of both hemi-
spheres, probably through callosal connections. Some long-
latency MEG responses from the ipsilateral SI have been
reported (254). However, transfer of stimulus- or movement-
related vibrations to the other hand has been demonstrated as
one possible cause of ipsilateral responses (255).
Interesting changes occur during the development in the
latencies, relative amplitudes, and polarities of the main SEF
deflections (Fig. 42.28) (256).
Brief bursts of about 600-Hz oscillations are superimposed
on the earliest cortical deflections, the 20-msec N20m response
(257,258). These high-frequency responses will be discussed in
detail in Chapter 37.
Quantification of source strengths may be very useful in
some applications. For example, Elbert et al. (259), determining
source strengths for tactile finger stimulation at SI cortex, found
increased cortical representations for the left hands of string
players, implying practice-related cortical reorganization (see
also Ref. (260)).
Figure 42.29 illustrates the use of SEFs in preoperative iden-
tification of the sensorimotor strip. Superposition of sources
for median and tibial nerve stimulation on the patient’s MRI
illustrates that her cavernous angioma is about 1.5 cm away
from the somatomotor hand area. The angioma was later suc-
cessfully operated, without causing any sensory or motor
impairments.
In Figure 42.30, the preoperative evaluation includes, in
addition to SEF recordings, identification of the primary motor
cortex, anterior to the central sulcus, by means of cortex–mus-
cle coherence (see Fig. 42.17 and the related text). Visualization
of the somatosensory and motor landmarks on a brain render-
ing with veins superimposed (right) helps the neurosurgeon to
navigate during the operation (58).
Responses from SII and Other Nonprimary
Cortical Areas
MEG recordings have indicated that peripheral stimuli also
activate bilaterally areas close to the upper lip of the Sylvian fis-
sure, that is, most probably at the second somatosensory area
SII (23,45,261,262). Although the SI/SII differentiation was
possible already with small MEG sensor arrays, the more recent
whole-scalp recordings have given more detailed information
about the cortical somatosensory network, with spatially and
temporally distinct activations also in the posterior parietal cor-
tex, in the mesial cortex, and in the frontal lobes (12,24–26).
The field patterns of Figure 42.31 illustrate activation of the
right (contralateral) SI, left and right SII, and the posterior pari-
etal cortex after electric stimulation of the left thumb. Tactile
stimulation activates essentially the same brain regions as does
electric stimulation, but with slightly different latencies and rel-
ative amplitudes (263). The MEG source locations in SI and SII
cortices agree with fMRI data collected from the same subjects
(264,265).
The representations of fingers of both hands overlap
strongly in the bilaterally activated SII cortices, whereas the SI
886 Part V ■ Complementary and Special Techniques
Figure 42.27 Top: Responses of one subject to electric stimulation of the
left tibial nerve at the ankle. The traces are averages of 800 single
responses. The inset shows two enlarged sets of 400 responses from suc-
cessive experiments superimposed. The passband is from 0.03 to 275 Hz.
Bottom: Magnetic field patterns and equivalent current dipoles of the same
subject as a function of time (indicated in msec). The isocontours are sep-
arated by 20 fT. (Adapted from Hari R, Nagamine T, Nishitani N, et al.
Time-varying activation of different cytoarchitectonic areas of the human
SI cortex after tibial nerve stimulation. NeuroImage. 1996;4:111–118.)
cortex (area 3b) has representations of fingers of one hand only
and with much weaker overlap (266). Bilateral median-nerve
stimulation elicits left-hemisphere-dominant SII activation,
suggesting handedness-independent functional specialization
of the human SII cortices (267).
Differentiation of various source areas of SEFs has started to
enlighten functions of the cortical somatosensory network
(12,23,159,268) and recordings of SEFs in patients with various
lesions (259), and analysis of, for example, phase locking
between areas (270) may bring new information about the
debate on parallel versus serial processing in different
somatosensory cortical areas.
SEFs in Multiple Sclerosis
In patients with clinically diagnosed multiple sclerosis (271),
the N20m deflection was dampened and delayed by 2.5 to
3 msec. The 50- to 80-msec responses were often of larger
amplitude than in control subjects, especially in patients who
had periventricular plaques in MRI. The source locations of
N20m and P60m in the somatosensory hand area did not differ
between the groups.
It is probable that enhancement of the corresponding elec-
tric SEPs has remained undetected with the typical high-pass
filter settings at 10 to 20 Hz and the analysis periods of 30 to
50 msec applied in many laboratories.
SEFs in Myoclonus Epilepsy
Karhu et al. (272) suggested that progressive myoclonus
epilepsy (PME; Unverricht-Lundborg type) is associated with
thalamocortical hyperreactivity in the sensorimotor but not in
the auditory system. The patients were adequately medicated
and their MEG did not show spikes. The amplitudes of the
30-msec SEFs were up to sixfold compared with the response of
normal subjects, but the source locations agreed with the acti-
vation of the somatosensory hand area. Patients of juvenile
neuronal ceroid lipofuscinosis (JNCL) and late INCL (CLN5)
also show enhanced SEFs (273,274).
“Giant” SEFs from SI have been recorded also from a genet-
ically homogeneous group of PME (Unverricht-Lundborg)
patients (275). Figure 42.32 shows, as a sign on hyperreactivity,
strongly enhanced responses at the contralateral SI of a patient.
Responses are also seen in the ipsilateral SI, in contrast to the
healthy control subject; the latency difference of about 20 msec
suggests facilitated callosal transfer to the ipsilateral side.
Moreover, despite the enhanced SI responses, no clear SII
responses are seen in the patient. Findings like this encourage
the clinical neurophysiologist to assess functions of the
somatosensory circuitry beyond the primary somatosensory
cortex, as the dysfunctions may have direct consequences on the
symptoms, prognosis, and/or rehabilitation of the patient.
Noxious Stimulation
MEG responses to painful stimuli imply activation around SII
and/or the frontal operculum to electric stimulation of the
tooth pulp, carbon dioxide stimulation of the nasal mucosa,
and noxious stimulation (by means of electric current or CO2
or thulium laser) of the skin (276–283). Whole-scalp record-
ings indicated a clear right- hemisphere predominance to
painful CO 2 stimulation of the nasal mucosa (284).
With laser heat it is possible to stimulate rather selectively
A - and C-fibers by applying different stimulus intensities to
stimulus areas of different sizes (280,283,284). Figure 42.33
shows that the evoked responses from the contralateral SII cor-
tex peak to A -fiber stimuli at about 160 msec and to C-fiber
stimuli around 800 msec, in agreement with the conduction
velocities of these two fiber types (283). Interestingly, monitor-
ing of the level of the 20-Hz motor–cortex activity strongly sug-
gests that the painful laser stimuli automatically activate the
primary motor cortex and therefore might be informative
 Chapter 42 ■ Magnetoencephalography: Methods and Applications 887

–50 0 50 200 ms –50 0 50 200 ms

50 fT/cm

N20m P30m
M30 M60

Newborn Adult

Figure 42.28 Somatosensory-evoked fields and field patterns to median nerve stimulation in a newborn and an adult.
The signals are from one planar gradiometer channel over the right somatosensory cortex. (Adapted from Lauronen L,
Nevalainen P, Wikström H, et al. Immaturity of somatosensory cortical processing in human newborns. Neuroimage.
2006;33:195–203.) The photograph on the left illustrates how with an adult-shape neuromagnetometer, just one hemi-
sphere at the time can be recorded from the small baby brain. (Courtesy of Elina Pihko.) (See color insert)

Figure 42.29 Left: Sources for responses to left

median and tibial nerve stimulation superposed
on the MRI reconstruction of the subject. The
section passes through both sources to illustrate
the central sulcus just anterior to them. Right: A
parallel section slightly deeper, with the cav-
ernous angioma indicated. The patient, a 50-
year-old female, was later successfully operated.

Figure 42.30 Preoperative functional localiza-

tion in a patient with a parietal-lobe tumor. The
postcentral somatosensory cortex has been iden-
tified by measuring SEFs to median-nerve (hand)
and tibial-nerve (foot) stimulation. In addition,
cortex–muscle coherence for hand and foot small
muscles has been used to identify the precentral
primary motor cortex. On the right, veins are
superimposed on the surface rendering.
(Courtesy of CliniMEG, Brain Research Unit, Low
Temperature Laboratory, Aalto University School
of Science and Technology.)
888 Part V ■ Complementary and Special Techniques
Figure 42.31 Field patterns of SEFs to electric stimulation of the
subject’s left thumb (ISI 4 seconds). Illustrative responses are shown
from four areas (contralateral SI, both SII regions, and the PPC; panels
A, B, C, and D, respectively) and the corresponding field patterns at the
time of maximum response from each region. (Adapted from Forss N,
Jousmäki V, Hari R. Interaction in the human somatosensory cortex
between afferent input from different fingers. Brain Res. 1995;685:68–76.)
about the mechanisms underlying pathological muscle contrac-
tion in tension pain (286–288). In patients with chronic pain,
enhanced activation of the motor cortex occurs after innocuous
tactile stimuli (289). MEG recordings have also demonstrated
the spread of a pain-related disorder of somatosensory process-
ing from one hemisphere to another in patients suffering from
complex regional pain syndrome (290).
Visual-Evoked Fields
Visual-evoked fields (VEFs) were reported for the first time in
1975 (291,292). With flash stimuli, VEFs and VEPs did not peak
simultaneously and the responses depended on luminance in
different manner, thereby suggesting separate sources for VEFs
and VEPs (292). Steady-state recordings with gratings reversing
at 13 Hz (291) allowed for the determination of the sources:
The field patterns evoked by half-field stimuli agreed with acti-
vation of the contralateral occipital cortex and the patterns to
full-field stimulation were the sums of the two half-field
patterns. The retinotopic organization of the occipital visual
cortex has been studied with MEG by presenting stimuli to dif-
ferent parts of the visual field (293–296).
Outside the occipital cortex the visual field patterns are very
complex and the utilization of VEFs as tools of clinical neuro-
physiology is still scarce. Figure 42.34 shows responses to
half-field pattern and luminance stimuli. The first responses are
generated in the occipital cortex but a later response occurs,
especially for luminance stimuli, in the region of the parieto-
occipital sulcus (POS).
The very same POS area close to the midline can be activated
by various visual stimuli (295) as well as by voluntary eyeblinks
(297,298) and saccades (299); the finding has been confirmed
in functional magnetic resonance recordings. Anatomically, the
posterior parietal midline region, which also produces the most
prominent MEG alpha activity, could well be the human homo-
logue of the monkey V6/V6A region, influenced by, for exam-
ple, saccades and eye position and proposed to be related to
spatial coding of extrapersonal visual space and to visuomotor
integration (300). The V6/POS region is a part of the dorsal
visual stream and thus of interest for both basic neuroscience
and for some neurological disorders affecting visuomotor
coordination.
The human homologue of the monkey MT/V5 cortex, a
brain area selective to visual motion, has also been identified by
MEG recordings (301–304). Better understanding of neural ori-
gin of different VEF components brings invaluable information
about temporal relationships in the activation of different
visual areas. Interestingly, the activation order of human visual
cortices can deviate from the simple serial activation pattern
(305), in agreement with both monkey recordings (306) and
simulation studies (307), thereby indicating the importance of
reciprocal interareal connections in a hierarchically connected
network.
Responses to Faces and Language Stimuli
In agreement with both clinical and other imaging data, MEG
responses to pictures of faces indicate activation of extrastriate
visual areas, predominantly in the right posterior fusiform
gyrus (308–311). Face-categorization-related activity has been
reported as early as 100 msec after the stimulus (312), whereas
the typical face-specific response peaks at 140 to 170 msec (60),
and the activation shows a systematic amplitude dependence on
the degree of “faceness” of the stimulus (311).
MEG’s excellent temporal resolution is of high value in
studies of language perception and production. Salmelin et al.
(313) were able to follow the cortical activation sequence
from the posterior brain areas toward more anterior regions
when the subjects named pictures of objects. Signs of consid-
erable parallel processing were evident and both hemispheres
were activated by the task. Interestingly, MEG studies of
incongruities in musical syntax suggest that Broca’s region
and its right-hemisphere homologue area process syntactic
information that is less language-specific than usually
believed (314).
More recently, MEG has been increasingly applied for stud-
ies of reading and naming, both in healthy subjects (315) and in
 Chapter 42 ■ Magnetoencephalography: Methods and Applications 889

Figure 42.32 Source locations (left) and source waveforms as a function of time (right) to electric stimulation of the left
median nerve in a healthy control subject and in a patient suffering from Unverricht-Lundborg-type progressive myoclonus
epilepsy. The black circles indicate source locations and the white circles indicate possible source locations that did not
show any activation; c and I refer to contralateral and ipsilateral hemispheres. (Adapted from Forss N, Silén T, Karjalainen
T. Lack of activation of human secondary somatosensory cortex in Unverricht-Lundborg type of progressive myoclonus
epilepsy. Ann Neurol. 2001;49:90–97.)

Figure 42.33 Top left: Averaged responses (N = 100)

of a single subject from right SII region to selective A -
(upper trace; weak pricking pain) and C-fiber (lower
trace; weak burning pain) thulium-laser stimuli pre-
sented to the dorsum of the left hand once every 4.5 to
5.5 seconds. Top right and bottom: The mean source
locations of responses in the SII (top) and PPC (bot-
tom) region to A -fiber stimuli. The ovals indicate the
mean SEM locations across eight subjects superim-
posed on slices of a mean brain of the subjects, calcu-
lated by means of elastic transformations of their
magnetic resonance images. (Modified from Forss N,
Raij TT, Seppä M, et al. Common cortical network for
first and second pain. Neuroimage. 2005;24:132-142.)
890 Part V ■ Complementary and Special Techniques

Figure 42.34 Left: Responses to stimulation of the right visual hemifield by luminance stimuli. Right: Responses of the
same subject to left and right hemifield (LVF, RVF) luminance and pattern stimuli (radius 8 deg) from the parietal region
(Par) and from left and right occipital areas (Locc, Rocc). Note the predominance of the parietal response to luminance
stimuli and the very clear LVF versus RVF difference in the occipital channels for pattern, but not for luminance, stimuli.
(Adapted from Portin K, Salenius S, Salmelin R, et al. Activation of the human occipital and parietal cortex by pattern and
luminance stimuli: neuromagnetic measurements. Cereb Cortex. 1998;18:253–260.)

aphasic patients (316), stuttering subjects (317), as well as in
dyslexic individuals (318–320). New interesting approaches
include the analysis of long-range connectivity (321) and the
study of language development (322).
Saccade-Related Activity
Figure 42.35 shows saccade-related MEG signals, averaged on
the basis of the saccade onsets (299). The MEG channels close
to the orbits record eyeball-related signals. Two main signal
maxima were observed, one in the parietal midline and another
in the lowest posterior channels. Source modeling suggested
that the former response was generated in the POS region and
the latter signal in cerebellar vermis. Both source activations
peaked about 170 msec after the saccade onset, and the cerebel-
lar signal started already 30 msec before the saccade. The POS
signal was abolished when the lights of the experimental room
were turned off whereas the cerebellar response was only
slightly delayed. The possibility to noninvasively follow the

Figure 42.35 Magnetoencephalographic signals

from three locations (shown on the schematic
head) during a visually-guided saccade task. The
signals were averaged with respect to the onset of
horizontal saccades (leftwards or rightwards);
the horizontal EOG used as the trigger is depicted
above. Signals recorded during darkness and
with lights on are illustrated with dashed and
solid lines, respectively. (Adapted from Jousmäki
V, Hämäläinen M, Hari R. Magnetic source imag-
ing during a visually guided task. Neuroreport.
1996;7:2961–2964.)
 Chapter 42 ■ Magnetoencephalography: Methods and Applications 891

A B
12 Hz

Spectral density
12 Hz 30

(fT/cm/ Hz)
15 Hz 15 Hz
20

10

0 10 20 30
Frequency (Hz)
Time (s)
–2 –1 0 1 2
C

Frequency (Hz)
15 Hz (face-tag) 20

10
12 Hz (vase-tag)

PERCEPT “vase” “faces”

Figure 42.36 An example of frequency tagging. (A) The Rubins vase/ faces figure used as the stimulus. Dynamical noise
was superimposed to the image and updated at 12 Hz in the vase area and at 15 Hz in the face areas. (B) Frequency
spectrum from one occipital magnetoencephalogram (MEG) channel, based on 10-minute recording. (C) Time–frequency
representation of MEG activity computed time-locked to the change in percept, indicated by button press. (Adapted
from Parkkonen L, Andersson J, Hämäläinen M, et al. Early visual areas reflect the percept of an ambiguous scene.
Proc Nat Acad Sci USA. 2008;105:20500–20504.)(See color insert)

temporal behavior of human cerebellar activation may open
new windows to the monitoring of neurologic patients. Early
cerebellar signals related to somatosensory stimulation have
been observed (323), as have oscillations anticipating omissions
of somatosensory stimuli (324).
Cortical Correlates of Perceptual Sw itches
Some percepts change so unexpectedly that it is not possible to
time-lock the analysis of MEG signals to the event. Figure 42.36
shows how frequency tagging can help to follow cortical activa-
tion changes related to the spontaneous switching of percepts of
the Rubin’s vase (325). Dynamic noise was superimposed to the
picture and updated at 12 Hz in the vase region and at 15 Hz in
the face region. The subjects were not able to tell the difference
between the update frequencies, but their occipital MEG signals
displayed distinct frequency peaks both at 12 and 15 Hz.
On the basis of the changes in the relative strengths of the
12- and 15-Hz frequencies during the percepts it was possible to
conclude that the activation of the early visual cortices covaried
with the percept, most likely because of top-down influences.
Eye tracking ruled out the possibility that the changes in the
MEG signals could have been related to gaze deviations (325).
Human Mirror-Neuron System
MEG recordings have been recently used for studies of the
human mirror-neuron system (MNS). Mirror neurons were
first described in monkey premotor cortical area F5 (326):
the neurons discharged both when a monkey executed a hand
action and when it observed a similar action made by another
monkey or by the experimenter. The mirror neurons have
been assumed to match action observation and execution
(for a review, see Ref. (327)). In humans, the MNS comprises
at least the primary motor cortex (see Fig. 42.16), inferior
frontal cortex (Broca’s area in the left hemisphere), and infe-
rior parietal lobe, with contribution from the superior
temporal sulcus that does not contain proper mirror neu-
rons. These areas are activated in a specific order within 250
to 400 msec during viewing of other person’s motor acts
(328,329). Moreover, the primary and secondary somatosen-
sory cortices react to observed hand movements. The pri-
mary motor cortex is activated before and during observed
actions (that the viewer already more or less knows) and is
stabilized afterwards, very similarly as happens during the
person’s own actions; these conclusions are based on the
reactivity of the motor-cortex 20-Hz rhythm (330). The human
MNS may play an important role both in action imitation
and in understanding the meaning of actions made by other
subjects, thereby also having relevance for social interaction
(for reviews, see Refs. (332,333)).
CONCLUSIONS
MEG has established a position as a tool of basic neuroscience,
and clinical applications have started to rapidly develop when
neuromagnetometers with whole-scalp coverage have been
installed in hospital environments. During the last few years,
combination of functional information, deduced from MEG,
and structural information, obtained from MRI, has become a
892 Part V ■ Complementary and Special Techniques
routine, and coregistration of MEG and fMRI data have started
to become popular as well. Instead of instrumental develop-
ment, which already has reached a very user-friendly stage,
the efforts of the neuromagnetism community are now
increasingly focusing on development of new signal analysis
approaches (14).
MEG’s clinical applications in neurology, neurosurgery, and
audiology are expanding rapidly and the method will probably
be useful also in developmental neuroscience and in neuropsy-
chiatry. In addition to conventional studies of the auditory,
visual, and somatomotor systems, stimulation techniques are
available to obtain selective MEG responses also from olfactory
and gustatory systems, and reliable data have been obtained
with real-life-like stimulation. With the whole-scalp neuromag-
netometers and with improved head-position monitoring tools
as well as sophisticated artifact suppression methods, MEG can
be applied in a variety of behavioral conditions, designed for
assessment of neural basis of cognitive functions.
MEG allows studies of the temporal activation order of
several cortical areas, thereby offering new noninvasive infor-
mation of brain dynamics and temporal hierarchies (334). A
clinical neurophysiologist naturally looks for practical applica-
tions, appropriate to the functional assessment of sensory path-
ways and the cortex. In this respect, MEG’s ability to locate
sources of various electric signals that are frequently measured
with scalp EEG is very helpful. Better knowledge of cortical-
evoked responses and spontaneous rhythms will further widen
the clinical scope of both EEG and MEG recordings. Necessity
to map and interpret complex dynamical activation sequences
of cortical circuitries requires the future clinical neurophysiol-
ogists to obtain good understanding about modern electro-
physiologic and imaging methods, but also about systems-level
basic and cognitive neuroscience.
ACKNOWLEDGMENTS
This study has been supported by the Academy of Finland and
the Sigrid Jusélius Foundation. The skill, support, and enthusi-
asm of the members of the Brain Research Unit of the Low
Temperature Laboratory, Helsinki University of Technology,
have been decisive for collecting the MEG data presented in this
chapter.
REFERENCES
1. Cohen D. Magnetoencephalography: evidence of magnetic field
produced by alpha-rhythm currents. Science. 1968;161:784–786.
2. Hari R, Lounasmaa OV. Recording and interpretation of cerebral
magnetic fields. Science. 1989;244:432–436.
3. Grandori F, Hoke M, Romani G-L, eds. Auditory Evoked Magnetic
Fields and Potentials. Advances in Audiology, Vol. 6. Basel:
Karger; 1990.
4. Sato S, eds. Magnetoencephalography. Advances in Neurology. Vol.
54. New York: Raven Press; 1990.
5. Hämäläinen M, Hari R, Ilmoniemi R, et al. Magnetoence-
phalography—theory, instrumentation, and applications to non-
invasive studies of the working human brain. Rev Mod Phys.
1993;65:413–497.
6. Näätänen R, Ilmoniemi R, Alho K. Magnetoencephalography in
studies of human cognitive brain function. Trends Neurosci.
1994;17:389–395.
7. Baumgartner C, Deecke L, Stroink G, et al., eds. Biomagnetism:
Fundamental Research and Clinical Applications. Amsterdam:
Elsevier; 1995.
8. Hashimoto I, Okada YC, Ogawa S, eds. Visualization of Information
Processing in the Human Brain. Amsterdam: Elsevier; 1996.
9. Lounasmaa OV, Hämäläinen M, Hari R, et al. Information pro-
cessing in the human brain – magnetoencephalographic approach.
Proc Natl Acad Sci USA. 1996;93:8809–8815.
10. Aine C, OkadaYC, Stroink G, et al., eds. Advances in Biomagnetism:
Biomag96. Springer Verlag; 1997.
11. Hari R, Salmelin R. Human cortical rhythms: a neuromagnetic
view through the skull. Trends Neurosci. 1997;20:44–49.
12. Hari R, Forss N. Magnetoencephalalography in the study of
human somatosensory cortical processing. Proc Royal Soc Lond B.
1999;354:1145–1154.
13. Forss N, Nakasato N, Ebersole J, et al. Clinical use of magnetoen-
cephalography. Suppl Clin Neurophysiol. 2000;53:287–297.
14. Baillet S, Mosher JC, Leahy RM. Electromagnetic brain mapping.
IEEE Signal Proc Magaz. 2001;18:14–30.
15. Del Gratta C, Pizzella V, Tecchio F, et al. Magnetoencephalography –
a noninvasive brain imaging method with 1 ms time resolution.
Rep Prog Phys. 2001;64:1759–1814.
16. Hämäläinen M, Hari R. Magnetoencephalographic characteriza-
tion of dynamic brain activation: basic principles, and methods
of data collection and source analysis. In: Toga AW, Mazziotta JC,
eds. Brain Mapping. The Methods, 2nd ed. Amsterdam: Academic
Press; 2002:227–253.
17. Lu ZL, Kaufman L, eds. Magnetic Source Imaging of the Human
Brain. Mahwah, New Jersey: Lawrence Erlbaum Associates; 2003.
18. Hansen P, Kringelbach M, Salmelin R, eds. MEG. An Introduction
to Methods. New York: Oxford University Press; 2010.
19. Pernier J, Perrin F, Bertrand O. Scalp current density fields: con-
cept and properties. Electroencephalogr Clin Neurophysiol. 1988;69:
385–389.
20. Nunez PL. Estimation of large scale neocortical source activity
with EEG surface Laplacians. Brain Topogr. 1989;2:141–154.
21. Cohen D, Cuffin B. Demonstration of useful differences between
MEG and EEG. Electroencephalogr Clin Neurophysiol. 1983;56:38–51.
22. Hari R, Reinikainen K, Kaukoranta E, et al. Somatosensory evoked
cerebral magnetic fields from SI and SII in man. Electroencephlogr
Clin Neurophysiol. 1984;57:254–263.
23. Hari R, Karhu J, Hämäläinen M, et al. Functional organization of
the human first and second somatosensory cortices: a neuromag-
netic study. Eur J Neurosci. 1993;5:724–734.
24. Forss N, Hari R, Salmelin R, et al. A novel source area in the human
parietal cortex activated by median nerve stimulation. Exp Brain
Res. 1994;99:309–315.
25. Forss N, Merlet I, Vanni S, et al. Activation of human mesial cortex
during somatosensory attention task. Brain Res. 1996;734:229–235.
26. Mauguière F, Merlet I, Forss N, et al. Activation of a distributed
somatosensory cortical network in the human brain. A dipole
modelling study of magnetic fields evoked by median nerve stim-
ulation. Part I: location and activation timing of SEF sources.
Electroencephalogr Clin Neurophysiol. 1997;104:281–289.
27. Hari R. On brain’s magnetic responses to sensory stimuli. J Clin
Neurophysiol. 1991;8:157–169.
28. Parkkonen L, Fujiki N, Mäkelä J. Sources of auditory brainstem
responses revisited: contribution by magnetoencephalography.
Hum Brain Mapp. 2009;30:1772–1782.
 Chapter 42
29. Tesche CD. Non-invasive imaging of neuronal population dynam-
ics in human thalamus. Brain Res. 1996;729:253–258.
30. Tesche CD. Non-invasive detection of ongoing neuronal population
activity in normal human hippocampus. Brain Res. 1997;749:53–60.
31. Numminen J, Ahlfors S, Ilmoniemi R, et al. Transformation of
multichannel magnetoencephalographic signals to standard grid
form. IEE Trans Biomed Eng. 1995;42:72–77.
32. Barth D, Sutherling W, Engel JJ, et al. Neuromagnetic localization
of epileptiform spike activity in the human brain. Science.
1982;218:891–894.
33. Duret D, Karp P. Figure of merit and spatial resolution of super-
conducting flux transformers. J Appl Phys. 1984;56:1762–1768.
34. Romani G, Leoni R. Localization of cerebral sources by neuromag-
netic measurements. In: Weinberg H, Stroink G, Katila T, eds.
Biomagnetism: Applications and Theory. Oxford: Pergamon;
1985:205–220.
35. Curio G, Drung D, Koch H, et al. Magnetometry of evoked fields
from human peripheral nerve, brachial plexus and primary
somatosensory cortex using a liquid nitrogen cooled superconduct-
ing quantum interference device. Neurosci Lett. 1996;206:204–206.
36. Kominis IK, Kornack TW, Allred JC, et al. A subfemtotesla multi-
channel atomic magnetometer. Nature. 2003;422:596–599.
37. Pannetier M, Fermon C, Le Goff G, et al. Femtotesla magnetic field
measurement with magnetoresistive sensors. Science. 2004;304:
1648–1650.
38. Hillebrand A, Barnes GR. A quantitative assessment of the sensi-
tivity of whole-head MEG to activity in the adult human cortex.
Neuroimage. 2002;16:638–650.
39. Sutherling W, Crandall P, Darcey T, et al. The magnetic and electric
fields agree with intracranial localizations of somatosensory cor-
tex. Neurology. 1988;38:1705–1714.
40. Tiihonen J, Hari R, Hämäläinen M. Early deflections of cerebral
magnetic responses to median nerve stimulation.
Electroencephalogr Clin Neurophysiol. 1989;74:290–296.
41. Baumgartner C. Clinical Neurophysiology of the Somatosensory
Cortex. Wien: Springer Verlag; 1993
42. Lounasmaa OV, Williamson SJ, Kaufman L, et al. Visually evoked
responses from non-occipital areas of human cortex. In: Weinberg
H, Stroing G, Katila T, eds. Biomagnetism: Applications & Theory.
New York: Pergamon Press; 1985:348–353.
43. Tesche CD, Uusitalo MA, Ilmoniemi RJ, et al. Signal-space projec-
tions of MEG data characterize both distributed and well-localized
neuronal sources. Electroencephalogr Clin Neurophysiol. 1995;95:
189–200.
44. Uusitalo M, Ilmoniemi R. Signal-space projection method for sep-
arating MEG and EEG into components. Med Biol Eng Comput.
1997;35:135–140.
45. Kaukoranta E, Hämäläinen M, Sarvas J, et al. Mixed and sensory
nerve stimulations activate different cytoarchitectonic areas in the
human primary somatosensory cortex SI. Exp Brain Res.
1986;63:60–66.
46. Hari R, Joutsiniemi SL, Sarvas J. Spatial resolution of neuromag-
netic records: theoretical calculations in a spherical model.
Electroencephalogr Clin Neurophysiol. 1988;71:64–72.
47. Scherg M. Fundamentals of dipole source potential analysis. In:
Grandori F, Hoke M, Romani GL, eds. Auditory Evoked Magnetic
Fields and Potentials. Advances in Audiology. Vol. 6. Basel: Karger;
1990:40–69.
48. Vvedensky V, Hari R, Ilmoniemi R, et al. Physical basis of neuro-
magnetic fields. Biophysica. 1985;30:154–158.
49. Okada YC, Papuashvili N, Xu C. Maximum current dipole
moment density as an important physiological constraint in MEG
■ Magnetoencephalography: Methods and Applications 893
inverse solutions. Tenth International Conference on Biomagnetism.
Santa Fe, NM, Volume of Abstracts; 1996:149.
50. Hari R. The neuromagnetic method in the study of the human
auditory cortex. In: Grandori F, Hoke M, Romani G, eds. Auditory
Evoked Magnetic Fields and Potentials. Advances in Audiology.
Vol. 6. Basel: Karger; 1990:222–282.
51. Bernander O, Douglas R, Martin K, et al. Synaptic background
activity determined spatial-temporal integration in pyramidal
cells. Proc Natl Acad Sci USA. 1991;88:11569–11573.
52. Hämäläinen M, Ilmoniemi R. Interpreting magnetic fields of the
brain: minimum norm estimates. Med Biol Eng Comput.
1994;32:35–42.
53. Ribary U, Ioannides A, Singh K, et al. Magnetic field tomography
of coherent thalamocortical 40-Hz oscillation in humans. Proc
Natl Acad Sci USA. 1991;88:11037–11041.
54. Volkmann J, Joliot M, Mogilner A, et al. Central motor loop oscil-
lations in parkinsonian resting tremor revealed by magnetoen-
cephalography. Neurology. 1996;46:1359–1370.
55. Uutela K, Hämäläinen M, Somersalo E. Visualization of magne-
toencephalographic data using minimum current estimates.
Neuroimage. 1999;10:173–180.
56. Stenbacka L, Vanni S, Uutela K, et al. Comparison of minimum cur-
rent estimate and dipole modeling in the analysis of simulated acti-
vity in the human visual cortices. Neuroimage. 2002;16:936–943.
57. Jack C, Marsh W, Hirschorn K, et al. EEG scalp electrode projec-
tions onto three-dimensional surface rendered images of the
brain. Radiology. 1990;176:413–418.
58. Mäkelä J, Kirveskari E, Seppä M, et al. Three-dimensional integration
of brain anatomy and function to facilitate intraoperative navigation
around the sensorimotor strip. Hum Brain Mapp. 2001;12:181–192.
59. Aine CJ. A conceptual overview and critique of functional neu-
roimaging techniques in humans: I. MRI/fMRI and PET. Crit Rev
Neurobiol. 1995;9:229–309.
60. Furey ML, Tanskanen T, Beauchamp MS, et al. Dissociation of face
selective cortical responses by attention. Proc Natl Acad Sci USA.
2006;103:1065–1070.
61. Nangini C, Hlushchuk , Hari R. Predicting stimulus-rate sensitivity
of human somatosensory fMRI signals by MEG. Human Brain
Mapp. 2009;30:1824–1832.
62. Hari R, Ilmoniemi RJ. Cerebral magnetic fields. CRC Crit Rev
Biomed Eng. 1986;14:93–126.
63. Tarkiainen A, Liljeström M, Seppä M, et al. The 3D topography of
MEG source localization accuracy: effects of conductor model and
noise. Clin Neurophysiol. 2003;114:1977–1992.
64. Hämäläinen MS, Sarvas J. Realistic conductivity geometry model
of the human head for interpretation of neuromagnetic data. IEEE
Trans Biomed Eng. 1989;36:165–171.
65. Cuffin B. Effects of fissures in the brain on electroencephalograms
and magnetoencephalograms. J Appl Phys. 1985;57:146–153.
66. Cooper R, Winter A, Crow H, et al. Comparison of subcortical,
cortical and scalp activity using chronically indwelling electrodes
in man. Electroencephalogr Clin Neurophysiol. 1965;18:217–228.
67. Ilmoniemi R. Radial anisotropy added to a spherically symmetric
conductor does not affect the external magnetic field due to inter-
nal sources. Europhys Lett. 1995;30:313–316.
68. Antervo A, Hari R, Katila T, et al. Magnetic fields produced by eye
blinking. Electroencephalogr Clin Neurophysiol. 1985;61:247–254.
69. Jousmäki V, Hari R. Cardiac artifacts in magnetoencephalogram.
J Clin Neurophysiol. 1996;13:172–176.
70. Samonas M, Petrou M, Ioannides A. Identification and elimina-
tion of cardiac contribution in single-trial magnetoencephalo-
graphic signals. IEEE Trans Biomed Eng. 1997;44:386–393.
894 Part V ■ Complementary and Special Techniques
71. Taulu S, Simola J, Kajola M. Applications of the signal space sepa-
ration method. IEEE Trans Sign Proc. 2005;53:3359–3372.
72. Taulu S, Simola J. Spatiotemporal signal space separation method
for rejecting nearby interference in MEG measurements. Phys Med
Biol. 2006;51:1759–1768.
73. Taulu S, Hari R. Removal of MEG artifacts with temporal signal-
space separation: demonstration with single-trial auditory evoked
responses. Hum Brain Mapp. 2008;30:1524–1534.
74. Mäkelä JP, Taulu S, Pohjola J, et al. Effects of subthalamic nucleus
stimulation on spontaneous sensorimotor MEG activity in a
Parkinsonian patient. Int Congr Ser. 2007;1300:345–348.
75. Mertens M, Lütkenhöner B. Effective neuromagnetic determina-
tion of landmarks in the somatosensory cortex. Clin Neurophysiol.
2000;111:1478–1487.
76. Jousmäki V, Nishitani N, Hari R. Brush stimulator for functional
brain imaging. Clin Neurophysiol. 2007;118:2620–2624.
77. Lütkenhöner B. Frequency-domain localization of intracerebral dipo-
lar sources. Electroencephalogr Clin Neurophysiol. 1992;82:112–118.
78. Tesche C, Kajola M. A comparison of the localization of sponta-
neous neuromagnetic activity in the frequency and time domains.
Electroencephalogr Clin Neurophysiol. 1993;87:408–416.
79. Salmelin R, Hämäläinen M. Dipole modelling of MEG rhythms in
time and frequency domains. Brain Topogr. 1995;7:251–257.
80. Salmelin R, Hari R. Spatiotemporal characteristics of rhythmic
neuromagnetic activity related to thumb movement. Neuroscience.
1994;60:537–550.
81. Pfurtscheller G, Aranibar A. Event-related desynchronization
detected by power measurements of scalp EEG. Electroenceph Clin
Neurophysiol. 1977;42:138–146.
82. Barth DS, Sutherling W, Engel J, et al. Neuromagnetic evidence of
spatially distributed sources underlying epileptiform spikes in the
human brain. Science. 1984;223:293–296.
83. Stefan H, Schneider S, Abraham-Fuchs K, et al. The neocortico to
mesio-basal limbic propagation of focal epileptic activity during
the spike-wave complex. Electroencephalogr Clin Neurophysiol.
1991;79:1–10.
84. Chapman RM, Ilmoniemi RJ, Barbanera S, et al. Selective localiza-
tion of alpha brain activity with neuromagnetic measurements.
Electroencephalogr Clin Neurophysiol. 1984;58:569–572.
85. Vvedensky V, Ilmoniemi R, Kajola M. Study of the alpha rhythm
with a 4-channel SQUID magnetometer. Med Biol Eng Comput.
1986;23:11–12.
86. Williamson S, Wang J-Z, Ilmoniemi R. Method for locating
sources of human alpha activity. In: Williamson S, Hoke M,
Stroink G, et al., eds. Advances in Biomagnetism. New York: Plenum
Press; 1989:257–260.
87. Lu ZL, Wang JZ, Williamson S. Neuronal sources of human
parieto-occipital alpha rhythm. In: Hoke M, Erne S, Okada Y, et al.,
eds. Biomagnetism. Clinical Aspects. Amsterdam: Excerpta Medica;
1992:33–37.
88. Salmelin R, Hari R. Characterization of spontaneous MEG
rhythms in healthy adults. Electroencephalogr Clin Neurophysiol.
1994;91:237–248.
89. Ilmoniemi R, Williamson S, Hostetler W. Method for the study of
spontaneous brain activity. In: Atsumi K, Kotani M, Ueno M, et al.,
eds. Biomagnetism ’87. Tokyo: Tokyo Denki University Press;
1987:182–185.
90. Kaufman L, Schwarz B, Salustri C, et al. Modulation of sponta-
neous brain activity during mental imagery. J Cogn Neurosci.
1990;2:124–132.
91. Michel CM, Kaufman L, Williamson SJ. Duration of EEG and
MEG suppression increases with angle in a mental rotation
task. J Cogn Neurosci. 1994;6:139–150.
92. Salenius S, Kajola M, Thompson WL, et al. Reactivity of magnetic
parieto-occipital alpha rhythm during visual imagery.
Electroencephalogr Clin Neurophysiol. 1995;95:453–462.
93. Portin K, Salenius S, Salmelin R, et al. Activation of the human
occipital and parietal cortex by pattern and luminance stimuli:
neuromagnetic measurements. Cereb Cortex. 1998;8:253–260.
94. Vanni S, Revonsuo A, Hari R. Modulation of the parieto-occipital
alpha rhythm during object detection. J Neurosci. 1997;17:7141–
7147.
95. Tuladhar AM, ter Huurne N, Schoffelen JM, et al. Parieto-
occipital sources account for the increase in alpha activity with
working memory load. Hum Brain Mapp. 2007;28:785–792.
96. Jeavons P. The electro-encephalogram in blind children. Br
J Opthalmol. 1964;48:83–101.
97. Salenius S, Schnitzler A, Salmelin R, et al. Modulation of human
rolandic rhythms during natural sensorimotor tasks.
Neuroimage. 1997;5:221–228.
98. Tiihonen J, Kajola M, Hari R. Magnetic mu rhythm in man.
Neuroscience. 1989;32:793–800.
99. Nagamine T, Kajola M, Salmelin R, et al. Movement-related slow
cortical magnetic fields and changes of spontaneous brain
rhythms. Electroencephalogr Clin Neurophysiol. 1996;99:274–296.
100. Schnitzler A, Salenius S, Salmelin R, et al. Involvement of primary
motor cortex in motor imagery: a neuromagnetic study.
Neuroimage. 1997;6:201–208.
101. Hari R, Forss N, Avikainen S, et al. Activation of human primary
motor cortex during action observation: a neuromagnetic study.
Proc Natl Acad Sci USA. 1998;95:15061–15065.
102. Salmelin R, Hämäläinen M, Kajola M, et al. Functional segrega-
tion of movement-related rhythmic activity in the human brain.
Neuroimage. 1995;2:237–243.
103. Jensen O, Goel P, Kopell N, et al. On the human motor-cortex beta
rhythm: Sources and modelling. Neuroimage. 2005;26:347–355.
104. Conway B, Halliday D, Farmer S, et al. Synchronization between
motor cortex and spinal motoneuronal pool during the perform-
ance of a maintained motor task in man. J Physiol. 1995;489:
917–924.
105. Salenius S, Salmelin R, Neuper C, et al. Human cortical 40-Hz
rhythm is closely related to EMG rhythmicity. Neurosci Lett.
1996;21:75–78.
106. Salenius S, Portin K, Kajola M, et al. Cortical control of human
motoneuron firing during isometric contraction. J Neurophysiol.
1997;77:3401–3405.
107. Hari R, Salenius S. Rhythmical corticomuscular communication.
Neuroreport. 1999;10:R1–R10.
108. Salenius S, Hari R. Synchronous oscillatory cortical activity
during motor action. Curr Opin Neurobiol. 2003;13:678–684.
109. Mima T, Hallett M. Corticomuscular coherence: a review. J Clin
Neurophysiol. 1999;16:501–511.
110. Kilner J, Baker S, Salenius S, et al. Human cortical muscle coher-
ence is directly related to specific motor parameters. J Neurosci.
2000;20:8838–8845.
111. Kilner JM, Salenius S, Baker SN, et al. Task-dependent modula-
tions of cortical oscillatory activity in human subjects during a
bimanual precision grip task. Neuroimage. 2003;18:67–73.
112. Brown P, Salenius S, Rothwell JC, et al. The cortical correlate of
the Piper rhythm in man. J Neurophysiol. 1998;80:2911–2917.
113. Brown P. Cortical drives to human muscle: the Piper and related
rhythms. Prog Neurobiol. 2000;60:97–108.
114. Gross J, Timmermann L, Kujala J, et al. The neural basis of inter-
mittent motor control in humans. Proc Natl Acad Sci USA.
2002;99:2299–2302.
 Chapter 42
115. Tiihonen J, Hari R, Kajola M, et al. Magnetoencephalographic
10-Hz rhythm from the human auditory cortex. Neurosci Lett.
1991;129:303–305.
116. Lehtelä L, Salmelin R, Hari R. Evidence for reactive magnetic
10-Hz rhythm in the human auditory cortex. Neurosci Lett.
1997;222:111–114.
117. Bente D. Die Insuffizienz des Vigilitätstonus. Habilitationsschrift.
University of Erlangen-Nürnberg; 1964.
118. Santamaria J, Chiappa K. The EEG of drowsiness in normal
adults. J Clin Neurophysiol. 1987;4:327–382.
119. Lu ST, Kajola M, Joutsiniemi SL, et al. Generator sites of sponta-
neous MEG activity during sleep. Electroencephalogr Clin.
Neurophysiol. 1992;82:182–196.
120. Niedermeyer E. Alpha-like rhythmical activity of the temporal
lobe. Clin Electroencephalogr. 1990;21:210–224.
121. Niedermeyer E. The ‘third rhythm’: further observations. Clin
Electroencephalogr. 1991;22:83–96.
122. Narici L, Forss N, Jousmäki V, et al. Evidence for a 7- to 9-Hz ‘sigma’
rhythm in the human SII cortex. Neuroimage. 2001;13:662–668.
123. Sasaki K, Tsujimoto T, Nambu A, et al. Dynamic activities of the
frontal association cortex in calculating and thinking. Neurosci.
Res. 1994;19:229–233.
124. Gray M, Singer W. Stimulus-specific neuronal oscillations in ori-
entation columns of cat visual cortex. Proc Natl Acad Sci USA.
1989;86:1698–1702.
125. Llinas R. Intrinsic electrical properties of mammalian neurons
and CNS function. In: Fidia Research Foundation Neuroscience
Award Lectures. New York: Raven Press; 1990;4:175–194.
126. Tallon-Baudry C, Bertrand O, Wienbruch C, et al. Combined
EEG and MEG recordings of visual 40 Hz responses to illusory
triangles in human. Prog Neurobiol. 2000;60:97–108.
127. Wyart V, Tallon-Baudry C. Neural dissociation between visual
awareness and spatial attention. J Neurosci. 2008;28:2667–2679.
128. Hughes JR, Hendrix DE, Cohen J, et al. Relationship of the mag-
netoencephalogram to the electroencephalogram. Normal wake
and sleep activity. Electroencephalogr Clin Neurophysiol. 1976;40:
261–278.
129. Nakasato N, Kado H, Nakanishi M, et al. Magnetic detection of
sleep spindles in normal subjects. Electroencephalogr Clin
Neurophysiol. 1990;76:123–130.
130. Ueno S, Iramina K. Modeling and source localization of MEG
activities. Brain Topogr. 1990;3:151–165.
131. Manshanden I, De Munck JC, Simon NR, et al. Source localiza-
tion of MEG sleep spindles and the relation to sources of alpha
band rhythms. Clin Neurophysiol. 2002;113:1937–1947.
132. Numminen J, Mäkelä J, Hari R. Distribution and sources of mag-
netoencephalographic K-complexes. Electroencephalogr Clin
Neurophysiol. 1996;99:544–555.
133. Buser P. Thalamocortical mechanisms underlying synchronized
EEG activity. In: Halliday A, Butler S, Paul R, eds. A Textbook of
Clinical Neurophysiology. Chichester: Wiley; 1987:595–621.
134. Niedermeyer E. Depth electroencephalography. In: Niedermeyer
E, Lopes da Silva F, eds. Electroencephalography. Basic Principles,
Clinical Applications and Related Fields. Baltimore-Munich:
Urban & Schwarzenberg; 1982:519–536.
135. Scheuler W, Kubicki S, Scholz G, et al. Two different activities in
the sleep spindle frequency band-discrimination based on the
topographical distribution of spectral power and coherence. In:
Horne J, ed. Sleep ’90. Bochum: Pontenagel Press; 1990:13–16.
136. Ishii R, Dziewas R, Chau W, et al. Current source density distri-
bution of sleep spindles in humans as found by synthetic aper-
ture magnetometry. Neurosci Lett. 2003;340:25–28.
■ Magnetoencephalography: Methods and Applications 895
137. Llinas R, Ribary U. Coherent 40-Hz oscillation characterizes dream
state in humans. Proc Natl Acad Sci USA. 1993;90:2078–2081.
138. Jäntti V, Baer G, Yli-Hankala A, et al. MEG burst suppression in
an anaesthesized dog. Acta Anaesth Scand. 1995;39:126–128.
139. Cohen D. Magnetoencephalography: detection of the brain’s
electrical activity with a superconducting magnetometer. Science.
1972;1175:664–666.
140. Modena I, Ricci GB, Barbanera S, et al. Biomagnetic measure-
ments of spontaneous brain activity in epileptic patients.
Electroencephalogr Clin Neurophysiol. 1982;54:622–628.
141. Rose DF, Smith PD, Sato S. Magnetoencephalography and
epilepsy research. Science. 1987;238:329–335.
142. Sutherling WW, Barth DS. Neocortical propagation in temporal
lobe spike foci on magnetoencephalography and electroen-
cephalography. Ann Neurol. 1989;25:373–381.
143. Paetau R, Kajola M, Hari R. Magnetoencephalography in the
study of epilepsy. Neurophysiol Clinique. 1990;20:169–187.
144. Paetau R, Kajola M, Karhu J, et al. Magnetoencephalographic
localization of epileptic cortex – impact on surgical treatment.
Ann Neurol. 1992;32:106–109.
145. Paetau R, Hämäläinen M, Hari R, et al. Presurgical MEG evaluation
of children with intractable epilepsy. Epilepsia. 1994;35:275–284.
146. Tiihonen J, Hari R, Kajola M, et al. Localization of epileptic foci
using a large-area magnetometer and functional brain anatomy.
Ann Neurol. 1990;27:283–290.
147. Ebersole JS. New applications of EEG/MEG in epilepsy evalua-
tion. Epilepsy Res Suppl. 1996;11:227–237.
148. Nakasato N, Seki K, Kawamura T, et al. Cortical mapping using an
MRI-linked whole head MEG system and presurgical decision
making. In: Hashimoto I, Okada Y, Ogawa S, eds. Visualization of
Information Processing in the Human Brain: Recent Advances in
MEG and Functional MRI (EEG Suppl. 47). Elsevier; 1996:333–341.
149. Watanabe Y, Fukao K, Watanabe M, et al. Epileptic events
observed by multichannel MEG. In: Hashimoto I, Okada Y,
Ogawa S, eds. Visualization of Information Processing in the
Human Brain: Recent Advances in MEG and Functional MRI (EEG
Suppl. 47). Elsevier; 1996:283–391.
150. Merlet I, Paetau R, Garcia-Larrea L, et al. Apparent asynchrony
between interictal electric and magnetic spikes. Neuroreport.
1997;8:1971–1076.
151. Lamusuo S, Forss N, Ruottinen H-M, et al. [ 18F]FDG-PET and
whole-scalp MEG localization of epileptogenic cortex. Epilepsia.
1999;40:921–930.
152. Baumgartner C, Pataraia E, Lindinger G, et al. Magneto-encephalog-
raphy in focal epilepsy. Epilepsia. 2001;41 (suppl 3): S39–S47.
153. Iwasaki M, Nakasato N, Shamoto H, et al. Surgical implications
of neuromagnetic spike localization in temporal lobe epilepsy.
Epilepsia. 2002;43:415–424.
154. Barkley GL, Baumgartner C. MEG and EEG in Epilepsy. J Clin
Neurophysiol. 2003;20:163–178.
155. Knowlton RC. Magnetoencephalography: clinical application in
epilepsy. Curr Neurol Neurosci Rep. 2003;3:341–348.
156. LinY-Y, Chang K-P, Hsieh J-C, et al. Magnetoencephalographic
analysis of bilaterally synchronous discharges in benign rolandic
epilepsy of childhood. Seizure. 2003;12:448–455.
157. Van’t Ent D, Manshanden I, Ossenblok P, et al. Spike cluster analysis
in neocortical localization related epilepsy yields clinically signifi-
cant equivalent source localization results in magnetoencephalo-
gram (MEG). Clin Neurophysiol. 2003;114:1948–1962.
158. Hari R, Ahonen A, Forss N, et al. Parietal epileptic mirror focus
detected with a whole-head neuromagnetometer. Neuroreport.
1993;5:45–48.
896 Part V ■ Complementary and Special Techniques
159. Forss N, Jousmäki V, Hari R. Interaction in the human
somatosensory cortex between afferent input from different
fingers. Brain Res. 1995;685:68–76.
160. Paetau R, Kajola M, Korkman M, et al. Landau-Kleffner syn-
drome: epileptic activity in the auditory cortex. Neuroreport.
1991;2:201–204.
161. Paetau R, Granström ML, Blomstedt G, et al. Magnetoence-
phalography in presurgical evaluation of children with the
Landau-Kleffner syndrome. Epilepsia. 1999;40:326–335.
162. Paetau R. Sounds trigger spikes in the Landau-Kleffner
syndrome. J Clin Neurophysiol. 1994;11:231–241.
163. Morrell F, Whisler WW, Smith MC, et al. Landau-Kleffner
syndrome. Treatment with subpial intracortical transection.
Brain. 1995;118:1529–1546.
164. Lau M, Yam D, Burneo JG. A systematic review on MEG and its
use in the presurgical evaluation of localization-related epilepsy.
Epilepsy Res. 2008;79:97–104.
165. Lewine JD. Commentary on Lau et al. A systematic review on
MEG and its use in the presurgical evaluation of localization
related epilepsy. Epilepsy Res. 2008;82:235–236.
166. Schwartz ES, Dlugos DJ, Storm PB, et al. Magnetoence-
phalography for pediatric epilepsy: how we do it. Am
J Neuroradiol. 2008;29:832–837.
167. Vieth J, Sack G, Schueler P, et al. Ischemic and epileptic lesions
measured by AC- and DC-MEG. In: Williamson S, Hoke M,
Stroink G, et al., eds. Advances in Biomagnetism. New York:
Plenum Press; 1989:307–310.
168. Gallen C, Schwartz B, Pantev C, et al. Detection and localization
of delta frequency activity in human stroke. In: Hoke M, Erné
SN, Okada YC, et al., eds. Biomagnetism. Clinical Aspects.
Amsterdam: Excerpta Medica; 1992;301–305.
169. Mäkelä JP. Neurological application of MEG. In: Hashimoto I,
Okada Y, Ogawa S, eds. Visualization of Information Processing in
the Human Brain: Recent Advances in MEG and Functional MRI
(EEG Suppl. 47). Elsevier; 1996;343–355.
170. Mäkelä JP, Salmelin R, Kotila M, et al. Modification of neuromag-
netic cortical signals by thalamic infarctions. Electroencephalogr
Clin Neurophysiol. 1998;106:433–443.
171. Paetau R, Santavuori P, Vanhanen S-L, et al. unpublished obser-
vations, 1991
172. Okada Y. Magnetoencephalography as a noninvasive tool for
electrophysiological characterization of auras in classic migraine.
In: Sato S, ed. Advances in Neurology, Vol 54: Magnetoence-
phalography. New York: Raven Press; 1990:133–140.
173. Barkley G, Moran J, Takanashi Y, et al. Techniques for DC
magnetoencephalography. J Clin Neurophysiol. 1991;8:189–199.
174. Hadjikhani N, Sanchez del Rio M, Wu O, et al. Mechanisms of
migraine aura revealed by functional MRI in human visual
cortex. Proc Natl Acad Sci USA. 2001;98:4687–4692.
175. Heikman P, Salmelin R, Mäkelä J, et al. Relation between the
frontal 3–7 Hz MEG activity and the efficacy of ECT in major
depression. J ECT. 2001;17:136–140.
176. Mäkelä JP, Salmelin R, Hokkanen L, et al. Neuromagnetic seque-
lae of herpes simplex encephalitis. Electroencephalogr Clin
Neurophysiol. 1998;106:251–258.
177. Mäkelä JP, Hari R, Karhu J, et al. Suppression of magnetic mu
rhythm during Parkinsonian tremor. Brain Res. 1993;617:189–193.
178. Salenius S, Avikainen S, Kaakkola S, et al. Defective cortical drive
to muscle in Parkinson’s disease and its improvement with
levodopa. Brain. 2002;125:491–500.
179. Brown P, Marsden J, Defebvre L, et al. Intermuscular coherence
in Parkinson’s disease: relationship to bradykinesia. Neuroreport.
2001;12:2577–2581.
180. Timmermann L, Gross J, Dirks M, et al. The cerebral oscillatory
network of parkinsonian resting tremor. Brain. 2003;126:199–212.
181. Schnitzler A, Timmermann L, Gross J. Physiological and patho-
logical oscillatory networks in the human motor system. J Physiol
Paris. 2005;99:3–7.
182. Elberling C, Bak C, Kofoed B, et al. Magnetic auditory responses
from the human brain. A preliminary report. Scand Audiol.
1980;9:185–190.
183. Hari R, Aittoniemi K, Järvinen ML, et al. Auditory evoked tran-
sient and sustained magnetic fields of the human brain.
Localization of neural generators. Exp Brain Res. 1980;40:237–240.
184. Kuriki S, Nogai T, Hirata Y. Cortical sources of middle latency
responses of auditory evoked magnetic field. Hering Res.
1995;92:47–51.
185. Pelizzone M, Hari R, Mäkelä JP, et al. Cortical origin of middle-
latency auditory evoked responses in man. Neurosci Lett. 1987;82:
303–307.
186. Mäkelä JP, Hämäläinen M, Hari R, et al. Whole-head mapping of
middle-latency auditory evoked fields. Electroencephalogr Clin
Neurophysiol. 1994;92:414–421.
187. McEvoy L, Mäkelä JP, Hämäläinen M, et al. Effect of interaural
time differences on middle-latency and late auditory evoked
magnetic fields. Hear Res. 1994;78:249–257.
188. Yoshiura Y, Ueno S, Iramina K, et al. Effects of stimulation site on
human middle latency auditory evoked magnetic fields. Neurosci
Lett. 1994;172:159–162.
189. Nishitani N, Nagamine T, Fujiwara N, et al. Cortical-hippocampal
auditory processing identified by magnetoencephalography.
J Cogn Neurosci. 1998;10:231–247.
190. Hari R, Kaila K, Katila T, et al. Interstimulus-interval dependence
of the auditory vertex response and its magnetic counterpart:
Implications for their neural generation. Electroencephalogr Clin
Neurophysiol. 1982;54:561–569.
191. Hari R, Pelizzone M, Mäkelä JP, et al. Neuromagnetic responses
of the human auditory cortex to on- and offsets of noise bursts.
Audiology. 1987;25:31–43.
192. Lu ZL, Williamson S, Kaufman L. Behavioral lifetime of human
auditory sensory memory predicted by physiological measures.
Science. 1992;258:1668–1670.
193. Mäkelä JP, Ahonen A, Hämäläinen M, et al. Functional differ-
ences between auditory cortices of the two hemispheres revealed
by whole-head neuromagnetic recordings. Hum Brain Mapp.
1993;1:48–56.
194. Sams M, Hari R, Rif J, et al. The human auditory sensory mem-
ory trace persists about 10 s: neuromagnetic evidence. J Cogn
Neurosci. 1993;5:363–370.
195. Pantev C, Hoke M, Lehnertz K, et al. Tonotopic organization of
the human auditory cortex revealed by transient auditory evoked
magnetic fields. Electroencephalogr Clin Neurophysiol. 1988;69:
160–170.
196. Pantev C, Hoke K, Lehnertz K, et al. Neuromagnetic evidence of
an amplitopic organization of the human auditory cortex.
Electroencephalogr Clin Neurophysiol. 1989;72:225–231.
197. Lütkenhöner B, Krumbholtz K, Seither-Preisler A. Studies of
tonotopy based on wave N100 of the auditory evoked field are
problematic. NeuroImage. 2003;9:935–949.
198. Mäkelä JP, Hari R, Linnankivi A. Different analysis of frequency
and amplitude modulations of a continuous tone in the human
auditory cortex: A neuromagnetic study. Hear Res. 1987;27:257–
264.
199. Kaukoranta E, Hari R, Lounasmaa OV. Responses of the human
auditory cortex to vowel onset after fricative consonants. Exp
Brain Res. 1987;69:19–23.
 Chapter 42
200. Mäkelä JP, Hari R, Leinonen L. Magnetic responses of the human
auditory cortex to noise/tone-transitions. Electroencephalogr Clin
Neurophysiol. 1988;69:423–430.
201. Hari R, Hämäläinen M, Kaukoranta E, et al. Selective listening
modifies activity of the human auditory cortex. Exp Brain Res.
1989;74:463–470.
202. Arthur D, Lewis P, Medwick P, et al. A neuromagnetic study of
selective auditory attention. Electroenceph. Clin Neurophysiol.
1991;78:348–360.
203. Kuriki S, Takeuchi F. Neuromagnetic responses elicited by
auditory stimuli in dichotic listening. Electroencephalogr Clin
Neurophysiol. 1991;80:406–411.
204. Rif J, Hari R, Hämäläinen M, et al. Auditory attention affects two
different areas in the human auditory cortex. Electroencephalogr
Clin Neurophysiol. 1991;79:464–472.
205. McEvoy L, Hari R, Imada T, et al. Human auditory cortical mech-
anisms of sound lateralization: II. Interaural time differences at
sound onset. Hear Res. 1993;67:98–109.
206. Sams M, Hämäläinen M, Hari R, et al. Human auditory cortical
mechanisms of sound lateralization: I. Interaural time differences
within sound. Hear Res. 1993;67:89–97.
207. Näätänen R, Picton T. The N1 wave of the human electric and
magnetic response to sound: a review and analysis of the compo-
nent structure. Psychophysiology. 1987;24:375–425.
208. Näätänen R. Attention and Brain Function. Hillsdale, New Jersey:
Erlbaum; 1992.
209. Näätänen R. Mismatch negativity: clinical research and possible
applications. Int J Psychophysiol. 2003;48:179–188.
210. Alho K. Cerebral generators of mismatch negativity (MMN) and
its magnetic counterpart (MMNm) elicited by sound changes.
Ear Hear. 1995;16:38–51.
211. Raij T, McEvoy L, Mäkelä J, et al. Human auditory cortex is activated
by omissions of auditory stimuli. Brain Res. 1997;745:134–143.
212. Sams M, Aulanko R, Hämäläinen M, et al. Seeing speech: visual
information from lip movements modifies activity in the human
auditory cortex. Neurosci Lett. 1991;127:141–145.
213. Calvert GA, Bullmore ET, Brammer MJ, et al. Activation of audi-
tory cortex during silent lipreading. Science. 1997;276:593–596.
214. Romani GL, Williamson SJ, Kaufman L. Tonotopic organization
of the human auditory cortex. Science. 1982;216:1339–1340.
215. Galambos R, Makeig S, Talmachoff PJ. A 40-Hz auditory poten-
tial recorded from the human scalp. Proc NY Acad Sci. 1981;78:
2643–2647.
216. Mäkelä J, Hari R. Evidence for cortical origin of the 40-Hz audi-
tory evoked response in man. Electroencephalogr Clin
Neurophysiol. 1987;66:539–546.
217. Ross B, Picton TW, Pantev C. Temporal integration in the human
auditory cortex as represented by the development of the steady-
state magnetic field. Hear Res. 2002;165:68–84.
218. Hari R, Hämäläinen M, Joutsiniemi SL. Neuromagnetic steady-
state responses to auditory stimuli. J Acoust Soc Am.
1989;86:1033–1039.
219. Gutschalk A, Mase R, Roth R, et al. Deconvolution of 40 Hz
steady-state fields reveals two overlapping source activities of the
human auditory cortex. Clin Neurophysiol. 1999;110:856–868.
220. Pastor MA, Artieda J, Arbizu J, et al. Activation of human cerebral
and cerebellar cortex by auditory stimulation at 40 Hz.
J Neurosci. 2002;22:10501–10506.
221. Joliot M, Ribary M, Llinas R. Human oscillatory brain activity
near 40 Hz coexists with cognitive temporal binding. Proc Natl
Acad Sci USA. 1994;91:11748–11751.
222. Fujiki N, Jousmäki V, Hari R. Neuromagnetic responses to fre-
quency-tagged sounds: a new method to follow inputs from each
■ Magnetoencephalography: Methods and Applications 897
ear to the human auditory cortex during binaural hearing.
J Neurosci. 2002;22:RC205.
223. Kaneko K, Fujiki N, Hari R. Binaural interaction in the human
auditory cortex revealed by neuromagnetic frequency-tagging:
no effect of stimulus intensity. Hearing Res. 2003;183:1–6.
224. Leinonen L, Joutsiniemi SL. Auditory evoked potentials and
magnetic fields in patients with lesions of the auditory cortex.
Acta Neurol Scand. 224;79:316–325.
225. Mäkelä JP, Hari R, Valanne L, et al. Auditory evoked magnetic
fields after ischemic brain lesions. Ann Neurol. 1991;30:76–82.
226. Mäkelä JP, Hari R. Neuromagnetic auditory evoked responses after
a stroke in the right temporal lobe. Neuroreport. 1992;3:94–96.
227. Pelizzone M, Hari R, Mäkelä J, et al. Cortical activity evoked by a
multichannel cochlear prosthesis. Acta Oto-laryngol. 1987;103:
632–636.
228. Hari R, Pelizzone M, Mäkelä J, et al. Neuromagnetic responses
from a deaf subject to stimuli presented through a multichannel
cochlear prosthesis. Ear Hear. 1988;9:148–152.
229. Hoke M, Pantev C, Lütkenhöner B, et al. Magnetic fields from the
auditory cortex of a deaf human individual occurring sponta-
neously or evoked by stimulation through a cochlear prosthesis.
Audiology. 1989;28:152–170.
230. Hari R. Neuromagnetic approach to human auditory cortical
functions, with emphasis on subjects with cochlear implants. In:
Honjo I, Takahashi H, eds. Cochlear Implant and Related Sciences
Update Adv. Otorhinolaryngol. Vol. 52. Basel: Karger; 1997:15–18.
231. Vasama JP, Mäkelä JP, Parkkonen L, et al. Auditory cortical
responses in humans with congential unilateral conductive hear-
ing loss. Hear Res. 1994;78:91–97.
232. Vasama JP, Mäkelä JP, Pyykkö I, et al. Modification of central
auditory pathways after unilateral permanent deafness due to
removal of acoustic neuroma. Neuroreport. 1994;6:961–964.
233. Vasama JP, Mäkelä JP. Auditory pathway plasticity in adult
humans after idiopathic sudden unilateral sensorineural hearing
loss. Hear Res. 1995;87:132–140.
234. Pekkonen E, Huotilainen M, Virtanen J, et al. Alzheimer’s disease
affects parallel processing between the auditory cortices.
Neuroreport. 1996;7:1365–1368.
235. Tiihonen J, Hari R, Naukkarinen H, et al. Auditory hallucinations
may modify activity of the human auditory cortex. Am
J Psychiatry. 1992;149:255–257.
236. Hoke M, Feldman H, Pantev C, et al. Objective evidence of tinni-
tus in auditory evoke magnetic fields. Hear Res. 1989;37:281–286.
237. Pantev C, Hoke M, Lütkenhöner B, et al. Tinnitus remission
objectified by neuromagnetic measurements. Hear Res. 1989;40:
261–264.
238. Jacobson G, Ahmad B, Moran J, et al. Auditory evoked cortical
magnetic field (M100/M200) measurements in tinnitus and nor-
mal groups. Hear Res. 1991;56:44–52.
239. Colding-Jørgensen E, Lauritzen M, Johnsen NJ, et al. On the evi-
dence of auditory evoked magnetic fields as an objective measure
of tinnitus. Electroencephalogr Clin Neurophysiol. 1992;83:
322–327.
240. Shiomi Y, Nagamine T, Fujiki N, et al. Tinnitus remission by lido-
caine demonstrated by auditory-evoked magnetoencephalo-
gram. Acta Otolaryngol (Stockh.) 1997;117:31–34.
241. Muhlnickel W, Elbert T, Taub E, et al. Reorganization of auditory
cortex in tinnitus. Proc Natl Acad Sci USA. 1998;95:10340–10343.
242. Nagarajan S, Mahncke H, Salz T, et al. Cortical auditory signal
processing in poor readers. Proc Natl Acad Sci USA. 1999;96:
6483–6488.
243. Renvall H, Hari R. Diminished auditory mismatch fields in
dyslexic adults. Ann Neurol. 2003;53:551–557.
898 Part V ■ Complementary and Special Techniques
244. Gootjes L, Raij T, Salmelin R, et al. Left-hemisphere dominance
for processing of vowels: a whole-scalp neuromagnetic study.
Neuroreport. 1999;10:2987–2991.
245. Brenner D, Lipton J, Kaufman L, et al. Somatically evoked mag-
netic fields of the human brain. Science. 1978;199:81–83.
246. Suk J, Ribary U, Cappell J, et al. Anatomical localization revealed
by MEG recordings of the human somatosensory system.
Electroencephalogr Clin Neurophysiol. 1991;78:185–196.
247. Yang TT, Gallen CC, Schwartz BJ, et al. Noninvasive
somatosensory homunculus mapping in humans by using a
large-array biomagnetometer. Proc Natl Acad Sci USA. 1993;90:
3098–3102.
248. Huttunen J, Kaukoranta E, Hari R. Cerebral magnetic responses
to stimulation of tibial and sural nerves. J Neurol Sci. 1987;79:
43–54.
249. Fujita S, Nakasato N, Matani A, et al. Short latency somatosen-
sory evoked field for tibial nerve stimulation: rotation of dipole
pattern over the whole head. In: Baumgartner C, Deecke L,
Stroink G, et al., eds. Biomagnetism: Fundamental Research and
Clinical Applications. Amsterdam: Elsevier; 1995:95–98.
250. Kakigi R, Koyama S, Hoshiyama M, et al. Topography of
somatosensory evoked magnetic fields following posterior tibial
nerve stimulation. Electroencephalogr Clin. Neurophysiol 1995;95:
127–134.
251. Hari R, Nagamine T, Nishitani N, et al. Time-varying activation
of different cytoarchitectonic areas of the human SI cortex after
tibial nerve stimulation. Neuroimage. 1996;4:111–118.
252. Hashimoto I, Sakuma K, Kimura T, et al. Serial activation of dis-
tinct cytoarchitectonic areas of the human SI cortex after poste-
rior tibial nerve stimulation. Neuroreport. 2001;12:1857–1862.
253. Schnitzler A, Salmelin R, Salenius S, et al. Tactile information
from the human hand reaches the ipsilateral primary somatosen-
sory cortex. Neurosci Lett. 1995;200:25–28.
254. Korvenoja A, Wikström H, Huttunen J, et al. Activation of ipsilat-
eral primary sensorimotor cortex by median nerve stimulation.
Neuroreport. 1995;6:2589–2593.
255. Hari R, Imada T. Ipsilateral movement-evoked fields (MEFs)
reconsidered. Neuroimage. 1999;10:582–588.
256. Lauronen L, Nevalainen P, Wikström H, et al. Immaturity of
somatosensory cortical processing in human newborns.
Neuroimage. 2006;33:195–203.
257. Curio G, Mackert B, Burghoff M, et al. Localization of evoked
neuromagnetic 600 Hz activity in the cerebral somato -
sensory system. Electroencephalogr Clin Neurophysiol. 1994;91:
483–487.
258.Hashimoto I, Mashiko T, Imada T. Somatic evoked high-frequency
magnetic oscillations reflect activity of inhibitory interneurons in
the human somatosensory cortex. Electroencephalogr Clin
Neurophysiol. 1996;100:189–203.
259. Elbert T, Pantev C, Wienbruch C, et al. Increased cortical repre-
sentation of the fingers of the left hand in string players. Science.
1995;270:305–307.
260. Pantev C, Engelien A, Candia V, et al. Representational cortex in
musicians. Plastic alterations in response to musical practice. Ann
NY Acad Sci. 2001;930:300–314.
261. Hari R, Hämäläinen M, Kaukoranta E, et al. Neuromagnetic
responses from the second somatosensory cortex in man. Acta
Neurol Scand. 1983;68:207–212.
262. Hari R, Hämäläinen H, Tiihonen J, et al. Separate finger repre-
sentations at the human second somatosensory cortex.
Neuroscience. 1990;37:245–249.
263. Forss N, Salmelin R, Hari R. Comparison of somatosensory
evoked fields to airpuff and electric stimuli. Electroencephalogr
Clin Neurophysiol. 1994;92:510–517.
264. Del Gratta C, Della Penna S, Ferretti A, et al. Topographic organ-
ization of the human primary and secondary somatosensory cor-
tices: comparison of fMRI and MEG findings. NeuroImage.
2002;17:1373–1383.
265. Tuunanen PI, Kavec M, Jousmäki V, et al. Comparison of BOLD
fMRI and MEG characteristics to vibrotactile stimulation.
Neuroimage. 2003;19:1778–1786.
266. Simoes C, Mertens M, Forss N, et al. Functional overlap of finger
representations in human SI and SII cortices. J Neurophysiol.
2001;86:1661–1665.
267. Simões C, Alary F, Forss N, et al. Left-hemisphere-dominant SII
activation after bilateral median nerve stimulation. Neuroimage.
2002;15:686–690.
268. Huttunen J, Wikström H, Korvenoja A, et al. Significance of the
second somatosensory cortex in sensorimotor integration:
enhancement of sensory responses during finger movements.
Neuroreport. 1996;7:1009–1012.
269. Forss N, Hietanen M, Hari R. Modified activation of somatosen-
sory cortical network in patients with right hemisphere stroke.
Brain. 1999;122:1989–1899.
270. Simões C, Jensen O, Parkkonen L, et al. Phase locking between
human primary and secondary somatosensory cortices. Proc Natl
Acad Sci USA. 2003;100:2691–2694.
271. Karhu J, Hari R, Mäkelä J, et al. Somatosensory evoked magnetic
fields in multiple sclerosis. Electroenceph Clin Neurophysiol.
1992;83:192–200.
272. Karhu J, Hari R, Paetau R, et al. Cortical reactivity in progressive
myoclonus epilepsy. Electroenceph Clin Neurophysiol.
1994;90:93–102.
273. Lauronen L, Heikkilä E, Autti T, et al. Somatosensory evoked
magnetic fields from primary sensorimotor cortex in juvenile
neuronal ceroid lipofuscinosis. J Child Neurol. 1997;12:
355–360.
274. Lauronen L, Huttunen J, Kirveskari E, et al. Enlarged SI and SII
somatosensory evoked responses in the CLN5 form of neuronal
ceroid lipofuscinosis. Clin Neurophysiol. 2002;113:1491–1500.
275. Forss N, Silén T, Karjalainen T. Lack of activation of human sec-
ondary somatosensory cortex in Unverricht-Lundborg type of
progressive myoclonus epilepsy. Ann Neurol. 2001;49:90–97.
276. Hari R, Kaukoranta E, Reinikainen K, et al. Neuromagnetic local-
ization of cortical activity evoked by painful dental stimulation in
man. Neurosci Lett. 1983;42:77–82.
277. Huttunen J, Kobal G, Kaukoranta E, et al. Cortical responses to
painful CO2 stimulation of the nasal mucosa. Electroencephalogr
Clin Neurophysiol. 1986;64:347–349.
278. Joseph J, Howland E, Wakai R, et al. Late pain-related magnetic
fields and electric potentials evoked by intracutaneous electric
finger stimulation. Electroencephalogr Clin Neurophysiol.
1991;80:46–52.
279. Kakigi R, Koyama S, Hoshiyama M, et al. Pain-related magnetic
fields following painful CO2 laser stimulation in man. Neurosci
Lett. 1995;192:45–48.
280. Kakigi R, Tran TD, Qiu Y, et al. Cerebral responses following
stimulation of unmyelinated C-fibers in humans: electro- and
magnetoencephalographic study. Neurosci Res. 2003;45:255–275.
281. Ploner M, Schmitz F, Freund HJ, et al. Parallel activation of
primary and secondary somatosensory cortices in human pain
processing. J Neurophysiol. 1999;81:3100–3104.
 Chapter 42
282. Kanda M, Nagamine T, Ikeda A, et al. Primary somatosensory
cortex is actively involved in pain processing in human. Brain
Res. 2000;853:282–289.
283. Forss N, Raij T, Seppä M, et al. Common cortical network for first
and second pain. Neuroimage. 2005;24:132–142.
284. Hari R, Portin K, Kettenmann B, et al. Right-hemisphere prepon-
derance of responses to painful CO2 stimulation of the human
nasal mucosa. Pain. 1997;72:145–151.
285. Bragard D, Chen AC, Plaghki L. Direct isolation of ultra-late (C-
fibre) evoked brain potentials by CO2 laser stimulation of tiny
cutaneous surface areas in man. Neurosci Lett. 1996;209:81–84.
286. Hari R, Forss N, Raij TT. Neuromagnetic exploration of the
connection between pain and the motor cortex. In: Kalso E,
Estlander A-M, Klockars M, eds. Psyche, Soma and Pain. Acta
Gyllenbergiana IV. Helsinki: The Signe and Ane Gyllenberg
Foundation; 2003:145–153.
287. Raij TT, Forss N, Stancak A, et al. Noxious input activates the
human motor cortex: implications for tension-type pain? Soc
Neurosci Ann Meet. 2003;238:3.
288. Raij TT, Forss N, Stancak A, et al. Modulation of motor-cortex
oscillatory activity by painful A - and C-fiber stimuli.
Neuroimage. 2004;23:569–573.
289. Juottonen K, Gockel M, Silén T, et al. Altered central sensorimo-
tor processing in patients with complex regional pain syndrome.
Pain. 2002;98:315–323.
290. Forss N, Kirveskari E, Gockel M. Mirror-like spread of chronic
pain. Neurology. 2005;65:748–750.
291. Brenner D, Williamson SJ, Kaufman L. Visually evoked magnetic
fields of the human brain. Science. 1975;190:480–481.
292. Teyler TJ, Cuffin BN, Cohen D. The visual magnetoencephalo-
gram. Life Sci. 1975;17:683–692.
293. Maclin E, Okada Y, Kaufman L, et al. Retinotopic map on the
visual cortex for eccentrically placed patterns: first noninvasive
measurement. Nuovo Cimento C. 1983;2D:410–419.
294. Ahlfors S, Ilmoniemi R, Hämäläinen M. Estimates of visually
evoked cortical currents. Electroencephalogr Clin Neurophysiol.
1992;82:225–236.
295. Aine C, Supek S, George J, et al. Retinotopic organization of
human visual cortex: Departures from the classical model. Cereb
Cortex. 1996;6:354–361.
296. Tabuchi H, Yokoyama T, Shimogawara M, et al. Study of the
visual evoked magnetic field with the m-sequence technique.
Invest Ophthalmol Vis Sci. 2002;43:2045–2054.
297. Hari R, Salmelin R, Tissari S, et al. Visual stability during eye-
blinks. Nature. 1994;367:121–122.
298. Bristow D, Frith C, Rees G. Two distinct neural effects of blinking
on human visual processing. Neuroimage. 2005;27:136–145.
299. Jousmäki V, Hämäläinen M Hari R. Magnetic source imaging
during a visually guided task. Neuroreport. 1996;7:2961–2964.
300. Galletti C, Battaglini P, Fattori P. Eye position influence on the
parieto-occipital area PO (V6) of the macaque monkey. Eur
J Neurosci. 1995;7:2486–2501.
301. Ffytche DH, Guy CN, Zeki S. The parallel visual motion inputs
into areas V1 and V5 of human cerebral cortex. Brain.
1995;118:1375–1394.
302. Anderson SJ, Holliday IE, Singh KD, Harding GFA. Localization
and functional analysis of human cortical area V5 using magneto
encephalography. Proc R Soc Lond B Biol Sci. 1996;263: 423–431.
303. Uusitalo M, JousmäkiV, Hari R. Activation trace lifetime of
human cortical responses evoked by apparent visual motion.
Neurosci Lett. 1997;224:45–48.
■ Magnetoencephalography: Methods and Applications 899
304. Uusitalo M, Virsu V, Salenius S, et al. Human cortical activation
related to perception of visual motion and movement after-
effect. Neuroimage. 1997;5:241–250.
305. Vanni S, Tanskanen T, Seppä M, et al. Coinciding early activation
of human primary visual cortex and anteromedial cuneus. Proc
Natl Acad Sci USA. 2001;98:2776–2780.
306. Schmolesky MT, Wang Y, Hanes DP, et al. Signal timing across the
macaque visual system. J Neurophysiol. 1998;79:3272–3278.
307. Petroni F, Panzeri S, Hilgetag CC, et al. Simultaneity of responses
in a hierarchical visual network. Neuroreport. 2001;12:2753–2759.
308. Lu S, Hämäläinen M, Hari R, et al. Seeing faces activates three
brain areas outside the occipital visual cortex in man.
Neuroscience. 1991;43:287–290.
309. Sams M, Hietanen J, Hari R, et al. Face-specific responses from
the human inferior occipitotemporal cortex. Neuroscience.
1997;77:49–55.
310. Linkenkaer-Hansen K, Palva JM, Sams M, et al. Face-selective
processing in human extrastriate cortex around 120 ms after
stimulus onset revealed by magneto- and electroencephalogra-
phy. Neurosci Lett. 1998;253:147–50.
311. Halgren E, Raij T, Marinkovic K, et al. Cognitive response profile
of the human fusiform area as determined by MEG. Cereb
Cortex. 2000;10:69–81.
312. Liu J, Harris A, Kanwisher N. Stages of processing in face percep-
tion: an MEG study. Nat Neurosci. 2002;5:910–916.
313. Salmelin R, Hari R, Lounasmaa OV, Sams M. Dynamics of brain
activation during picture naming. Nature. 1994;368:463–465.
314. Maess B, Koelsch S, Gunter TC, et al. Musical syntax is processed
in Broca’s area: an MEG study. Nat Neurosci. 2001;4:540–545.
315. Cornelissen P, Tarkiainen A, Helenius P, et al. Cortical effects of
shifting letter position in letter strings of varying length. J Cogn
Neurosci. 2003;15:731–746.
316. Cornelissen K, Laine M, Tarkiainen A, et al. Adult brain plasticity
elicited by anomia treatment. J Cogn Neurosci. 2003;15:444–461.
317.Salmelin R, Schnitzler A, Schmitz F, et al. Single word reading in devel-
opmental stutterers and fluent speakers. Brain. 2000;123:1184–202.
318. Salmelin R, Service E, Kiesilä P, et al. Impaired visual word pro-
cessing in dyslexia revealed with magnetoencephalography. Ann
Neurol. 1996;40:157–162.
319. Salmelin R, Helenius P, Service E. Neurophysiology of fluent and
impaired reading: a magnetoencephalographic approach. J Clin
Neurophysiol. 2000;17:163–174.
320. Helenius P, Salmelin R, Service E, et al. Semantic cortical activa-
tion in dyslexic readers. J Cogn Neurosci. 1999;11:535–550.
321.Salmelin R, Kujala J. Neural representation of language: activation
versus long-range connectivity. Trends Cogn Sci. 2006;10:519–525.
322. Imada T, Zhang Y, Cheour M, et al. Infant speech perception acti-
vates Broca’s area: a developmental magnetoencephalography
study. Neuroreport. 2006;17:957–962.
323. Tesche CD, Karhu J. Somatosensory evoked magnetic fields aris-
ing from sources in the human cerebellum. Brain Res.
1997;744:23–31.
324. Tesche CD, Karhu JY. Anticipatory cerebellar response during
somatosensory omission in man. Hum Brain Mapp. 2000;9:
119–142.
325. Parkkonen L, Andersson J, Hämäläinen M, et al. Early visual areas
reflect the percept of an ambiguous scene. Proc Nat Acad Sci USA.
2008;105:20500–20504.
326. Rizzolatti G, Fadiga L, Gallese V, et al. Premotor cortex and the
recognition of motor actions. Brain Res Cogn Brain Res.
1996;3:131–141.
900 Part V ■ Complementary and Special Techniques
327. Rizzolatti G, Fogassi L, Gallese V. Neurophysiological mecha-
nisms underlying the understanding and imitation of action. Nat
Rev Neurosci. 2001;2:661–670.
328. Nishitani N, Hari R. Temporal dynamics of cortical representa-
tion for action. Proc Natl Acad Sci USA. 2000;97:913–918.
329. Nishitani N, Hari R. Viewing lip forms: cortical dynamics.
Neuron. 2002;36:1211–1220.
330. Caetano G, Jousmäki V, Hari R. Actor’s and viewer’s primary
motor cortices stabilize similarly after seen or heard motor
actions. Proc Natl Acad Sci USA. 2007;104:9058–9062.
331. Hari R, Nishitani N. From viewing of movement to imitation and
understanding of other persons’ acts: MEG studies of the human
mirror-neuron system. In: Kanwisher N, Duncan J, eds.
Functional Neuroimaging of Visual Cognition. Attention and
Performance XX. Oxford: Oxford University Press; 2004:463–479.
332. Hari R, Kujala MV. Brain basis of human social interaction. From
concepts to brain imaging. Physiol Reviews. 2009;89:453–479.
333. Hari R, Parkkonen L, Nangini C. The brain in time: Insights from
neuromagnetic recordings. Ann NY Acad Sci. 2010;1191:89–109.
 Part VI EEG: Neuropharmacology and Anesthesia
CHAPTER
EEG, Drug Effects, and
Central Nervous System Poisoning
GERHARD BAUER AND RICHARD BAUER
43
A
large number of patients examined in an electroen- observed. At the deepest level, EEG activity ceases. The burst-
cephalographic (EEG) laboratory are under treatment suppression pattern exhibits no differences to those seen with
with drugs that alter EEG activity. Therefore, it is deep comatose states under pathologic circumstances (see
important to know changes due to the particular drug(s) used, Chapter 23) or intoxication with sedative drugs (see below).
chronic overdosage, the patterns of overt intoxication, and
withdrawal effects. The abundant literature of the topic makes
a selection inevitable. This chapter discusses EEG findings of HYPNOTICS AND SEDATIVES
drugs most frequently used in modern medicine, provides an
Barbiturates
overview of toxic encephalopathies of clinical importance, and
concentrates on visual EEG reading. Pharmacogenomics of Barbiturates remained on the market only as anesthetics and
drug actions and toxicities are not addressed in detail (for sum- antiepileptic drugs, especially in developing countries (8). They
mary see Ref. 1). apparently act at all levels of the neuraxis by binding to a specific
Superimposed fast frequencies (Fig. 43.1) are suggestive for site on the -aminobutyric acid (GABAA) receptor, different
sedating drugs. Otherwise, EEG abnormalities due to substance from the binding site of benzodiazepines (9). Low blood concen-
intake are unspecific. Diffuse slow activities and coma patterns trations produce inhibition of higher cortical functions and dis-
in cases of severe intoxication cannot be distinguished from inhibition of more primitive behavior. With increasing blood
those seen with other etiologies. This is also true for epilepti- levels, clinical signs are determined by generalized inhibition.
form activities and triphasic waves regardless whether they are
encountered in intoxicated patients with or without acute EEG Changes with Therapeutic Doses
symptomatic seizures. With localized EEG abnormalities an When administered in small doses, the barbiturates produce an
additional focal lesion should be excluded (2). increase in fast activities in the range of 25 to 35 cycles/sec (Hz),
soon shifting to 15 to 25 Hz. This activity is predominant over
DRUGS ACTING ON THE
CENTRAL NERVOUS SYSTEM
General Anesthetics
Molecular and cellular mechanisms of general anesthesia have
been reviewed by Franks and Lieb (3) and Campagna et al. (4).
Anesthetics exert their effects primarily at the neuronal ion chan-
nels via neurotransmitter receptors and by binding directly to the
protein site. Initially, anesthetics induce amnesia, euphoria, anal-
gesia, excitation, and hyperreflexia. Surgical anesthesia consists of
deep sedation, muscle relaxation, and diminished or abolished
motor and autonomous responses to noxious stimuli. Specific
effects of different substances (nitrous oxide, halothane, enflu-
rane, isoflurane, sevoflurane, desflurane, propofol, etc.) are due to
specific binding sites and genetic peculiarities of the patient.
Correlation studies of EEG and the stage of anesthesia have
been summarized by Winters (5) and Sloan (6). The initial
phase is dominated by the appearance of frontal fast activity,
gradually becoming more generalized and associated with dis-
solution of the alpha rhythm. During excitation, epileptiform
activities become discernible in nonepileptic and epileptic
patients with different anesthetic drugs (for summary see Ref. Figure 43.1 St.St., a 14-year-old female teenager, who took an
7). As anesthesia deepens, activity becomes slower and the unknown amount of diazepam. Clinically slight drowsiness. Twelve- to
voltage increases. Eventually, a burst-suppression pattern can be 13-Hz rhythmical activities, maximal over the anterior regions.

901
902 Part VI ■ EEG: Neuropharmacology and Anesthesia
Figure 43.2 W.M., a 33-year-old female and chronic abuser of alcohol and barbiturates (Optalidon). EEG at the sixth day
after admission for withdrawal. No seizures. Paroxysmal 2- to 3-Hz waves with intermingled small spikes.
the frontal cortex and spreads to parietal and occipital areas.
Increasing doses are accompanied by intermixed slow activities
and dissolution of the alpha rhythm, indicating drowsiness and
sleep. In children, the elderly, and patients with organic brain
lesions, barbiturates even in low doses may induce irritability,
hyperactivity, or delirium (10). Lack of barbiturate-induced fast
activity indicates a diffuse organic brain lesion; asymmetries are
suggestive of a focal lesion (11).
Chronic Overdosage
Signs of chronic barbiturate intoxication in seizure patients
might be misinterpreted as personality changes attributed to
the epileptic disorder. Signs include general sluggishness,
lethargy, difficulty in thinking, poor memory, learning disabili-
ties, and cerebellar deficits (10). Chronic neurotoxicity may
result in insidiously developing disturbances of higher cognitive
functions without concomitant abnormalities (12). Compared
with other antiepileptic drugs, cognitive adverse effects are pro-
nounced with phenobarbital (for review see Ref. 13). EEG
changes are similar to those seen in the first stage of acute intox-
ication. In epileptic patients, an increase in slow activities in
serial recordings without increase or even with reduction in
seizure frequency is indicative of overdosage and should
prompt an evaluation of the serum level.
EEG Changes after Withdrawal
After long-term ingestion of barbiturates, abrupt withdrawal
leads to paroxysmal abnormalities. Myoclonic jerks, even as
myoclonic status, generalized tonic–clonic seizures, and delir-
ium are major complications. The EEG frequently shows gener-
alized paroxysmal activities and spikes, especially with photic
stimulation (14). These changes are usually transient in nature
and most often occur within the first few days after withdrawal,
but they may occasionally persist for 3 to 4 weeks. They also
occur without clinical seizures and even with medically con-
trolled gradual withdrawal (Fig. 43.2).
Acute Intoxication
Sedatives such as barbiturates, anxiolytic drugs, phenoth-
iazines, and tricyclic antidepressants can all produce coma.
They directly depress cellular oxidative metabolism and do not
permanently damage neuronal functions. The mere functional
character of CNS depression after an acute overdose indicates a
less grave prognosis even in the presence of very serious clinical
and EEG signs.
According to the multilevel action of barbiturates, the clin-
ical syndrome encompasses cortical, reticular, vestibular, and
other brainstem dysfunctions. Coma due to overdose with
sedatives is fairly characteristic if one considers the depth of
 Chapter 43 ■ EEG, Drug Effects, and Central Nervous System Poisoning 903

Figure 43.3 E.A., a 24-year-old female.

Suicidal attempt by ingestion of a combina-
tion of cyclobarbital, hexobarbital, meproba-
mate, and carbromal (= Somnupan, banned
from the Austrian market for several years).
Coma with decerebrate posturing (with fast-
acting barbiturates cerebral functions can be
depressed in a rostro-caudal fashion and
motor signs may initially evolve) (16). EEG
with diffuse slow activity superimposed by
high-voltage 10- to 13-Hz activities. Patient
recovered completely.

unresponsiveness in combination with flaccid muscle tone, Miscellaneous Hypnotic Drugs

absent plantar responses, and preserved pupillary reactions Melatonin
(15). With fast-acting barbiturates cerebral functions can be
Melatonin has been extensively discussed in the popular media.
depressed in a rostro-caudal fashion and flexor and extensor
It is the hormone of darkness (22). The substance may act as a
postures may initially evolve (Fig. 43.3) (16). In severe cases,
phase-setter for sleep–wake cycles in subjects with a delayed
circulatory effects lead to a typical shock syndrome.
sleep phase syndrome (23). However, no evidence was found
Subsequent cerebral hypoxia may turn the functional distur-
that melatonin is effective in treating secondary sleep disorders
bance into structural damage with serious prognostic implica-
or jet lag and shift work disorders (24). For difficulties in sleep
tions.
EEG signs in different stages of acute intoxication closely
resemble those observed with other CNS depressant substances.
Initial dissolution of the alpha rhythm and appearance of inter-
spersed theta frequencies are followed by predominant slow
activities superimposed by 10- to 16-Hz rhythmical activity
maximal over the anterior regions. Such an EEG is suggestive of
intoxication with depressant drugs of any type. However, a so-
called spindle coma after head injury and the alpha coma pat-
tern with cerebral hypoxia may look similar. Superimposed fast
activities after intoxications have also been termed “drug-
induced alpha coma” (17). The separation from alpha coma
pattern with hypoxic states has prognostic significance with a
much more favorable outlook with intoxications (17). With
deepening coma, fast frequencies disappear and diffuse delta
activity becomes prominent. With impending breakdown of
vegetative functions, periods of flattening and, eventually, a
burst-suppression pattern appears (18). Preceding or following
the development of a burst-suppression pattern, the record may
be characterized by bi- or triphasic sharp transients (Fig. 43.4)
(19). Electrocerebral silence signifies the most advanced cases.
The prognostic meaning of this otherwise ominous sign is less
grave with intoxication. Several patients survived without per-
manent neurologic sequels (20).
Attempts have been made to correlate the EEG with clinical
signs and blood levels (14,19,21). In the earliest phases, clinical
signs are superior to the EEG in the assessment of the severity
of the functional disturbance, whereas the EEG permits grading Figure 43.4 B.E., a 33-year-old female. Coma after suicidal attempt by
of stages during the later phases. ingestion of barbiturates. Diffuse slow activity with triphasic waves.
904 Part VI ■ EEG: Neuropharmacology and Anesthesia
onset, ramelteon, a melatonin receptor agonist was recom-
mended (25). Effects on the conventional EEG have not been
reported for melatonin or the agonist.
Bromides
Bromide was the drug of choice as anticonvulsant and sedative
during the second half of the 19th century. It is still used in the
treatment of therapy-resistant tonic–clonic seizures (26). With
acute bromide intoxication, the EEG shows mixed slow and fast
activities. Very pronounced EEG slowing is found in chronic
bromide encephalopathies. Drug level determinations have
improved the handling of the drug, so the typical bromism of
the old days has disappeared.
Antipsychotic Drugs
Antipsychotic drugs are used for the treatment of schizophrenia.
They are listed as typical, classic or first-generation antipsychotic
agents (chlorpromazine, chlorprothixen, flupentixol, melperone,
fluphenazine, perphenazine, trifluoperazine, thiothixene,
haloperidol among 18 drugs developed between the 1950s and the
1970s), and atypical or second-generation drugs (clozapine,
risperidone, olanzapine, quetiapine, ziprasidone, aripiprazole,
amisulpride). Major progress has been made to elucidate their
mechanism of action (for review see Ref. 27). Typical antipsy-
chotic drugs act by blocking the dopamine D2 receptor and are
connected with parkinsonian side effects. The newer antipsychotic
compounds bind at many different receptors including members
of the dopamine receptor family and 5-HT2A receptors.
Typical Antipsychotic Drugs
EEG changes due to neuroleptics are not particularly different
for the numerous substances under this pharmacologic heading
and therefore are discussed jointly.
EEG Changes with Therapeutic Doses
A great number of quantitative analyses of EEG changes due to
therapeutic doses of neuroleptics have been summarized by
Saletu (28). Neuroleptics increase the alpha activity with a slight
shift to the lower frequency range and increase the amount of
slow activity and the general voltage output, but decrease per-
cent time of beta activity, variability of frequencies, and the
average frequency. Mild to moderate diffuse abnormalities have
been observed on visual EEG reading with considerable but
inconsistent variations among the different substances (29).
The EEG abnormalities in treated patients exceeded the occur-
rence in psychiatric patients not treated with antipsychotics for
at least 5 days.
Epileptogenic Potency
Although the risk of seizures under treatment with typical
antipsychotic drugs does not differ substantially from the gen-
eral population (30), with high doses seizures and spikes in
the EEG may occur (31). EEG and clinical seizures can deteri-
orate in chronic epileptic patients. Nonconvulsive status
epilepticus has been reported under treatment with typical
neuroleptics (32).
Acute Intoxication
Overdosage with neuroleptics usually leads to sedation and
coma. In milder cases, patients can be agitated, delirious, or
confused. Involuntary extrapyramidal movements, parkinson-
ism, and generalized tonic–clonic seizures can be prominent.
Fatalities are rare if neuroleptics alone are taken; death is attrib-
utable to cardiac effects. The EEG is dominated by slow waves,
frequently occurring as generalized paroxysmal activities. Fast
frequencies as in barbiturate poisoning are not recorded (14).
Atypical Antipsychotic Drugs
A long list of newer antipsychotic drugs is on the market (cloza-
pine, amisulpride, olanzapine, quetiapine, risperidone, sertin-
dole, ziprasidone, zotepine, aripiprazole). In a systematic
meta-analysis only clozapine exhibited fewer parkinsonian
symptoms and a higher antipsychotic effect than conventional
drugs (33). Compared with the classic antipsychotics, the newer
drugs are less extensively studied with quantitative EEG methods.
Clozapine
Clozapine has affinity for serotonin (5-HT2a, 5-HT6, 5-HT7,
5-HT2c, and 5-HT3), alpha-adrenergic, and dopamine D4
receptors but weak affinity for the D2 receptor (34). It causes no
extrapyramidal symptoms and tardive dyskinesia (33). The
major adverse effect of clozapine relates to its potential for
damaging the granulocyte cell line. A cumulative risk figure for
agranulocytosis of 0.8% at 1 year and 0.9% at 18 months was
calculated (35).
EEG power spectra show an increase in delta, theta, and above-
21-Hz beta activities. The EEG became abnormal in up to 53% of
patients as a function of clozapine serum levels (36). Increased
theta and delta activities were prominent over the frontal, central,
and parietal areas (37). EEG abnormalities included paroxysmal
slowing and spikes. Clozapine can provoke seizures. Seizure fre-
quency amounts up to 4% (38), increases with doses over
600 mg/day, and seems to be greater than with other typical and
atypical antipsychotic drugs (39). Besides generalized tonic–clonic
seizures, myoclonic jerks have been observed (40).
Risperidone
Risperidone is a benzisoxazole derivative with combined
dopamine D2 receptor and serotonin 5-HT2 receptor blocking
properties (41). The substance has an established efficacy in
acute psychotic states. Risperidone does not induce EEG changes
in the waking state (42). The risk of risperidone-related seizures
amounts to 0.3%, a low number compared with the 3.5% risk of
clozapine, that ranks highest among all antipsychotic drugs (39).
Other Atypical Antipsychotic Drugs
Little is known about EEG changes of other new antipsychotic
drugs. After Alper et al. (39) the seizure risk extracted from
approval reports is 0.9% for olanzapine, 0.8% for quetiapine,
0.4% to 0.5% for ziprasidone, and 0.4% for aripiprazole.
Antidepressant Drugs
Since the introduction of the first-generation tricyclic anti-
depressants, several new drugs have been marketed. New
 Chapter 43 ■ EEG, Drug Effects, and Central Nervous System Poisoning
antidepressants might be classified according to their central
mode of action into selective serotonin reuptake inhibitors (flu-
oxetine, fluvoxamine, paroxetine, citalopram, sertraline), dual
serotoninergic antidepressants (reuptake inhibitors plus recep-
tor antagonism; nefazodone), selective serotonin and noradren-
alin reuptake inhibitors (venlafaxine), noradrenergic and
specific serotonergic antidepressants (mirtazapine), selective
noradrenalin reuptake inhibitors (reboxetine), and reversible
specific monoamine oxidase inhibitors (moclobemide) (43).
Antidepressants, especially amitriptyline, and several
antiepileptic drugs (see below) are also important in the treat-
ment of neuropathic pain (44).
Tricyclic Antidepressants
EEG Changes with Therapeutic Doses
Tricyclic antidepressants such as imipramine, amitriptyline,
doxepin, desipramine, nortriptyline, and protriptyline increase
the amount of slow and fast activities along with instability of
frequencies and voltage. Furthermore, they slow down the fre-
quency of the alpha rhythm (for review see Ref. 28).
Epileptogenic Potency
Paroxysmal slow waves, spikes, and polyspikes occur with ther-
apeutic doses (45,46). The seizure frequency may increase in
chronic epileptic patients. Furthermore, single or multiple
seizures occur in nonepileptic patients, especially with high
doses (47,48). Several cases of absence status have been thought
to be due to treatment with tricyclic agents (49,50).
Acute Intoxication
Unlike the phenothiazines, tricyclic antidepressants have much
narrower therapeutic ranges and quickly reach toxic levels.
Overdosage may result in serious life-threatening conditions.
This is of great concern, because depression is notorious for
suicidal attempts. The clinical picture is characterized by
905
Figure 43.5 P.M., a 22-year-old male.
Comatose due to intoxication with an
unidentified amount with tricyclic antide-
pressants. Unresponsive slowing with super-
imposed fast activities. The patient survived
and was transferred to the psychiatric ward
the next day.
hyperpyrexia, hypertension, seizures, and coma. The prognosis
depends largely on the effects on the cardiovascular system.
Even with therapeutic doses, there is an increased tendency
toward cardiac arrhythmias, and there have been several reports
of unexpected death. Greater than 10-fold differences in the
number of deaths per million prescriptions have been shown
comparing tricyclic antidepressants with the newer antidepres-
sants (51). The EEG during acute intoxication shows wide-
spread, poorly reactive, irregular 8- to 10-Hz activity and
paroxysmal abnormalities including spikes as well as unspecific
coma patterns (Fig. 43.5) (14).
New Antidepressants
Seizures have also been reported with the newer antidepres-
sants. There are differences in seizure propensity among the dif-
ferent substances, particularly pronounced with maprotiline
and bupropion (1.5% seizure risk) and low with trazodone,
nefazodone, mirtazapine, and the selective serotonin reuptake
inhibitors (for review see Refs. 39, 43, and 52). However,
second-generation antidepressants other than bupropion
apparently have antiepileptic properties (39).
In general, the new antidepressant drugs have a higher ther-
apeutic index than the tricyclic compounds. Overdosage with
several of the new drugs has been reported, but deaths seem to
be an exception (for review see Ref. 43). Fatal toxicity index for
venlafaxine is higher than that for other serotoninergic antide-
pressants (51). EEG signs during the intoxicated state are rarely
demonstrated and show no particular features.
The Serotonin Syndrome
The serotonin syndrome is an adverse drug reaction associated
with a number of drug–drug interactions (53). Among the drugs
at risk are selective serotonin reuptake inhibitors, other antide-
pressant drugs, and monoamine oxidase inhibitors. The protean
manifestations of the syndrome include mental status changes,
autonomic hyperactivity, and neuromuscular abnormalities.
906 Part VI ■ EEG: Neuropharmacology and Anesthesia
Figure 43.6 E.R., a 43-year-old female with intoxication from lithium carbonate. A: Delirious, multiple hyperkinesia, gen-
eral rigidity. Blood serum level of Li 2.95 mval/ L. Diffuse slowing with rhythmical 1- to 3-Hz activity maximal over anterior
regions. Some triphasic-like waves. B: Nine days after A. No abnormal results at neurologic examination. No more lithium
could be found in the blood. The EEG was also normal.
Since the condition is potentially life-threatening, an early diag-
nosis with immediate removal of the responsible drugs is essen-
tial. Case reports of the syndrome have noted EEG abnormalities
with diffuse slow activities, spikes, and triphasic waves (54).
Lithium
Lithium is still used in the prophylactic treatment of bipolar
mood disorders and has not been entirely replaced by valproate
or lamotrigine. EEG changes during the treatment are frequent
and marked (55,56). In general, they parallel the blood serum
levels, but there are also remarkable discrepancies.
Acute intoxication with lithium salts can be a life-threaten-
ing condition. Early symptoms include fatigue, muscular weak-
ness, and tremor. When plasma concentrations rise above
2 mEq/L, more serious toxic effects occur. Disturbances in renal
elimination or fever with liberation of tissue-bound lithium are
the most frequent causes of intoxication. Obtundation, stupor,
or delirium is always present. Neurologic signs show a bewilder-
ing variety of cortical and subcortical dysfunctions, optomotor
disturbances, and peripheral neuropathies (57,58). Movement
disorders are especially dramatic. Myoclonic jerks (59), convul-
sions, choreiform hyperkinesis, other complex extrapyramidal
movements, and several other signs may be observed. If accom-
panied by triphasic waves in the EEG, such conditions have
been called lithium-induced Creutzfeldt–Jakob syndrome (60)
or nonconvulsive status epilepticus (61). However, the epileptic
or encephalopathic nature of these confusional states remains
undecided (62). Permanent neurologic sequels due to lithium
toxicity have also been reported (63).
The EEG is always markedly abnormal with lithium intoxi-
cation and shows diffuse slowing, paroxysmal abnormalities
including spikes and triphasic waves. Focal slowing also occurs
and does not have to be taken as a sign of focal brain lesion
(64). Improvement of electrical activity parallels clinical
improvement, but the EEG abnormalities regularly outlast the
abnormal serum levels (Fig. 43.6A and B).
ANXIOLYTIC DRUGS
Benzodiazepines
The main actions of benzodiazepines can be described as hyp-
notic, anxiolytic, anticonvulsant, myorelaxant, and amnesic.
They exert their actions by binding at specific sites at the GABA-
A receptor chloride ionophore (65). The distribution of GABA-
A receptors within brain regions and changing compositions of
receptor subunits might contribute to differences in the efficacy
of the numerous benzodiazepines (9,66). Furthermore, pharma-
cokinetic properties like rapidity of absorption, half-life time,
binding to fat deposits, and activities of metabolites are impor-
tant differences among the substances. In general, however, all
benzodiazepines exert actions mentioned above and share the
same side effects. Therefore, those on the list of benzodiazepines
(clobazam, clonazepam, diazepam, estazolam, flunitrazepam,
flurazepam, lorazepam, nitrazepam, quazepam, temazepam,
triazolam among others) are considered together.
EEG Changes with Therapeutic Doses
The benzodiazepine derivatives are potent activators of beta
activity, which persist in the EEG as long as 2 weeks after the last
ingestion. As with barbiturates, benzodiazepine-induced fast
activities are reduced over the site of a cerebral lesion (11).
Furthermore, benzodiazepines produce a decrease in alpha
activity and general voltage, and a slight increase of 4- to 7-Hz
activity (28).
Effect on Seizures
Benzodiazepines reduce generalized spikes but have no effect on
interictal focal spikes (67) except for discharges in rolandic
 Chapter 43 ■ EEG, Drug Effects, and Central Nervous System Poisoning
epilepsy (68). They are drugs of first choice for status epilepticus
and repetitive acute symptomatic seizures (69). Despite their
antiepileptic action, they may provoke tonic status epilepticus if
intravenously administered to children with absence status in
Lennox–Gastaut syndrome (70,71). Lorazepam is superior in the
treatment of status epilepticus because it is less extensively bound
to fat (72). Withdrawal of benzodiazepines has been considered
an etiologic factor of “de novo” absence status of late onset (50).
Side Effects of Benzodiazepines
All benzodiazepines share a long list of side effects consisting of
tiredness during the day following ingestion (hangover),
rebound anxiety, anterograde amnesia, rebound insomnia, low-
dose dependency, and withdrawal syndromes. The profile of
side effects depends on the specific binding site, the dose, the
half-life time, and other pharmacokinetic variables. With short-
acting substances, the amnesic effects are comparatively marked
(73,74). In general, benzodiazepines have significantly fewer
side effects than barbiturates, and it is rational to prefer them as
hypnotics.
Acute Intoxication
The clinical picture resembles that seen with other CNS depres-
sant drugs and is not truly specific for benzodiazepines (75).
The prognosis is generally good, although in patients with
decreased respiratory reserve and in the very young even thera-
peutic doses may dangerously depress cardiorespiratory func-
tion (76). The EEG shows prominent fast activity with no
response to stimuli. With larger doses, coma patterns, as in
other intoxications, are recorded (14).
So-called Z-drugs
Zaleplon, zolpidem, and zopiclone are non-benzodiazepine hyp-
notics. Zolpidem belongs to the class of imidazopyridines (77).
It binds to the alpha 1 unit of the GABA-A receptor complex (9).
Similar side effects as with benzodiazepines have been reported.
907
Figure 43.7 M.C., a 17-year-old male. Had
a severe brain trauma 1 year ago. Marked
spasticity. Now comatose due to an over-
dose of baclofen (failure of the intrathecal
drug delivery system). Had left-sided focal
motor seizures the first day after intoxica-
tion. The EEG taken the second day exhibits
periodic paroxysmal slow and spiky waves
on a severely abnormal slow background
activity. Recovered up to the habitual condi-
tion after the brain trauma. No further
seizures.
Sleep architecture is claimed to resemble normal patterns more
closely. Zopiclone belongs to the cyclopyrrolones. The final
pathway constitutes an opening of the chloride ionophore. The
most frequent adverse effects include bitter taste, dry mouth,
and complaints comparable to the benzodiazepines (78).
Microstructural analysis of sleep architecture and decrease
of EEG arousals allow discrimination of benzodiazepines,
zolpidem, and zopiclone (79). There was no consistent pattern
of superiority in therapeutic efficacy among the Z-drugs.
Baclofen
Baclofen is a selective GABA-B receptor agonist and is used for
treatment of spasticity (80). Oral and intrathecal baclofen have
been associated with epileptic seizures (Fig. 43.7) (81,82). After
intrathecal baclofen overdose the EEG can exhibit coma pat-
terns characterized by periodic generalized epileptiform dis-
charges without overt seizures (83).
Psychotogenic Drugs
The substances under the heading of psychotogenic drugs are
intimately connected with addiction. Exciting progress has been
made elucidating the neurobiology of addiction and its relation
to various intrinsic neurotransmitter systems (84). This scien-
tifically as well as clinically important issue can only be touched
on a few references.
Lysergic Acid Diethylamide (LSD) and Mescaline
These agents cause decreased amplitude and depression of slow
waves as well as acceleration of the dominant frequencies dur-
ing the drug-induced psychotic state (45,85).
Marijuana
Cannabis exerts its central effects through the CB1 cannabinoid
receptor (for review see Ref. 86). These effects include disrup-
tion of psychomotor behavior, short-term memory impair-
ment, stimulation of appetite, and antinoceptive and antiemetic
908 Part VI ■ EEG: Neuropharmacology and Anesthesia
actions. Neuropsychologic deficits have been noted after years
of heavy frequent cannabis use (87), and the risk of psychosis in
later life seems to be increased (88). Cannabis is used in the
treatment of neuropathic pain and spasticity (89), but the risk
to benefit ratio has to be clarified.
Marijuana smoking produces no visible effects on the con-
ventional EEG (64,90). The EEG of chronic marijuana users
was associated with reduced power in the alpha and beta bands
at posterior sites (91). These EEG changes continued for the
month of abstinence.
CNS Stimulants
Central stimulants potentiate central dopamine activity.
Different modes of dopaminergic potentiation have been
shown in such drugs as the amphetamines, methylphenidate,
and cocaine (92).
Low -Dose Effects on EEG
CNS stimulants increase the amounts of beta and alpha activi-
ties, reduce the general voltage and the amount of slow waves,
and tend to suppress seizure discharges in pertinent cases, espe-
cially the 3-Hz spike-and-wave complexes (28). Abnormal slow
waves during stupor and coma are diminished by the adminis-
tration of methylphenidate. So-called recreational club drugs
illegally offered as “ecstasy” contain CNS stimulants, especially
methamphetamines and dioxymethamphetamine (MDMA)
(93). A number of acute and long-lasting medical complica-
tions are associated with abuse of these drugs. Therefore, most
CNS stimulants have been banned from the market. Modafinil,
a pharmacologically unique wake-promoting agent, has
replaced the traditional amphetamines for the treatment of nar-
colepsy (94). Psychostimulants are also used for treatment of
attention-deficit hyperactivity disorder (ADHD). Increased
slow activities in unmedicated ADHD girls are normalized
under psychostimulant therapy (95). With atomoxetine a non-
stimulant pharmacotherapy is now available for ADHD (96).
Acute Intoxication
Intoxications with CNS stimulants have increased due to their
illegal use as “ecstasy.” In a 1996 survey it was reported that
nearly 5 million Americans have used methamphetamine at
some time in their lives (97). Symptoms of mild intoxications
are those of sympathetic activity. In more severe poisoning,
hypertension, confusion, cardiac anomalies, and, finally, hyper-
thermia, convulsions, circulatory failure, and coma occur.
Death due to amphetamines is related either to direct pharma-
cologic effects or to secondary complications in drug addicts
(98). The EEG with overdose of stimulant drugs shows the
usual coma pattern without any particular features.
Methamphetamines can damage dopaminergic and serotonin-
ergic neurons (99) and produce corresponding permanent neu-
rologic abnormalities.
Cocaine
Cocaine has become a major substance in the field of drug
addiction. It binds strongly to the dopamine reuptake trans-
porter and blocks such reuptake after normal neuronal activity.
High dopamine concentrations at the synapse produce the
characteristic cocaine “high” (100). The induction of long-term
synaptic plasticity in neurons of the brain’s reward system
seems to represent a common pathway for the addictive poten-
tial of cocaine, morphine, nicotine, ethanol, and amphetamine
(101). Variants of the DRD2 gene have been associated with
cocaine, nicotine, and opioid dependency (102).
Berger’s historic experiments on the EEG effects of cocaine
intake have been confirmed. Cocaine increases beta power cor-
related with the area under the cocaine plasma versus time
curve (103). Several neurologic complications are associated
with chronic cocaine intake. It induces strokes (104–107),
orbital infarction (108), subarachnoid and intracerebral hem-
orrhages (109,110), vasospasms after aneurysmal subarachnoid
hemorrhage (111), cerebral vasculitis (112), persistent dyskine-
sias (113), oculomotor nerve palsies (114), and spinal and
medullary vascular syndromes (106), and can provoke seizures
or exacerbate a preexisting seizure disorder (92). Seizure activ-
ity can present as status epilepticus (115).
Antiepileptic Drugs
Drug therapy of epilepsies represents the long-term therapy par
excellence. Dose- or interaction-related CNS toxic effects are
common. Delayed toxic effects and drug-induced diseases are
not dose-related, occasionally life-threatening and not always
reversible. Antiepileptic drugs act directly on voltage-gated ion
channels and by influencing GABA-mediated effects. They can
bind at the GABA-A receptor at different binding sites, or can
inactivate GABA-metabolizing enzymes and inhibit uptake of
GABA into nerve cells and glia cells. Other antiepileptic drugs
influence different transporter systems or the glutaminergic
system via NMDA/AMPA receptors. For detailed and compre-
hensive information, the reader is referred to Levy et al. (116).
Several antiepileptic drugs are also used in the treatment of
neuropathic pain, migraine, and alcohol dependency, and as so-
called mood stabilizers.
EEG changes due to antiepileptic drugs have to be divided
into effects of therapeutic doses on background activity and on
frequency and morphology of preexisting spikes as well as into
the effects of overdoses, overt intoxication, and withdrawal. In
general, the standard antiepileptic drugs slow down the fre-
quency of the occipital basic rhythm even with nontoxic serum
levels and increase the percentage of power in the theta and
delta bands with visual (117,118) and with quantitative analysis
(119). These changes correlate with cognitive effects and sub-
jective complaints. The effect on interictal spikes varies consid-
erably with a positive correlation between seizure frequency
and the number of spikes in some patients (120).
Hydantoins
Signs of cerebellovestibular dysfunction signal initial hydan-
toin overdosage. Cerebellar atrophy occurs with chronic
hydantoin use but also with acute intoxication (121,122).
Overt hydantoin intoxication is further characterized by
altered higher cognitive function, pyramidal signs, and several
extrapyramidal movement disorders. Epileptic seizures may be
exacerbated (123,124). Cardiovascular toxicity is rare unless
the substance has been given parenterally for treatment of
status epilepticus.
 Chapter 43 ■ EEG, Drug Effects, and Central Nervous System Poisoning
In contrast to barbiturates and benzodiazepines, the EEG
shows no increase in fast activities with visual EEG reading
(125). Phenytoin increases the power in the theta and delta
bands with blood levels in the usual range and without clinical
signs of overt intoxication (117,118). There are conflicting
reports on the influence of phenytoin on interictal epileptiform
discharges. No changes (126), an increase (127), or a decrease of
spikes (128) have been reported.
No changes in background activity occurred with reduction
of phenytoin dosage (Duncan et al., 1989). Withdrawal of
antiepileptic drugs routinely is used as a seizure-provoking
method in intensive epilepsy monitoring. No misleading infor-
mation was gained with this procedure localizing the seizure
onset zone (129). The same was true withdrawing carba-
mazepine and valproate.
At toxic levels, phenytoin can cause marked diffuse delta
activity and paroxysmal slow-wave abnormalities (125). This is
also true in cases of chronic encephalopathy with near-normal
serum levels (130).
Carbamazepine
Carbamazepine is chemically related to the tricyclic antidepres-
sants. It may augment diffuse slow activities (118,126). Besser et
al. (131) found no correlation between serum levels of carba-
mazepine or the epoxide and the increase of power in the theta
and delta bands. Furthermore, EEG slowing was not associated
with rise in seizure frequency (120,132). Severe hyponatremia
occurred in 2.8% of carbamazepine-treated patients (133). This
adverse effect might contribute to the development of diffuse
slow activities in the EEG. Generalized paroxysmal activities
including spikes can be increased (126). Preexisting interictal
focal spikes are either increased or unchanged by carba-
mazepine (126). New appearance of generalized epileptiform
discharges after introduction of carbamazepine treatment was
observed, and spikes can be accompanied by seizure exacerba-
tion (134–136). Carbamazepine overdose may also exacerbate
909
Figure 43.8 K.B., a 17-year-old female.
Temporal lobe epilepsy with rare seizures.
Self-poisoning with a huge amount of a
slow release preparation of carbamazepine.
CBZ level the next day 16 g/ mL. Soporos.
Diffuse slowing and 5- to 6-Hz rhythmical
activities over the posterior regions.
Recovered promptly.
seizures (123,137). However, several other antiepileptic drugs
can worsen seizures. Overdosage represents only one of the
three possible causes (124).
Carbamazepine was claimed to have fewer adverse neu-
ropsychological effects than phenobarbital (138). However, car-
bamazepine has cognitive side effects without clinically
significant differences to phenytoin (139). Clinical signs of
acute poisoning include ataxia, nystagmus, diplopia, drowsi-
ness, and diffuse slowing in EEG (Fig. 43.8). Cardiovascular
complications are common (140). Coma indicates a severe
intoxication, and EEG can deteriorate to burst-suppression pat-
tern with accompanying myoclonus (137).
Oxcarbazepine
Oxcarbazepine is an analog of carbamazepine and shares its
efficacy against focal seizures (for summary see Ref. 141).
Oxcarbazepine is not metabolized via the epoxide-diol pathway,
possibly resulting in fewer adverse events. However, hypona-
tremia is more frequent and more severe with oxcarbazepine
than with carbamazepine (133). EEG changes were character-
ized by generalized spikes not seen prior to oxcarbazepine,
occasionally accompanied by worsening of generalized seizure
types (142,143).
Succinimides
The use of succinimides directed against absence seizures has
declined in recent years due to antiepileptic drugs with a
broader efficacy spectrum (valproate, lamotrigine). The succin-
imides sometimes cause somnolence, lethargy, and emotional
instability accompanied by signs of altered vigilance in EEG.
Acute intoxication has been reported rarely (see Ref. 144).
Valproic Acid
Valproate in the usual doses did not significantly change EEG
background activity (117). With processed EEG analysis a
decreased EEG synchronization in the delta and theta frequencies
910 Part VI ■ EEG: Neuropharmacology and Anesthesia
was observed (145). The most important change consists of
reduction or even disappearance of generalized spikes along with
seizure reduction (146–149; Villareal et al., 1978). Photosensitive
spikes likewise disappear with valproate treatment (150).
Intoxication with valproate was occasionally accompanied
by marked diffuse slowing considered to be partly caused by
drug interactions (151). Enormous interindividual differences
have been observed with acute valproate poisoning.
Drowsiness, stupor, and coma occur with normal doses of val-
proic acid, with or without evident metabolic changes such as
hyperammonemia and low carnitine (for review see Refs. 152
and 153). In the diagnosis of valproate-associated encephalopa-
thy the EEG plays an important role. It is characterized by a dra-
matic change in sequential recordings with the occurrence of
bilaterally synchronous high-voltage slow-wave activity. Case
reports exist about valproic acid-induced encephalopathy with
triphasic waves (154).
Fatal liver failure with valproate therapy has been reported
(155–157). There are no indications that repeated EEG records
can herald this insidious complication.
Although valproate is increasingly used for treatment of sta-
tus epilepticus, the substance by itself can trigger tonic status
epilepticus (158).
Clonazepam
Clonazepam is a benzodiazepine and produces the correspon-
ding EEG changes (159). It has little influence on focal interictal
epileptic activity (160). Given intravenously, the substance is a
powerful blocker of status epilepticus, but the effect remains
temporary in absence status occurring in patients suffering
from Lennox–Gastaut syndrome. Like diazepam, clonazepam
can produce tonic seizures if given intravenously in this setting
(161).
Vigabatrin
Vigabatrin binds irreversibly to GABA transaminase and
inhibits the catabolism of the neurotransmitter. GABA is
increased in the brain after vigabatrin exposure (for review see
Ref. 162).
In humans, one study reports diffuse slowing with vigaba-
trin (163), whereas several other studies found no change in
background EEG activity (164–166). No consistent influence
on interictal spikes was reported by Ben-Menachem and
Treiman (167) and Marciani et al. (163). On the other hand,
development of generalized spikes was seen, occasionally
accompanied by myoclonic jerks or absence seizures, even in
patients with focal epilepsies (168–171). An excess GABA con-
tent in the brain was held responsible for the induction of gen-
eralized spikes (see also the section “Tiagabine”).
Acute encephalopathies characterized by EEG abnormalities
occurred after the start of vigabatrin treatment (171,172).
These anecdotical case reports do not allow a decision whether
these encephalopathies are related to comedication or to a pre-
existing cerebral anomaly. There are few reports of overdoses
with vigabatrin (see Ref. 144). The development of psychoses in
the course of vigabatrin treatment was not uniformly related to
overdose (173).
Observations of concentric and irreversible visual field
deficits (174) limited the use of the substance to the West syn-
drome and to otherwise untreatable focal epilepsies (175). The
maximum daily dose was the single most important factor for
visual field deficits (176).
Lamotrigine
Lamotrigine blocks voltage-gated sodium channels (for review
see Ref. 177). Rash is the most common cause for withdrawal of
lamotrigine treatment. Comedication with valproic acid and
rapid titration are risk factors.
No slowing in background activity was seen in volunteers
(178) and in epileptic patients (179). Lamotrigine has a marked
decreasing effect on the occurrence and frequency of sponta-
neous and photosensitive generalized spikes and waves
(180,181) as well as on interictal epileptiform activity in young
patients with drug-resistant epilepsy (182).
Lamotrigine can lead to exacerbation of myoclonic epilep-
sies (183,184), and triggers convulsive (185) and nonconvulsive
status epilepticus (186). One report deals with self-poisoning
with lamotrigine (187). No serious toxicity was observed. There
was a prolongation of QRS in ECG, but no EEG examination
was reported.
Gabapentin
Gabapentin is a chemical derivative of GABA, which penetrates
the blood brain barrier. The substance has proven antiepileptic
properties. There are several lines of evidence for its mecha-
nisms of action in a variety of animal models, which suggest
that gabapentin differs pharmacologically from other
antiepileptic drugs (for review see Ref. 188). Gabapentin is
extensively used in other indications than epilepsy like migraine
(189), chronic daily headache (190), painful neuropathy (191),
and psychiatric disorders (for summary see Ref. 192).
Gabapentin toxicity is low (for review see Ref. 193). The
most common adverse events are somnolence, dizziness, and
ataxia. Overdoses showed no serious toxicity (194).
Prolonged therapy with gabapentin induced EEG slowing
that correlates with cognitive complaints (195). Gabapentin-
treated subjects had an increase in slow-wave sleep compared
with baseline (196).
Pregabalin
Like gabapentin pregabalin is a structural analog of GABA
(197). Its action is not entirely clarified. Compared with
gabapentin, pregabalin seems to be more potent against
seizures (198) and pain disorders (199). To our knowledge no
systematic EEG studies have been published. Myoclonic jerks
have been observed associated with pregabalin treatment
(200,201). The EEG obtained in these patients exhibited no
changes time-locked to the jerks, so their cortical origin was not
proven. In two patients treated with pregabalin for chronic pain
generalized myoclonic status epilepticus occurred (202).
Pregabalin withdrawal led to a delirious encephalopathy in a
postherpetic-neuralgia patient (203). MRI showed a splenial
edema, but no EEG report was added.
 Chapter 43
Felbamate
Felbamate is the first drug with proven efficacy against seizures
occurring in Lennox–Gastaut syndrome (204). Observations of
aplastic anemia and liver failure including fatalities have been
reported after the drug was marketed in the United States
(Kaufman et al., 1997). The use of felbamate now is limited to
patients with otherwise intractable epilepsies (French et al.,
1999) (205). Some overdoses have been reported with low
toxicities. Published EEG investigations are limited to animal
experiments.
Tiagabine
Tiagabine has an established mode of anticonvulsant action via
inhibition of GABA uptake. The substance has a proven
antiepileptic potency in focal seizures and a low toxicity (for
review see Ref. 206). Overdose has been reported (207).
The effects of tiagabine on electroencephalogram and spike-
wave discharges have been studied in animals (208,209). In a rat
model of absence epilepsy spike waves and other forms of parox-
ysmal activity were facilitated by tiagabine. As with vigabatrin, an
increase of GABA content in the brain can result in generalized
EEG abnormalities and seizures in epilepsy patients. Several cases
of nonconvulsive status epilepticus have been reported during
therapy with tiagabine, in seven patients electroclinically con-
firmed (revisited by Koepp et al. (210)). The FDA safety informa-
tion and adverse event reporting program warned that seizures
and status epilepticus are known to occur with Tiagabin.
Topiramate
Topiramate is a broad-spectrum antiepileptic drug with multi-
ple modes of action (for summary see Ref. 211). Cognitive and
behavioral effects are slightly worse than with valproic acid
(212) or with lamotrigine (213). Language deficits are a specific
side effect (214,215). No changes were seen on EEG background
measures or on daytime vigilance (Salinsky et al., 2007) (216).
Levetiracetam
Levetiracetam is chemically related to piracetam and was
screened for second-generation nootropic substances. It is
devoid of anticonvulsant activities in traditional seizure screen-
ing models but exhibits a broad antiepileptic spectrum in
epileptic patients. So-called antiepileptogenic properties in
experimental investigations have been claimed to be clinically
relevant (for review see Ref. 217). Levetiracetam has become a
major antiepileptic drug for partial epilepsies. Antimyoclonic
efficacy has been reported recently (218–220).
Levetiracetam reduces the corticospinal response to mag-
netic stimulation (221). If related to serum levels, there was a
substantial time lag to these responses (222). This effect might
correspond to a sustained efficacy against seizures in refractory
epilepsies (223). The amount of interictal spikes in epilepsy
patients was reduced by levetiracetam. (224). It has no major
effects on sleep nature (225).
In general, levetiracetam is well tolerated. Adverse psychi-
atric effects occur in a group of patients generally prone to such
complications during therapy with antiepileptic drugs (226). A
■ EEG, Drug Effects, and Central Nervous System Poisoning 911
reversible encephalopathy with marked delta slowing in EEG
was reported under levetiracetam (227).
Zonisamide
Zonisamide is a 1,2-benzisoxazole with a sulfonamide side
chain. Its antiepileptic profile is similar to hydantoins and car-
bamazepine with multiple mechanisms of action (for summary
see Ref. 228). The development of the substance started in
Europe and the United States and was stopped because of high
incidence of renal calculi (229). Afterwards, zonisamide was
further used in Japan (230), and finally reintroduced in the
Western countries (231).
Zonisamide decreases cortical excitability in patients with
idiopathic generalized epilepsy (232). After zonisamide an
almost complete resolution of generalized spike-and-wave dis-
charges and EEG absence seizures have been noted in a patient
with juvenile myoclonic epilepsy (233). To our knowledge, no
systematic study deals with EEG changes after zonisamide.
Lacosamide
Lacosamide was recently approved as an antiepileptic drug in
Europe. It is a member of a series of amino acids and has a dual
mode of action. It selectively enhances slow inactivation of volt-
age-gated sodium channels, and has a collapsin response medi-
ator protein as a binding partner (234). Lacosamide employs
proven antiepileptic properties in adults with partial-onset
seizures (235). EEG effects are not investigated.
Other Drugs
A great number of drugs produce EEG changes and CNS intox-
ications (for review see Ref. 14). The following part selects some
drugs with particular interest for neurologists.
Morphine and the Opiates
Morphine and the opiates produce only mild to moderate effects
on the EEG; the frequency of the alpha rhythm may be slowed
down, and paroxysmal changes may occasionally appear. REM
sleep depression has been observed. In neonates, morphine pro-
duces profound reversible EEG alterations, which have to be
taken into consideration in the interpretation of abnormalities in
these age groups (236). With opioid anesthesia spike discharges
resembling benign epileptiform transients of sleep appeared
within 3 minutes of the first opioid dose in 19 of 20 patients
undergoing coronary artery revascularization (237). Abrupt ces-
sation of spikes after administration of midazolam was held an
argument for an epileptogenic mechanism of the opioid-induced
activity. Morphine and heroin addicts usually show unremark-
able EEG tracings unless they present in a comatose state after
overdose, which produces diffuse slowing. EEG abnormalities
observed in an alert addict should prompt the suspicion of a
HIV-related brain disease.
Aminophylline
Aminophylline is a bronchodilator commonly used for the
treatment of chronic destructive pulmonary disease. Agitation,
tremor, delirium, and even coma occur with therapeutic doses.
Acute convulsions may complicate a comatose state (238). Focal
motor status epilepticus (239,240) might be accompanied by
912 Part VI ■ EEG: Neuropharmacology and Anesthesia

Figure 43.9 K.H., a 58-year-old male. Chronic asthmatic and alcoholic. Relapsing episodes of focal motor status with
epilepsia partialis continua-like features with twitchings of left facial and upper extremity muscles. Euphyllin overdose
due to self-medication. No abnormalities with CT and MRI scan. The EEG exhibited constantly repeated episodes of left
parieto-temporal slow and superimposed fast activities accompanied by periodic twitches (see muscle artifacts).

periodic lateralized epileptiform discharges (PLEDs) in the
EEG (Fig. 43.9). Aminophylline-associated seizures are difficult
to stop by diazepines. Besides the removal of the responsible
substance, the use of barbiturates as the drug of first choice was
recommended (241).
Isoniazid (INH)
INH, a highly effective tuberculostatic drug, is known to inter-
act with several antiepileptic drugs. This interaction may lead to
intoxication from previously well-tolerated doses of anticon-
vulsants. Overdose of INH can produce coma complicated by
acute convulsions, metabolic acidosis, hyperglycemia, and ace-
tonuria (242). The interictal EEG is characterized by diffuse
slowing and generalized paroxysmal slow waves with intermin-
gled bilateral sharp waves (243). With a large overdose of INH,
status epilepticus has been observed (244).
ANTIBIOTICS
Penicillin
The parenteral administration of very large doses of penicillin
G (40 to 80 million units/day) may produce jerks, generalized
seizures, or even status epilepticus (245). In the treatment of
status epilepticus of other etiologies, penicillin should not be
administered as an antibiotic drug.
Cephalosporins
Cephalosporin antibiotics comprise a varied group of beta-
lactam antibiotics. They produce seizures and nonconvulsive
status epilepticus with corresponding EEG abnormalities
(246,247). The confusional states after cephalosporin have also
been called encephalopathy with diffuse rhythmic nonreactive
triphasic sharp waves in the EEG (248).
Immunosuppressant Drugs
Cyclosporine is a lipophillic cyclic undecapeptide, which today
is the major immunosuppressant after organ transplantations.
It is also used in the treatment of rheumatic and other autoim-
mune diseases. Cyclosporine neurotoxicity is one of the most
significant clinical side effects occurring in up to 60% of
patients (for summary see Ref. 249). The clinical symptoms
consist of encephalopathies, seizures, and stroke-like episodes
among others. In these cases the EEG exhibited diffuse and focal
slowing and epileptiform discharges (see Ref. 249). A significant
risk for seizure recurrence was associated with persistent EEG
abnormalities (250).
 Chapter 43 ■ EEG, Drug Effects, and Central Nervous System Poisoning
Antineoplastic Agents
A rapidly growing number of antineoplastic substances increase
the armamentarium fighting cancer. Many, if not all, of these
drugs cause encephalopathies with personality changes, confu-
sions, hallucinations, and coma (see also Chapter 17). Singh et al.
(251) reviewed seizures and epilepsy in people with cancers other
than primary brain tumors. Among different other causes anti-
neoplastic agents trigger seizures through varied mechanisms.
Detailed reports exist for ifosfamide. It causes nonconvulsive sta-
tus epilepticus (252,253), diazepam-sensitive encephalopathy
(254), and encephalopathy characterized by periodic generalized
triphasic waves and negative myoclonus (255).
Salicylate
Salicylate overdose occurs with suicidal attempts or acciden-
tally. Most persons lack prominent neurologic symptoms, but
severe illness with coma and seizures can occur (15). Two
reports mention EEG changes (256,257). The records showed
diffuse slowing accentuated over the anterior regions and occa-
sionally in rhythmical trains.
Contrast Media
Metrizamide (Amipaque) is a nonionic water-soluble radio-
logic contrast agent. Metrizamide is used for myelography and
cisternography. Nausea, vomiting, myoclonus, seizures, epilep-
sia partialis continua (258), stroke-like episodes, and aphasia
are symptoms of metrizamide-induced encephalopathy. Several
cases of nonconvulsive status epilepticus have also been
reported (Fig. 43.10) (259–262). Ropper et al. (263) reported
EEG changes after metrizamide for myelography and posterior
fossa cisternography in 61 patients. The EEG exhibited abnor-
malities in 34% of patients. A number of changes required 24 to
48 hours to develop.
After myelography with iopamidol, sold under the trade
name Omnipaque, status epilepticus was reported (264). The
EEG demonstrated the evolution of lateralized rhythmic ictal
913
activity to PLEDs and, with resolution of the status, to contin-
uous lateralized slow activities. Seizures have also been seen
after intravenous application of several different contrast media
(265). In a case of gadolinium-related encephalopathy the EEG
showed generalized atypical triphasic waves (266).
Levodopa
Levodopa is the main drug used to treat symptoms of
Parkinson disease. Reports on EEG changes are sparse. Neufeld
(267) described four patients in subacute confusional states fol-
lowing an increase in the dose of levodopa. The EEG revealed
periodic generalized triphasic waves.
Sildafenil
Sildafenil (Viagra R) is the first drug approved by drug adminis-
trations to treat erectile dysfunction. In mice, sildafenil induces
a proconvulsant effect, probably due to the release of nitric oxide
(268). Generalized tonic–clonic seizures are reported after
sildafenil intake in two patients (269). Seizures also occurred
after vardenafil, a substance related to sildafenil (270). In both
reports the EEG even after sleep deprivation has been normal.
Toxic Encephalopathies
A selection of toxic encephalopathies relevant for neurologists
is given in the following section. For a comprehensive review
the reader is referred to textbooks of toxicology and to Mellerio
and Kubicki (271).
Toxic encephalopathies present themselves as acute, suba-
cute, or chronic disorders. Neurologic and EEG abnormalities
indicate alterations of neuronal structures, receptor composi-
tion and sensitivity, or neuronal death. In general, clinical signs
may be classified as coma, organic mental impairment, seizures,
movement disorders, involvement of cranial and spinal periph-
eral nerves, and neuromuscular dysfunction. EEG changes
occur with coma, acute encephalopathies, and severe chronic
cerebral deficit syndromes.
Figure 43.10 E.A., a 56-year-old female.
Comatose after myelography with metriza-
mide. Continuous repetitive triphasic
waves. These conditions have also been
called symptomatic nonconvulsive status
(see text).
914 Part VI ■ EEG: Neuropharmacology and Anesthesia
Lead
The neurotoxic actions of lead include apoptosis, excitotoxicity,
and influences on neurotransmitter storage and release processes,
mitochondria, second messengers, cerebrovascular endothelial
cells, and glia cells (for review see Ref. 272). Chronic exposure to
lead is associated with intellectual impairment. Declines in IQ
have been correlated to blood lead concentrations, even those
below 10 g/dL (273). Neurodegeneration linked to cumulative
lead exposure can be measured by MRI (274).
In acute intoxications, the EEG shows the usual signs of dif-
fuse encephalopathies. Fejerman et al. (275) reported lead
encephalopathy as a cause of a Lennox–Gastaut syndrome with
slow spikes and waves. In chronic forms, the EEG is inconclu-
sive (Burchfield et al., 1980) (276).
Mercury
Exposure to high doses of methyl mercury causes devastating
damage to the nervous system, resulting in abnormalities in
motor function and impairment in the visual, auditory, and
somatosensory systems (for review see Ref. 277). Chronic toxi-
city is related to industrial exposure and mainly affects the kid-
ney. Exposure to mercury from dental amalgams and fish
consumption has been a concern for decades. The hypothesis of
prenatal mercury as a neurodevelopmental risk factor could not
be supported (278). Ayurvedic medicines contain detectable
lead, mercury, or arsenic (279). A negative impact to human’s
health, however, is not substantiated.
The EEG abnormalities consist of unspecific diffuse slowing
and reflect the clinical state (280). Quantified EEG may docu-
ment early effects of exposure to mercury vapor (281).
Manganese
The most characteristic for manganese poisoning is parkinson-
ism. After prolonged exposure spasticity, epileptic seizures, and
dementia may be associated (282). Myoclonus and high-inten-
sity signals in the globus pallidus on T1-weighted MR images
have been reported (283). EEG slowing and fast activities (271)
as well as normal records were observed.
Aluminum
Toxic effects of aluminum have been recognized particularly
with hemodialysis treatment for chronic renal failure. The EEG
in the dialysis encephalopathy syndrome (284) shows diffuse
slow activities and spikes. However, the disease virtually has dis-
appeared with improvement of dialysis technique.
Thallium
Thallium has been widely used as a raticide. Ingestion of thal-
lium occurs with suicidal or murderous intention. The clinical
picture and the EEG signs are essentially of nonspecific charac-
ter (285). Figure 43.11A and B demonstrates the EEG course in
a case of criminal intoxication.
Carbon Monoxide
Carbon monoxide intoxication is a frequent incident in coun-
tries with traditional heating. It is still the leading cause of death
by poisoning in the United States (for review see Ref. 286).
Motor vehicle exhaust gases are the most common source. The
CNS effects are related to its extremely strong affinity to hemo-
globin. If the acute intoxication with different comatose stages
is survived, vegetative states, diffuse encephalopathies, focal
abnormalities, and extrapyramidal syndromes may persist and
may be accompanied by corresponding MRI findings. A delayed
encephalopathy appears after a period free of symptoms
(287,288).
Diffuse or focal epileptiform abnormalities may be seen in
acute and protracted encephalopathies (289). Leweke et al.
(1999) found a good correlation between the clinical and EEG
course. An improvement of the occipital alpha activity was
observed by hyperbaric oxygen therapy (290). This effect was
held an indicator for preventing delayed neuropsychiatric
sequels of carbon monoxide poisoning.
Methyl Alcohol (Methanol)
Methyl alcohol is metabolized to formaldehyde and formic
acid. Formaldehyde is particularly toxic for retinal cells, and
formic acid causes acidosis, generally considered the main cause
of CNS involvement (291). Intoxications mostly occur with low
social status and may acquire an epidemic character. Leading
symptoms are visual disturbances and consecutive permanent
blindness. Epileptic seizures, stupor, and coma are signs of
severe acute intoxications. Necrosis of the putamen and cere-
bellar cortex and corresponding parkinsonian syndromes (292)
have been reported.
The EEG shows marked slowing that correlates mainly with
acidosis rather than with blood and cerebrospinal fluid
methanol level (293).
Ethylene Glycol
Ethylene glycol is used as antifreeze. Intoxications occur in
developing countries, if the substance is used for elixir prepara-
tions of drugs, with suicidal attempts, and if the substance is
illegally added to wine. Glycol causes renal failure and death
(294,295) as well as peripheral nerve palsies mimicking
polyradiculitis (296–298). In comatose states with acute intoxi-
cation the ubiquitous EEG patterns without any specific signs
can be observed.
Ethyl Alcohol (Ethanol)
Hallucinosis, delirium tremens, and epileptic seizures are usu-
ally understood as withdrawal signs following dependence and
habituation. EEG activity is desynchronized, voltages are low,
and generalized spikes may be encountered, if recording occurs
within the first 48 hours after withdrawal (299,300). Seizures
and spikes reflect reduced neurotransmission in GABA and
NMDA pathways (301). A constantly abnormal EEG in an alco-
holic with seizures suggests epilepsy or symptomatic seizures
unrelated to alcohol (302). Chronic alcohol use was not consid-
ered a risk factor for a first seizure symptomatic to other etiolo-
gies (303).
Wernicke encephalopathy, Korsakov syndrome, sensorimo-
tor polyneuropathy, retrobulbar neuritis, and cerebral pellagra
are related to nutritional deficiencies, mainly aneurine defi-
ciency, which serves as a cofactor in many enzyme systems. In
 Chapter 43 ■ EEG, Drug Effects, and Central Nervous System Poisoning
the Wernicke encephalopathy the EEG is variable, showing
mainly diffuse slowing. With other alcohol-related cerebral dis-
turbances severe dysfunctions are reflected in EEG slowing.
A subacute encephalopathy in alcoholics may lead to focal
seizures, transient focal cortical deficits, and PLEDs in the EEG
(304). Moreover, alcohol withdrawal has been shown to activate
PLEDs in chronic focal cerebral lesions (305). Although still
controversial, in some instances seizures may be directly related
to alcohol intoxication (306).
Organic Solvents
Trichloroethylene and methylbenzene (toluene) are used as
industrial solvents. Workers are at an occupational hazard for
intoxication. A major source for intoxication represents the sol-
vent abuse by children and young adults in developing coun-
tries (307). Neurotoxicity leads to cranial and peripheral
neuropathies, cerebellar, pyramidal, and extrapyramidal signs
as well as cognitive decline. MR images revealed white matter
lesions, brain atrophy, and thalamic hypointensity (308). A
chronic toxic encephalopathy can be recognized after excessive
occupational exposure to solvents (309). Furthermore,
trichloroethylene acts as a risk factor for the development of
915
Figure 43.11 A: N.R., 41 years. Was
transferred to the neurologic ICU in coma-
tose state with some irregular myoclonic
jerks. Diffuse 1- to 3-Hz activities with
triphasic waves, more on the left side.
Note current artifact due to high imped-
ance T4. Some muscle artifacts due to
involuntary movements. B: Ten days after
A. Investigations by police detected a mur-
derous attempt with thallium. The condi-
tion of the patient changed to a vegetative
state. Diffuse slow activities with continu-
ously superimposed muscle artifacts and
some eye movement artifacts. The patient
died due to intercurrent infections.
parkinsonism (310). Persistent EEG alterations with severe
paroxysmal abnormalities have been reported (311).
Organophosphate Pesticides
Organophosphates exhibit anticholinesterase activity. Sources of
organophosphate pesticides are garden sheds, room sprays, baits
for insects, shampoos against head lice, and pet preparations
(312). Organophosphates were used as sarin by terrorist attacks
in Japan and found and destroyed in the Gulf war. Three different
syndromes can be recognized (313). The acute cholinergic crisis
includes muscle weakness, respiratory failure, massive bronchor-
rhea, epileptic seizures, coma, and death. An intermediate syn-
drome is characterized by pre- and postsynaptic neuromuscular
junction failure. Late syndromes show polyneuropathy and
parkinsonism. A highly significant prolongation of P300 latency
highlighted the possibility of the development of long-lasting
cognitive deficits following poisoning (314).
EEG changes in organophosphorus poisoning show nonspe-
cific enhancement of slow activity and paroxysmal discharges
(14,243). According to Okonek and Rieger (315), there is a char-
acteristic sequence of EEG patterns in alkyl phosphate poisoning.
Unlike other acute intoxications, in a first stage of deep coma fast
916 Part VI ■ EEG: Neuropharmacology and Anesthesia
rhythmical activities were recorded. Paroxysmal discharges were
found to be exceptional although massive myoclonic jerks and
other convulsive manifestations occurred.
Further details about environmental toxins may be gleaned
from Lockwood (316).
REFERENCES
1. Nakamura Y. Pharmacogenomics and drug toxicity. New Engl J
Med. 2008;359:856–858.
2. Blume WT, Kaibara M, Young GB. Atlas of Adult
Electroencephalography. 2nd ed. Philadelphia: Lippincott Williams
& Wlikins; 2002:266.
3. Franks NP, Lieb WR. Molecular and cellular mechanisms of gen-
eral anesthesia. Nature. 1994;367:607–613.
4. Campagna JA, Miller KW, Forman SA. Mechanisms of action of
inhaled anesthetics. New Engl J Med. 2003;348:2110–2124.
5. Winters WD. Effects of drugs on the electrical activity of the brain:
anesthetics. Annu Rev Pharmacol Toxicol. 1976;16:413–426.
6. Sloan TB. Anesthetic effects on electrophysiologic recordings
[review articles]. J Clin Neurophysiol. 1998;15:217–226.
7. Chatrian GE. Intraoperative electrocorticography. In: Ebersole JS,
Pedley TA, eds. Current Practice in Clinical Electroencephalography.
3rd ed. Lippincott Williams & Wilkins; 2003:681–712.
8. Kwan P, Brodie MJ. Phenobarbital for the treatment of epilepsy in
the 21st century: a critical review. Epilepsia. 2004;45:1141–1149.
9. McKernan RM, Whiting PJ. Which GABA-A receptor subtypes
really occur in the brain? Trends Neurosci. 1996;19:139–143.
10. Ban TA, Amin M. Hypnotics, minor tranquilizers and sedatives. In:
Vinken PJ, Bruyn GW, eds. Handbook of Clinical Neurology. Vol 37.
Amsterdam: North Holland; 1979;347–364.
11. Gotman J, Gloor P, Quesney LF, et al. Correlations between EEG
changes induced by diazepam and the localization of epileptic
spikes and seizures. Electroencephalogr Clin Neurophysiol.
1982;54:614–621.
12. Trimble MR. Neurobehavioral effects of anticonvulsants. JAMA.
1991;13:1307–1308.
13. Cramer JA, Mattson RH. Phenobarbital toxicity. In: Levy RH,
Mattson RH, Meldrum BS, eds. Antiepileptic Drugs. 4th ed. New
York: Raven Press; 1955:409–420.
14. Kurtz D. The EEG in acute and chronic drug intoxications. In:
Glaser GH, ed. Metabolic, Endocrine and Toxic Diseases/Handbook
of Electroencephalography and Clinical Neurophysiology. Vol 15.
Amsterdam: Elsevier; 1976:88–104.
15. Plum F, Posner JP. Diagnosis of Stupor and Coma. 2nd ed.
Philadelphia: F.A. Davis; 1972.
16. Greenberg DA, Simon RP. Flexor and extensor postures in sedative
drug-induced coma. Neurology. 1982;32:448–451.
17. Kuroiwa Y, Furukawa T, Inaki K. Recovery from drug-induced
alpha coma. Neurology. 1981;31:1359–1361.
18. Ellington AL. Electroencephalographic pattern of burst suppres-
sion in a case of barbiturate coma. Electroencephalogr Clin
Neurophysiol. 1968;25:491–493.
19. Kubicki S, Rieger H, Busse G, et al. Elektroencephalographische
Befunde bei schweren Schlafmittelvergiftungen. Z EEG-EMG.
1970;1:80–93.
20. Bennett DR, Hughes JR, Korein J, et al. Atlas of
Electroencephalography in Coma and Cerebral Death. New York:
Raven Press; 1976.
21. Haider J, Matthew H, Oswald J. Electroencephalographic changes
in acute drug poisoning. Electroencephalogr Clin Neurophysiol.
1971;30:23–31.
22. Utiger RD. Melatonin—the hormone of darkness. New Engl J Med.
1992;327:1377–1379 [editorial].
23. Dahlitz M, Alvarez B, Vignau J, et al. Delayed sleep phase syndrome
response to melatonin. Lancet. 1991;337:1121–1124.
24. Buscemi N, Vandermeer B, Hooton N, et al. Efficacy and safety of
exogenous melatonin for secondary sleep disorders and sleep dis-
orders accompanying sleep restriction: metaanalysis. Br Med J.
2006;332:385–393.
25. Simpson D, Curran M. Ramelton: a review of its use in insomnia.
Drugs. 2008;68:1901–1919.
26. Dreifuss FE, Bertram EH. Bromide therapy for intractable
seizures. Epilepsia. 1986;27:593.
27. Freedman R. Schizophrenia. New Engl J Med. 2003;349:1738–1749.
28. Saletu B. Psychopharmaka, Gehirntätigkeit und Schlaf. Bibliotheka
Psychiatrica. No. 155. Basel: S. Karger; 1976.
29. Centorrino F, Price BH, Tuttle M, et al. EEG abnormalities during
treatment with typical and atypical antipsychotics. Am J
Psychiatry. 2002;159:109–115.
30. Alldredge BK. Seizure risk associated with psychotropic drugs:
clinical and pharmacokinetic considerations. Neurology.
1999;53(suppl 2):S68–S75.
31. Pisani F, Oteri G, Costa C, et al. Effects of psychotropic drugs on
seizure threshold. Drug Saf. 2002;25:91–110.
32. Van Sweden B. Toxic “ictal” confusion in middle age: treatment with
benzodiazepines. J Neurol Neurosurg Psychiatry. 1985;48:472–476.
33. Leucht St, Wahlbeck K, Hamann J, et al. New generation antipsy-
chotics versus low-potency conventional antipsychotics: a system-
atic review and meta-analysis. Lancet. 2003;361:1581–1589.
34. Schmidt CJ, Sorensen SM, Kehne JH, et al. The role of 5-HT2a
receptors in antipsychotic activity. Life Sci. 1995;56:2209–2222.
35. Alvir JMJ, Liebermann JA, Saffermann AZ, et al. Clozapine-
induced agranulocytosis: incidence and risk factors in the United
States. New Engl J Med. 1993;329:162–167.
36. Freudenreich O, Weiner RD, McEvoy JP. Clozapine-induced elec-
troencephalogram changes as a function of clozapine serum levels.
Biol Psychiatry. 1997;42:132–137.
37. Joutsiniemi S-L, Gross A, Appelberg B. Marked clozapine-induced
slowing of EEG background over frontal, central, and parietal
scalp areas in schizophrenic patients. J Clin Neurophysiol.
2001;18:9–13.
38. Baldessarini RJ, Frankenburg FR. Clozapine: a novel antipsychotic
agent. N Engl J Med. 1991;324:746–754.
39. Alper K, Schwartz KA, Kolts RL, et al. Seizure incidence in psy-
chopharmacological clinical trials: an analysis of Food and Drug
Administration (FDA) summary basis of approval reports. Biol
Psychiatry. 2007;62:345–354.
40. Sajatovic M, Meltzer HY. Clozapine-induced myoclonus and gen-
eralized seizures. Biol Psychiatry. 1996;39:367–370.
41. Mortimer AM. Newer and older antipsychotics. A comparative
review of appropriate use. CNS Drugs. 1994;2:381–396.
42. Cunnigham Owens DG. Adverse effects of antipsychotic agents.
Drugs. 1996;51:895–930.
43. Frazer A. Pharmacology of antidepressants. J Clin Pharmacol.
1997;17(suppl 1):S1–S18.
44. Sindrup SH, Jensen TS. Pharmacologic treatment of pain in
polyneuropathy. Neurology. 2000;55:915–920.
45. Itil TM. Convulsive and anticonvulsive properties of neuro-psy-
cho-pharmaca. In: Niedermeyer E, ed. Epilepsy. Modern Problems
in Pharmacopsychiatry. Vol 4. Basel: Karger; 1970:270–305.
46. Kugler J, Lorenzi E, Spatz R, et al. Drug-induced paroxysmal EEG
activities. Pharmacopsychiatry. 1979;12:165–172.
47. Jabbari B, Byran GE, Marsh EE, et al. Incidence of seizures with tri-
cyclic and tetracyclic antidepressants. Arch Neurol. 1985;42:480–481.
 Chapter 43 ■ EEG, Drug Effects, and Central Nervous System Poisoning
48. Skowron DM, Stimmel GL. Antidepressants and the risk of
seizures. Pharmacotherapy. 1992;12:18–22.
49. Rumpl E, Hinterhuber H. Unusual “spike-wave stupor” in a
patient with manic-depressive psychosis treated with amitripty-
line. J Neurol. 1981;226:131–135.
50. Thomas P, Beaumanoir A, Genton P, et al. “De novo” absence sta-
tus of late onset: report of 11 cases. Neurology. 1992;42:104–110.
51. Buckley NA, McManus PR. Fatal toxicity of serotoninergic and
other antidepressant drugs: analysis of United Kingdom mortality
data. Br Med J. 2002;325:1332–1333.
52. Schmitz B. Antidepressant drugs: indications and guidelines for
use in epilepsy. Epilepsia. 2002;43(suppl 2):14–18.
53. Boyer EW, Shannon M. The serotonin syndrome. N Engl J Med.
2005;352:1112–1120.
54. Dike GL. Triphasic waves in serotonin syndrome. J Neurol
Neurosurg Psychiatry. 1997;62:200.
55. Fetzer J, Kader G, Danahy S. Lithium encephalopathy: a clinical,
psychiatric and EEG evaluation. Am J Psychiatry. 1981;138:
1622–1623.
56. Mayfield D, Brown RG. The clinical, laboratory and electroen-
cephalographic effects of lithium. J Psychiatr Res. 1966;4:207–219.
57. Lee MS, Lessel S. Lithium-induced periodic alternating nystagmus.
Neurology. 2003;60:344.
58. Sansone MEG, Ziegler DK. Lithium toxicity: a review of neuro-
logic complications. Clin Neuropharmacol. 1985;8:242–248.
59. Caviness JN, Evidente JN. Cortical myoclonus during lithium
exposure. Arch Neurol. 2003;60:401–404.
60. Casanova B, de Entrambasaguas M, Perla C, et al. Lithium-induced
Creutzfeldt–Jakob syndrome. Clin Neuropharmacol. 1996;19:356–359.
61. Lee SI. Nonconvulsive status epilepticus. Ictal confusion in later
life. Arch Neurol. 1985;42:778–781.
62. Kaplan PW, Birbeck G. Lithium-induced confusional states: non-
convulsive status epilepticus or triphasic encephalopathy.
Epilepsia. 2006;47:2071–2074.
63. Kores B, Lader MH. Irreversible lithium neurotoxicity: an
overview. Clin Neuropharmacol. 1997;20:283–299.
64. Low MD. Evaluation of psychiatric disorders and the effects of
psychotherapeutic and psychomimetic agents. In: Klass DW, Daly
DD, eds. Current Practice of Clinical Electroencephalography.
1979:395–410.
65. Gardner CR, Tully RW, Hedgecock JR. The rapidly expanding
range of neuronal benzodiazepine receptor ligands. Prog
Neurobiol. 1993;40:1–61.
66. Doble A, Martin IL. Multiple benzodiazepine receptors: no reason
for anxiety. TiPS. 1992;13:76–81.
67. Niedermeyer E. Electroencephalographic studies on the anticon-
vulsive action of diazepam. Eur Neurol. 1970;3:88–96.
68. Mitsudone A, Ohfu M, Yasumoto S, et al. The effectiveness of clon-
azepam on the rolandic spikes. Brain Dev. 1997;19:274–278.
69. Riss J, Cloyd J, Gates J, et al. Benzodiazepines in epilepsy: pharma-
cology and pharmacokinetics. Acta Neurol Scand. 2008;118:69–86.
70. Prior PF, Maclaine GN, Scott DF, et al. Tonic status epilepticus pre-
cipitated by intravenous diazepam in a child with petit mal status.
Epilepsia. 1972;13:467–472.
71. Tassinari CA, Dravet C, Roger JA, et al. Tonic status epilepticus by
intravenous benzodiazepine in five patients with Lennox–Gastaut
syndrome. Epilepsia. 1972;13:421–435.
72. Treiman DM, Meyers PD, Walton NY, et al., for the Veterans Affairs
Status Epilepticus Cooperative Study Group. A comparison of four
treatments for generalized convulsive status epilepticus. N Engl J
Med. 1998;339:792–798.
73. Bixler EO, Kales A, Manfredi RL, et al. Next-day memory impair-
ment with triazolam use. Lancet. 1991;337:827–831.
917
74. Greenblatt DJ, Harmatz JS, Shapiro L, et al. Sensitivity to triazolam
in the elderly. N Engl J Med. 1991;324:1961–1698.
75. Bruni J, Wilder BJ. The toxicology of antiepileptic drugs. In:
Vinken PJ, Bruyn GW, eds. Handbook of Clinical Neurology. Vol 37.
Amsterdam: North Holland; 1979:199–222.
76. Mattson RH. Other antiepileptic drugs. Benzodiazepines. In:
Woodbury DM, Penry JK, Schmidt RP, eds. Antiepileptic Drugs.
New York: Raven Press; 1972:497–518.
77. Langtry HD, Benfield P. Zolpidem: a review of its pharmacody-
namic and pharmacokinetic properties and therapeutic potential.
Drugs. 1990;40:291–313.
78. Allain H, Delahaye CH, LeCoz F, et al. Postmarketing surveillance
of zopiclone in insomnia: analysis of 20513 cases. Sleep.
1991;14:408–413.
79. Parrino L, Boselli M, Spaggiari MC, et al. Multidrug comparison
(lorazepam, triazolam, zolpidem, and zopiclone) in situational
insomnia: polysomnographic analysis by means of the cyclic alter-
nating pattern. Clin Neuropharmacol. 1997;20:253–263.
80. Misgeld U, Bijak M, Jarolimek W. A physiological role for GABA-
B receptors and the effects of baclofen in the mammalian central
nervous system. Prog Neurobiol. 1995;46:423–462.
81. Kofler M, Kronenberg MF, Rifici C, et al. Epileptic seizures associated
with intrathecal baclofen application. Neurology. 1994;44:25–27.
82. Schuele SU, Kellinghaus C, Shook SJ, et al. Incidence of seizures in
patients with multiple sclerosis treated with intrathecal baclofen.
Neurology. 2005;63:1086–1087.
83. Fakhoury T, Abou-Khalil B, Blumenkopf B. EEG changes in
intrathecal baclofen overdose: a case report and review of the liter-
ature. EEG Clin Neurophysiol. 1998;107:339–342.
84. Chou I, Narasimhan K. Neurobiology of addiction. Nat Neurosci.
2005;8:1427.
85. Brown BB. Subjective and EEG responses to LSD in visualizer and
nonvisualizer subjects. Electroencephalogr Clin Neurophysiol.
1968;25:372–379.
86. Iversen L. Cannabis and the brain. Brain. 2003;126:1252–1270.
87. Messinis L, Kyprianidou A, Malefaki S, et al. Neuropsychological
deficits in long-term frequent cannabis users. Neurology.
2006:737–739.
88. Nordentoft M, Hjorthoj C. Cannabis use and risk of psychosis in
later life. Lancet. 2007;370:293–294.
89. Karst M, Salim K, Burstein S, et al. Analgesic effect of the synthetic
cannabinoid CT-3 on chronic neuropathic pain. JAMA.
2003;290:1757–1762.
90. Rodin EA, Domino EF, Porzak JP. The marihuana-induced “social
high.” Neurological and electroencephalographic concomitants.
JAMA. 1970;213:1800.
91. Herning RI, Better W, Cadet JL. EEG in chronic marijuana users dur-
ing abstinence: relationship to years of marijuana use, cerebral blood
flow and thyroid function. Clin Neurophysiol. 2008;119:321–331.
92. Nausieda PA. Central stimulant toxicity. In: Vinken PJ, Bruyn GW,
eds. Handbook of Clinical Neurology. Vol 37. Amsterdam: North
Holland; 1979:223–297.
93. Ricaurte GA, McCann UD. Recognition and management of compli-
cations of new recreational drug use. Lancet. 2005;365:2137–2145.
94. US Modafinil in Narcolepsy Multicenter Study Group.
Randomized trial of modafinil as a treatment for the excessive day-
time somnolence of narcolepsy. Neurology. 2000;54:1166–1175.
95. Clarke AR, Barry RJ, McCarty R, et al. Effects of stimulant medica-
tions on the EEG of girls with attention-deficit/hyperactivity dis-
order. Clin Neurophysiol. 2007;118:2700–2708.
96. Allen AJ, Kurlan RM, Gilbert DL, et al. Atomoxetine treatment in
children and adolescents with ADHD and comorbid tic disorders.
Neurology. 2005;65:1941–1949.
918 Part VI ■ EEG: Neuropharmacology and Anesthesia
97. Office of Applied Studies. Preliminary Results from the 1996
National Household Survey on Drug Abuse. Substance Abuse and
Mental Health Administration; 1996.
98. Schifano F, Oyefeso A, Webb L, et al. Review of deaths related to
taking ecstasy, England and Wales, 1997–2000. Br Med J.
2003;326:80–81.
99. Ernst Th, Chang L, Leonido-Yee M, et al. Evidence for long-term
neurotoxicity associated with methamphetamine abuse.
Neurology. 2000;54:1344–1349.
100. Leshner AI. Molecular mechanisms of cocaine addiction. N Engl
J Med. 1996;335:128–129.
101. Lüscher Ch, Bellone C. Cocaine-evoked synaptic plasticity: a key
to addiction? Nat Neurosci. 2008;11:737–738.
102. Noble EP. D2 dopamine receptor gene in psychiatric and neuro-
logic disorders and its phenotypes. Am J Med Genet.
2003;116B:103–125.
103. Herning RI, Jones RT, Hooker WD, et al. Cocaine increases EEG
beta: a replication and extension of Hans Berger’s historic
experiments. Electroencephalogr Clin Neurophysiol. 1985;60:
470–477.
104. Caplan LR, Heir DB, Banks G. Stroke and drug abuse. Stroke.
1982;13:869–872.
105. Globe LI, Merkin M. Cerebral infarction in a user of free-base
cocaine (“crack”). Neurology. 1986;36:1602–1604.
106. Mody CK, Miller BL, McIntyre HB, et al. Neurologic complica-
tions of cocaine abuse. Neurology. 1988;38:1189–1193.
107. Qureshi AI, Akbar MS, Czander E, et al. Crack cocaine use and
stroke in young patients. Neurology. 1997;48:341–345.
108. Van Stavern GP, Gorman M. Orbital infarction after cocaine use.
Neurology. 2002;59:642–643.
109. Aggarwal SK, Williams V, Levine SR, et al. Cocaine-associated
intracranial hemorrhage: absence of vasculitis in 14 cases.
Neurology. 1996;46:1741–1743.
110. Nolte KB, Brass LM, Fletterick CF. Intracranial hemorrhage asso-
ciated with cocaine abuse: a prospective autopsy study.
Neurology. 1996;46:1291–1296.
111. Conway JE, Tamergo RJ. Cocaine use is an independent risk fac-
tor for cerebral vasospasm after aneurysmal subarachnois hem-
orrhage. Stroke. 2001;32:2338–2343.
112. Krendel DA, Ditter SM, Frankel MR, et al. Biopsy-proven cerebral
vasculitis associated with cocaine abuse. Neurology.
1990;40:1092–1094.
113. Weiner WJ, Rabinstein A, Levin B, et al. Cocaine-induced persist-
ent dyskinesias. Neurology. 2001;56:964–965.
114. Migita DS, Devereaux MW, Tomsak RL. Cocaine and pupillary-
sparing oculomotor nerve palsies. Neurology. 1997;49:
1466–1467.
115. Ogunyemi AO, Locke GE, Kramer LD, et al. Complex partial sta-
tus epilepticus provoked by “crack” cocaine. Ann Neurol.
1989;26:785–786.
116. Levy RH, Mattson RH, Meldrum BS, et al., eds. Antiepileptic
Drugs. 5th ed. Philadelphia: Lippincott Williams & Wilkins;
2002.
117. Herkes GK, Lagerlund TD, Sharbrough FW, et al. Effects of
antiepileptic drug treatment on the background frequency of
EEGs in epileptic patients. J Clin Neurophysiol. 1993;10:210–216.
118. Salinsky MC, Oken BS, Morehead L. Intraindividual analysis of
antiepileptic drug effects on EEG background rhythms.
Electroencephalogr Clin Neurophysiol. 1994;90:186–193.
119. Salinsky MC, Oken BS, Storzbach D, et al. Assessment of CNS
effects of antiepileptic drugs by using quantitative EEG measures.
Epilepsia. 2003;44:1042–1050.
120. Duncan JS. Antiepileptic drugs and the electroencephalogram.
Epilepsia. 1987;28:259–266.
121. Kuruvilla Th, Bharucha NE. Cerebellar atrophy after acute
phenytoin intoxication. Epilepsia. 1997;38:500–502.
122. Luef G, Burtscher J, Kremser Ch, et al. Magnetic resonance vol-
umetry of the cerebellum in epileptic patients after phenytoin
overdosage. Eur Neurol. 1996;36:273–277.
123. Bauer J. Seizure-inducing effects of antiepileptic drugs. Acta
Neurol Scand. 1996;4:367–377.
124. Perucca E, Gram L, Avanzini G, et al. Antiepileptic drugs as a
cause of worsening seizures. Epilepsia. 1998;39:5–17.
125. Roseman E. Dilantin toxicity: a clinical and electroencephalo-
graphic study. Neurology (Minneapolis). 1961;11:912–921.
126. Wilkus RJ, Dodrill CB, Troupin AS. Carbamazepine and the elec-
troencephalogram of epileptics: a double-blind study in compar-
ison to phenytoin. Epilepsia. 1978;25:467–475.
127. Carrie JRG. Computer-assisted EEG sharp transient detection
and quantification during overnight recordings in an epileptic
patient. In: Kellaway P, Petersen I, eds. Quantitative Analytic
Studies in Epilepsy. New York: Raven Press; 1976:225–235.
128. Buchthal F, Svensmark O, Schiller PJ. Clinical and electroen-
cephalographic correlations with serum levels of diphenylhydan-
toin. Arch Neurol. 1960;2:624–630.
129. Spencer SS, Spencer DD, Williamson PD, et al. Ictal effects of
anticonvulsant medication withdrawal in epileptic patients.
Epilepsia. 1981;22:297–307.
130. Ambrosetto C, Tassinari CA, Baruzzi A, et al. Phenytoin
encephalopathy as probable idiosyncratic reaction: case report.
Epilepsia. 1977;18:405–408.
131. Besser R, Hornung K, Theison M, et al. EEG changes in patients
during the introduction of carbamazepine. Electroencephalogr
Clin Neurophysiol. 1992;83:19–23.
132. Saunders MG, Westmoreland BF. The EEG in the evaluation of
disorders affecting the brain diffusely. In: Klass DW, Daly DD,
eds. Current Practice of Clinical Electroencephalography. New
York: Raven Press; 1979:343–379.
133. Dong X, Leppik IE, White J, et al. Hyponatremia from oxcar-
bazepine and carbamazepine. Neurology. 2005;65:1976–1978.
134. Kochen S, Giagante B, Oddo S. Spike-and-wave complexes and
seizure exacerbation caused by carbamazepine. Eur J Neurol.
2002;9:41–47.
135. Prasad AN, Stefanelli M, Nagarajan L. Seizure exacerbation and
developmental regression with carbamazepine. Can J Neurol Sci.
1998;25:287–294.
136. Talwar D, Arora MS, Sher PK. EEG changes and seizure exacerba-
tion in young children treated with carbamazepine. Epilepsia.
1994;35:1154–1159.
137. De Rubeis AA, Young GB. Continuous EEG monitoring in a
patient with massive carbamazepine overdose. J Clin
Neurophysiol. 2001;18:166–168.
138. Meador KJ, Loring DW, Huh K, et al. Comparative cognitive
effects of anticonvulsants. Neurology (Minneapolis).
1990;40:391–394.
139. Dodrill CB, Troupin AS. Neuropsychological effects of carba-
mazepine and phenytoin: a reanalysis. Neurology. 1991;41:
141–143.
140. Leslie PH, Heyworth R, Prescott LF. Cardiac complications of
carbamazepine intoxication: treatment by hemoperfusion. Br J
Med. 1983;286:1018.
141. Grant SM, Faulds D. Oxcarbazepine. A review of its pharmacol-
ogy and therapeutic potential in epilepsy, trigeminal neuralgia
and affective disorders. Drugs. 1992;43:873–888.
 Chapter 43 ■ EEG, Drug Effects, and Central Nervous System Poisoning
142. Gelisse P, Genton P, Kuate C, et al. Worsening of seizures by
oxcarbazepine in juvenile generalized epilepsies. Epilepsia.
2004;45:1282–1286.
143. Venrdrame M, Khurana DS, Cruz M, et al. Aggravation of
seizures and/or EEG features in children treated with oxcar-
bazepine monotherapy. Epilepsia. 2007;48:2116–2120.
144. Jones AL, Proudfoot AT. Features and management of poisoning
with modern drugs used to treat epilepsy. Q J Med. 1998;91:
325–332.
145. Clemens B. Valproate decreases EEG synchronization in a use-
dependent manner in idiopathic generalized epilepsy. Seizure.
2008;17:224–233.
146. Adams DJ, Lüders H, Pippenger Ch. Sodium valproate in
the treatment of intractable seizure disorders: a clinical and
electroencephalographic study. Neurology (Minneapolis).
1978;28:152–157.
147. Braathen G, Theorell K, Persson A, et al. Valproate in the treat-
ment of absence epilepsy in children: a study of dose–response
relationship. Epilepsia. 1988;29:548–552.
148. Bruni J, Wilder BJ, Baumann AW, et al. Clinical efficacy and long-
term effects of valproic acid therapy on spike-and-wave dis-
charges. Neurology. 1980;30:42–46.
149. Sato S, White BG, Penry JK, et al. Valproic acid versus ethosux-
imide in the treatment of absence seizures. Neurology.
1982;32:157–163.
150. Harding GFA, Herrick CE, Jeavons PM. A controlled study of the
effect of sodium valproate on photosensitive epilepsy and its
prognosis. Epilepsia. 1978;19:555–565.
151. Simon D, Penry JK. Sodium-di-N-propylacetate (DPA) in the
treatment of epilepsy. A review. Epilepsia. 1975;16:549–573.
152. Dreifuss FE. Valproic acid. Toxicity. In: Levy RH, Mattson RH,
Meldrum BS, eds. Antiepileptic Drugs. 4th ed. Philadelphia:
Lippincott-Raven; 1995:641–648.
153. Segura-Bruna N, Rodriguez-Campello A, Puente V, et al.
Valproate-induced hyperammonemic encephalopathy. Acta
Neurol Scand. 2006;114:1–7.
154. Kifune A, Kubota F, Shibata N, et al. Valproic acid-induced hyper-
ammonemic encephalopathy with triphasic waves. Epilepsia.
2000;41:909–912.
155. Dreifuss FE, Santili N, Langer DH, et al. Valproic acid fatalities: a
retrospective review. Neurology. 1987;37:379–385.
156. Koenig SA, Buessing D, Longin E, et al. Valproic acid-induced
hepatopathy: nine new fatalities in Germany from 1994 to 2003.
Epilepsia. 2006;47:2027–2031.
157. Scheffner D, König S, Rauterberg-Ruland I, et al. Fatal liver fail-
ure in 16 children with valproate therapy. Epilepsia. 1988;29:
530–542.
158. Capocchi G, Balducci A, Cecconi M, et al. Valproate-induced
epileptic tonic status. Seizure. 1998;7:237–241.
159. Pindar RM, Brogden RN, Speight TM, et al. Clonazepam: a
review of its pharmacological properties and therapeutic efficacy
in epilepsy. Drugs. 1976;12:321–361.
160. Petsche H. Zum Nachweis des kortikalen Angriffspunktes des
antikonvulsiven Benzodiazepinderivats Clonazepam (Ro 5–
4023). Z EEG-EMG. 1972;3:145.
161. Martin D, Hirt HR. Clinical experience with clonazepam
(Rivotril) in the treatment of epilepsies in infancy and childhood.
Neuropediatrie. 1973;4:245–266.
162. Grant SM, Heel RC. Vigabatrin. A review of its pharmacody-
namic and pharmacokinetic properties, and therapeutic poten-
tial in epilepsy and disorders of motor control. Drugs.
1991;41:889–926.
919
163. Marciani MG, Stantione P, Maschio M, et al. EEG changes
induced by vigabatrin monotherapy in focal epilepsy. Acta Neurol
Scand. 1997;95:115–120.
164. Hammond EJ, Wilder BJ. Effects of gamma-vinyl-GABA on the
human electroencephalogram. Neuropharmacology. 1985;24:
975–984.
165. Kälviäinen R, Keränen T, Riekkinen PJ. Place of newer
antiepileptic drugs in the treatment of epilepsy. Drugs.
1993;46:1009–1024.
166. Mervaala E, Partanen J, Nousiainen U, et al. Electrophysiologic
effects of gamma-vinyl GABA and carbamazepine. Epilepsia.
1989;30:189–193.
167. Ben-Menachem E, Treiman DM. Effect of gamma-vinyl GABA
on interictal spikes and sharp waves in patients with intractable
complex partial seizures. Epilepsia. 1989;30:79–83.
168. Luna D, Dulac O, Pajot N, et al. Vigabatrin in the treatment of
childhood epilepsies: a single-blind placebo-controlled study.
Epilepsia. 1989;30:430–437.
169. Marciani MG, Maschio M, Spanedda F, et al. Development of
myoclonus in patients with partial epilepsy during treatment
with vigabatrin: an electroencephalographic study. Acta Neurol
Scand. 1995;91:1–5.
170. Michelucci R, Tassinari CA. Response to vigabatrin in relation to
seizure type. Br J Clin Pharmacol. 1989;27:119–124.
171. Sälke-Kellermann A, Baier H, Rambeck B, et al. Acute
encephalopathy with vigabatrin. Lancet. 1993;342:185 [letter].
172. Sharief MK, Sander JWA, Shorvon SD. Acute encephalopathy
with vigabatrin. Lancet. 1993;342:619 [letter].
173. Sander JW, Hart YM, Trimble MR, et al. Vigabatrin and psy-
chosis. J Neurol Neurosurg Psychiatry. 1991;54:435–439.
174. Malmgren K, Ben-Menachem E, Frisén L. Vigabatrin and visual
toxicity: evolution and dose dependence. Epilepsia. 2001;42:
609–615.
175. Camposano SE, Major P, Halpern E, et al. Vigabatrin in the treat-
ment of childhood epilepsy: a retrospective chart review of effi-
cacy and safety profile. Epilepsia. 2008;49:1186–1191.
176. Conway M, Cubbidge RP, Hosking SL. Visual field severity
indices demonstrate dose-dependent visual loss from vigabatrin
therapy. Epilepsia. 2008;49:108–116.
177. Fitton A, Goa KL. Lamotrigine. An update of its pharmacology
and therapeutic use in epilepsy. Drugs. 1995;50:691–713.
178. Mervaala E, Koivisto K, Hänninen T. Electrophysiological and
neuropsychological profiles of lamotrigine in young and age-
associated memory impairment (AAMI) subjects. Neurology.
1995;46(suppl 4):S259.
179. Foletti G, Volanschi D. Influence of lamotrigine addition on
computerized background EEG parameters in severe epilepto-
genic encephalopathies. Eur Neurol. 1994;34(suppl 1):S87–S89.
180. Besag FM. Use of the “monolog” spike wave monitor to evaluate
lamotrigine for absence seizures. Epilepsia. 1991;32(suppl
1):S89–S90 [abstract].
181. Binnie CD, van Emde-Boas W, Kasteleijn-Nolste-Trenite DG.
Acute effects of lamotrigine (BW 430C) in persons with epilepsy.
Epilepsia. 1986;27:248–254.
182. Erikson AS, Knudson E, Nergardth A. The effect of lamotrigine
on epileptiform discharges in young patients with drug-resistant
epilepsy. Epilepsia. 2001;42:230–236.
183. Biraben A, Allain H, Scarabin JM, et al. Exacerbation of juvenile
myoclonic epilepsy with lamotrigine. Neurology. 2000;55:1758.
184. Guerrini R, Dravet C, Genton P, et al. Lamotrigine and seizure
aggravation in severe myoclonic epilepsy. Epilepsia. 1998;39:
508–512.
920 Part VI ■ EEG: Neuropharmacology and Anesthesia
185. Dinnerstein E, Jobst BC, Williamson PD. Lamotrigine intoxica-
tion provoking status epilepticus in an adult with localization-
related epilepsy. Arch Neurol. 2007;64:1344–1346.
186. Trinka E, Dilitz E, Unterberger I, et al. Non convulsive status
epilepticus after replacement of valproate with lamotrigine. J
Neurol. 2002;249:1417–1422.
187. Buckley NA, Whyte IM, Dwason AH. Self-poisoning with lamot-
rigine. Lancet. 1993;342:1552–1553 [letter].
188. Taylor CP. Gabapentin. Mechanisms of action. In: Levy RH,
Mattson RH, Meldrum BS, et al., eds. Antiepileptic Drugs. 5th ed.
Philadelphia: Lippincott Williams & Wilkins; 2002:321–339.
189. Mathew NT, Rapoport A, Saper J, et al. Efficacy of gabapentin in
migraine prophylaxis. Headache. 2001;41:119–128.
190. Spira PJ, Beran RG. Gabapentin in the prophylaxis of chronic
daily headache. Neurology. 2003;61:1753–1759.
191. Backonja M, Beydoun A, Edwards KR, et al. Gabapentin for the
symptomatic treatment of painful neuropathy in patients with
diabetes mellitus. JAMA. 1998;280:1831–1836.
192. Greist JH. Gabapentin. Clinical efficacy and use in psychiatric
disorders. In: Levy RH, Mattson RH, Meldrum BS, et al., eds.
Antiepileptic Drugs. 5th ed. Philadelphia: Lippincott Williams &
Wilkins; 2002:349–353.
193. Ramsey RE, Pryor FM. Gabapentin. Toxicity. In: Levy RH,
Mattson RH, Meldrum BS, et al., eds. Antiepileptic Drugs. 5th ed.
Philadelphia: Lippincott Williams & Wilkins; 2002:354–359.
194. Fischer JH, Barr AN, Rogers SL, et al. Lack of serious toxicity fol-
lowing gabapentin overdose. Neurology. 1994;44:982–983.
195. Salinsky MC, Binder LM, Oken BS, et al. Effects of gabapentin
and carbamazepine on the EEG and cognition in healthy volun-
teers. Epilepsia. 2002;43:482–490.
196. Foldvary-Schaefer N, De Leon Sanchez I, Karafa M, et al.
Gabapentin increases slow-wave sleep in normal adults.
Epilepsia. 2002;43:1493–1497.
197. Taylor CP, Angelotti T, Fauman E. Pharmacology and mechanism
of action of pregabalin: the calcium channel 2– (alpha2–delta)
subunit as a target for antiepileptic drug discovery. Epilepsy Res.
2007;73:137–150.
198. Beydoun A, Uthman BM, Kugler AR, et al. Safety and efficacy of
two pregabalin regimens for add-on treatment of partial epilepsy.
Neurology. 2005;64:475–480.
199. Lesser H, Sharma U, LaMoreaux L, et al. Pregabalin relieves
symptoms of painful diabetic neuropathy. Neurology.
2004;63:2104–2110.
200. Asconape JJ, Hartman LM, Salanova V. Pregabalin-associated
myoclonus. Epilepsia. 1999;40(suppl 7):143.
201. Huppertz HJ, Feuerstein TJ, Schulze-Bonhage A. Myoclonus in
epilepsy patients with anticonvulsive add-on therapy with prega-
balin. Epilepsia. 2001;42:790–792.
202. Knake S, Klein KM, Hattemer K, et al. Pregabalin-induced gener-
alized myoclonic status epilepticus in patients with chronic pain.
Epilepsy Behav. 2007;11:471–473.
203. Oaklander AL, Buchbinder BR. Pregabalin-withdrawal
encephalopathy and splenial edema: a link to high-altitude ill-
ness? Ann Neurol. 2005;58:309–312.
204. Felbamate Study Group in the Lennox–Gastaut Syndrome.
Efficacy of felbamate in childhood epileptic encephalopathy
(Lennox–Gastaut syndrome). N Engl J Med. 1993;328:29–33.
205. Pellock JM, Faught E, Leppik IE, et al. Felbamate: consensus of
current clinical practice. Epilepsy Res. 2006;71:89–101.
206. Adkins JC, Noble St. Tiagabine. A review of its pharmacody-
namic and pharmacokinetic properties and therapeutic potential
in the management of epilepsy. Drugs. 1998;55:437–460.
207. Laech JP, Stolarek I, Brodie MJ. Deliberated overdose with the
novel anticonvulsant tiagabine. Seizure. 1995;4:155–157.
208. Coenen AML, Blezer EHM, van Luijtelaar ELJM. Effects of the
GABA-uptake inhibitor tiagabine on electroencephalogram,
spike-wave discharges and behavior of rats. Epilepsy Res.
1995;21:89–94.
209. Walton NY, Gunawan S, Treiman DM. Treatment of experimen-
tal status epilepticus with the GABA uptake inhibitor, tiagabine.
Epilepsy Res. 1994;19:237–244.
210. Koepp MJ, Edwards M, Collins J, et al. Status epilepticus and
tiagabine revisited. Epilepsia. 2005;46:1625–1632.
211. Cross JH. Topiramate. In: Shorvon S, Perucca E, Fish D, et al., eds.
The Treatment of Epilepsy. 2nd ed. Oxford: Blackwell; 2003:
535–547.
212. Meador KJ, Loring DW, Hulihan JF, et al. Differential cognitive
and behavioral effects of topiramate and valproate. Neurology.
2003;60:1483–1488.
213. Blum D, Meador K, Biton V, et al. Cognitive effects of lamotrigine
compared with topiramate in patients with epilepsy. Neurology.
2006;67:400–406.
214. Gross-Tour V, Shalev RS. Reversible language regression as an
adverse effect of topiramate treatment in children. Neurology.
2004;62:299–300.
215. Mula M, Trimble MR, Thompson P, et al. Topiramate and word-
finding difficulties in patients with epilepsy. Neurology.
2003;60:1104–1107.
216. Bonanni E, Galli R, Maestri M, et al. Daytime sleepiness in
epilepsy patients receiving topiramate monotherapy. Epilepsia.
2004;45:333–337.
217. Dooley M, Plosker GL. Levetiracetam. A review of its adjunctive
use in the management of partial seizures. Drugs. 2000;60:
871–893.
218. Sharpe DV, Patel AD, Abou-Khalil B, et al. Levetiracetam
monotherapy in juvenile myoclonic epilepsy. Seizure.
2008;17:64–68.
219. Specchio N, Boero G, Michelucci R, et al. Effects of levetirazetam
on EEG abnormalities in juvenile myoclonic epilepsy. Epilepsia.
2008;49:663–669.
220. Striano P, Sofia V, Capovilla G, et al. A pilot trial of levetiracetam
in eyelid myoclonia with absences (Jeavons syndrome). Epilepsia.
2008;49:425–430.
221. Sohn YH, Kaelin-Lang A, Jung HY, et al. Effect of levetiracetam
on human corticospinal excitability. Neurology. 2001;57:858–863.
222. Epstein CM, Girad-Siqueira L, Ehrenberg JA. Prolonged neuro-
physiologic effects of levetiracetam after oral administration in
humans. Epilepsia. 2008;49:1169–1173.
223. French J, di Nicola S, Arrigo C. Fast and sustained efficacy of lev-
etiracetam during titration and the first 3 months of treatment in
refractory epilepsy. Epilepsia. 2005;46:1304–1307.
224. Stodieck S, Steinhoff BJ, Kolmsee S, et al. Effect of levetiracetam
in patients with epilepsy and interictal epileptiform discharges.
Seizure. 2001;10:583–587.
225. Bazil CW, Battista J, Basner RC. Effects of levetiracetam on sleep
in normal volunteers. Epilepsy Behav. 2005;7:539–542.
226. Mula M, Trimble MR, Sander JW. Are psychiatric adverse events
of antiepileptic drugs a unique entity? A study on topiramate and
levetiracetam. Epilepsia. 2007;48:2322–2326.
227. Bauer J. Encephalopathy induced by levetiracetam added to val-
proate. Acta Neurol Scand. 2008;117:374–376.
228. Macdonald RL. Zonisamide. Mechanism of action. In: Levy RH,
Mattson RH, Meldrum BS, et al., eds. Antiepileptic Drugs. 5th ed.
Philadelphia: Lippincott Williams & Wilkins; 2002:867–872.
 Chapter 43 ■ EEG, Drug Effects, and Central Nervous System Poisoning
229. Leppik IE, Willmore LJ, Homan RW, et al. Efficacy and safety of
zonisamide: results of a multicenter study. Epilepsy Res.
1993;14:165–173.
230. Seino M, Fujitani B. Zonisamide. Clinical efficacy and use in
epilepsy. In: Levy RH, Mattson RH, Meldrum BS, et al., eds.
Antiepileptic Drugs. 5th ed. Philadelphia: Lippincott Williams &
Wilkins; 2002:885–891.
231. Baulac M, Leppik IE. Efficacy and safety of adjunctive zon-
isamide therapy for refractory partial seizures. Epilepsy Res.
2007;75:75–83.
232. Joo EY, Kim SH, Seo DW et al. Zonisamide decreases cortical
excitability in patients with idiopathic generalized epilepsy. Clin
Neurophysiol. 2008;119:1385–1392.
233. Szaflarski JP. Effects of zonisamide on the electroencephalogram
of a patient with juvenile myoclonic epilepsy. Epilepsy Behav.
2004;5:1024–1026.
234. Beyreuther BK, Freitag J, Heers C, et al. Lacosamide: a review of
preclinical properties. CNS Drug Rev. 2007;13:21–42.
235. Ben-Menachem E, Biton V, Jatuzis D, et al. Efficacy and safety of
oral lacosamide as adjunctive therapy in adults with partial-onset
seizures. Epilepsia. 2007;48:1308–1317.
236. Young GB, Da Silva OP. Effects of morphine on the electroen-
cephalograms of neonates: a prospective, observational study.
Clin Neurophysiol. 2000;111:1955–1960.
237. Kearse LA, Koski G, Husain MV, et al. Epileptiform activity
during opioid anesthesia. EEG Clin Neurophysiol. 1993;87:
374–379.
238. Jakobs MH, Senior RM. Theophylline toxicity due to impaired
theophylline degradation. Am Rev Respir Dis. 1974;110:342–345.
239. Schwartz KA, Scott DF. Aminophylline-induced seizures.
Epilepsia. 1974;15:501–505.
240. Yarnell PR, Chu NS. Focal seizures and aminophylline. Neurology
(Minneapolis). 1975;25:819–822.
241. Yoshikawa H. First-line therapy for theophylline-associated
seizures. Acta Neurol Scand. 2007;115:57–61.
242. Katz BE, Carver MW. Acute poisoning with isoniazid treated with
exchange transfusions. Pediatrics. 1956;18:77.
243. Mellerio F. L’électroencéphalographie Dans Les Intoxications
Aigues. Paris: Masson; 1964.
244. Terman D, Teitelbaum DT. Isoniazid self-poisoning. Neurology
(Minneapolis). 1970;20:299–304.
245. Weinstein L. Penicillins and cephalosporins. In: Goodman LS,
Gilman A, eds. The Pharmacological Basis of Therapeutics. New
York: Macmillan; 1975:1130–1166.
246. Fernández-Torre JL, Martínez-Martínez M, González-Rato J, et
al. Cephalosporin-induced nonconvulsive status epilepticus:
clinical and electroencephalographic features. Epilepsia.
2005;46:1550–1552.
247. Maganti R, Jolin D, Rishi D, et al. Nonconvulsive status epilepti-
cus due to cefepime in a patient with normal renal function.
Epilepsy Behav. 2006;8:312–314.
248. Jallon P, Fankhauser L, Du Pasquier R, et al. Severe but reversible
encephalopathy associated with cefepime. Neurophysiol Clin.
2000;30:383–386.
249. Serkova NJ, Christians U, Benet LZ. Biochemical mechanisms of
cyclosporin neurotoxicity. Mol Intervent. 2004;4:97–107.
250. Gleeson JG. Cyclosporin A acute encephalopathy and seizure
syndrome in childhood: clinical features and risk of seizure
recurrence. J Child Neurol. 1998;13:336–344.
251. Singh G, Ress JH, Sander JW. Seizures and epilepsy in oncological
practice: causes, course, mechanisms and treatment. J Neurol
Neurosurg Psychiatry. 2007;78:342–349.
921
252. Kilickap S, Cakar M, Onal IK, et al. Nonconvulsive status epilep-
ticus due to ifosfamide. Ann Pharmacother. 2006;40:332–335.
253. Wengs WJ, Talwar D, Bernard J. Ifosfamide-induced nonconvul-
sive status epilepticus. Arch Neurol. 1993;50:1104–1105.
254.Simonian NA, Gilliam FG, Chiappa KH. Ifosfamide causes a
diazepam-sensitive encephalopathy. Neurology. 1993;43:2700–2702.
255. Meyer Th, Ludolph AC, Münch C. Ifosfamide encephalopathy
presenting with asterixis. J Neurol Sci. 2002;1999:85–88.
256. Brown GL, Wilson WP. Salicylate intoxication and the CNS with
special reference to EEG findings. Dis Nerv Syst. 1971;32:135–140.
257. Münthe ThF, Knehans A. EEG-Befunde bei rezidivierender
Salizylat-Intoxication. Z EEG-EMG. 1997;28:52–54.
258. Shiozawa Z, Sasaki H, Ozaki Y, et al. Epilepsia partialis continua fol-
lowing metrizamide cisternography. Ann Neurol. 1981;10:400–401.
259. Elian M, Fenwick P. Metrizamide and the EEG: three case reports
and a review. J Neurol. 1985;232:341–345.
260. Obeid T, Yaqub B, Panayiotopoulos Ch, et al. Absence status
epilepticus with computed tomographic brain changes following
metrizamide myelography. Ann Neurol. 1988;24:582–584.
261. Russel D, Anke IM, Nyberg-Hansen R, et al. Complex partial sta-
tus epilepticus following myelography with metrizamide. Ann
Neurol. 1980;8:325–327.
262. Vollmer ME, Weiss H, Beanland C, et al. Prolonged confusion
due to absence status following metrizamide myelography. Arch
Neurol. 1985;42:1005–1008.
263. Ropper AH, Chiappa KH, Young RR. The effect of metrizamide
on the electroencephalogram: a prospective study in 61 patients.
Ann Neurol. 1979;6:222–226.
264. Klein KM, Shiratori K, Knake S, et al. Status epilepticus and
seizures induced by iopamidol myelography. Seizure.
2004;13:196–199.
265. Nelson M, Bartlett RJV, Lamb JT. Seizures after intravenous con-
trast media for cranial computed tomography. J Neurol
Neurosurg Psychiatry. 1989;52:1170–1175.
266. Maramatton BV, Manno EM, Wijdicks EFM, et al. Gadolinium
encephalopathy in a patient with renal failure. Neurology.
2005;64:1276–1278.
267. Neufeld MY. Periodic triphasic waves in levodopa-induced
encephalopathy. Neurology. 1992;42:444.
268. Riazi K, Roshanpour M, Rafiei-Tabatabaei N, et al. The procon-
vulsant effect of sildafenil in mice: role of nitric oxide–cGMO
pathway. Br J Pharmacol. 2006;147:935–943.
269. Gilad R, Lampl Y, Eshel Y, et al. Drug points: tonic–clonic seizures
in patients taking sildafenil. Br Med J. 2002;325:869.
270. Striano P, Zara F, MinettiC, et al. Epileptic seizures can follow
high doses of oral vardenafil. Br Med J. 2006;333:785.
271. Mellerio F, Kubicki S. Encephalopathy due to poisoning. In: Remond
A, ed.-in-chief. Handbook of Electroencephalography and Clinical
Neurophysiology. Vol 15A. Amsterdam: Elsevier; 1977:108–135.
272. Lidsky TI, Schneider JS. Lead neurotoxicity in children: basic
mechanisms and clinical correlates. Brain. 2003;126:5–19.
273. Canfield RL, Henderson CR, Cory-Slechta DA, et al. Intellectual
impairment in children with blood lead concentrations below
10 g per deciliter. N Engl J Med. 2003;348:1517–1526.
274. Stewart WF, Schwartz BS, Davatzikos C, et al. Past adult lead
exposure is linked to neurodegeneration measured by brain MRI.
Neurology. 2006;66:1474–1484.
275. Fejerman N, Gimenez ER, Vallejo NE, et al. Lennox–Gastaut syn-
drome and lead intoxication. Pediatrics. 1973;52:227–234.
276. Benignus VA, Otto DA, Müller KE, et al. Effects of age and body
lead burden on CNS functions is young children. II. EEG spectra.
Electroencephalogr Clin Neurophysiol. 1981;52:240–248.
922 Part VI ■ EEG: Neuropharmacology and Anesthesia
277. Mahaffey K. Methylmercury: a new look at the risk. Public Health
Rep. 1999;114:397–413.
278. Myers GJ, Davidson PW, Cox C, et al. Prenatal methylmercury
exposure from ocean fish consumption in the Seychelles child
development study. Lancet. 2003;361:1686–1692.
279. Saper RB, Phillips RS, Sehgal A, et al. Lead, mercury, an arsenic in
US- and Indian-manufactured ayurvedic medicines sold via
internet. J Am Med Assoc. 2008;300;915–923.
280. Brenner RP, Snijder RD. Late EEG-findings and clinical status
after organic mercury poisoning. Arch Neurol. 1980;37:282–284.
281. Pirkivi L, Tolonen U. EEG findings in chlor-alkali workers sub-
jected to low long term exposure to mercury vapour. Br J Ind
Med. 1989;46:370–375.
282. Lee JW. Manganese intoxication. Arch Neurol. 2000;57:597–599.
283. Ono K, Komai K, Yamada M. Myoclonic involuntary movement
associated with chronic manganese poisoning. J Neurol Sci.
2002;199:93–96.
284. Alfrey AC, Le Gendre GR, Kachny WD. The dialysis encephalopa-
thy syndrome. Possible aluminium intoxication. N Engl J Med.
1976;294:184–188.
285. Prick JJG. Thallium poisoning. In: Vinken PJ, Bruyn GW, eds.
Handbook of Neurology. Vol 36. Amsterdam: North Holland;
1979:239–278.
286. Prockop LD, Chichkova RI. Carbon monoxide intoxication: an
updated review. J Neurol Sci. 2007;262:122–130.
287. Kim J, Chang K, Song IC, et al. Delayed encephalopathy of acute
carbon monoxide intoxication: diffusivity of cerebral white mat-
ter lesions. Am J Neuroradiol. 2003;24:1592–1597.
288. Zagami AS, Lethlan AK, Mellick R. Delayed neurological deterio-
ration following carbon monoxide poisoning: MRI findings. J
Neurol. 1993;240:113–116.
289. Bokonjic N. Stagnant Anoxia and Carbon Monoxide Poisoning.
Amsterdam: Elsevier; 1963.
290. Murata M, Suzuki M, Hasegawa Y, et al. Improvement of occipital
alpha activity by repetitive hyperbaric oxygen therapy in patients
with carbon monoxide poisoning: a possible indicator for treat-
ment efficacy. J Neurol Sci. 2005;235:69–74.
291. Schneck SA. Methyl alcohol. In: Vinken PJ, Bruyn GW, eds.
Handbook of Neurology. Vol 36. Amsterdam: North Holland;
1979:351–360.
292. Ley CO, Gali FG. Parkinsonian syndrome after methanol intoxi-
cation. Eur Neurol. 1983;22:405–409.
293. Jameson HD, Kane R. EEG records during an epidemic of
methanol intoxication. Electroencephalogr Clin Neurophysiol.
1969;26:112 [abstract].
294. Hanif M, Mobarak MR, Ronan A, et al. Fatal renal failure caused
by diethylene glycol in paracetamol elixir: the Bangladesh epi-
demic. Br Med J. 1995;311:88–91.
295. O Brien KL, Selanikio JD, Hecdivert Ch, et al., for the Acute
Renal Failure Investigation Team. Epidemic of pediatric deaths
from acute renal failure caused by diethylene glycol poisoning.
JAMA. 1998;279:1175–1180.
296. Lewis LD, Smith BW, Mamourian AC. Delayed sequelae after
acute overdoses or poisonings: cranial neuropathy related to
ethylene glycol ingestion. Clin Pharmacol Ther. 1997;61:692–699.
297. Rollins YD, Filley CM, McNutt JT, et al. Fulminant ascending
paralysis as a delayed sequela of diethylene glycol (Sterno) inges-
tion. Neurology. 2002;59:1460–1463.
298. Zhou L, Zabad R, Lewis RA. Ethylene glycol intoxication: electro-
physiological studies suggest a polyradiculopathy. Neurology.
2002;59:1809–1810.
299. Chan AWK. Alcoholism and epilepsy. Epilepsia. 1985;26:323–333.
300. Krauss GL, Niedermeyer E. Electroencephalogram and seizures
in chronic alcoholism. Electroencephalogr Clin Neurophysiol.
1991;78:97–104.
301. Davis KM, Wu JY. Role of glutaminergic and GABAergic systems
in alcoholism. J Biomed Sci. 2000;8:7–19.
302. Sand T, Brathen G, Michler R, et al. Clinical utility of EEG in
alcohol-related seizures. Acta Neurol Scand. 2002;105:18–24.
303. Leone M, Tonini C, Bogliun G, et al., for the ARES (Alcohol
Related Seizures) Study Group. Chronic alcohol use and first
symptomatic epileptic seizure. J Neurol Neurosurg Psychiatry.
2002;73:495–499.
304. Niedermeyer E, Freund G, Krumholz A. Subacute encephalopa-
thy with seizures in alcoholics: a clinical electroencephalographic
study. Clin Electroencephalogr. 1981;12:113–129.
305. Mani J, Sitajayalakshmi S, Borgohain R, et al. Subacute
encephalopathy with seizures in alcoholism. Seizure.
2003;12:126–129.
306. Ng SKC, Hauser WA, Brust YCM, et al. Alcohol consumption and
withdrawal in new-onset seizures. N Engl J Med. 1988;319:637–666.
307. Ashton CH. Solvent abuse. Br Med J. 1990;20:135–136.
308. Aydin K, Sencer S, Demir T, et al. Cranial MR findings in chronic
toluene abuse by inhalation. Am J Neuroradiol. 2002;23:1173–1179.
309. Baker EL. Chronic toxic encephalopathy caused by occupational
solvent exposure. Ann Neurol. 2008;63:545–547.
310. Gash DM, Rutland K, Hudson NL, et al. Trichloroethylene:
parkinsonism and complex 1 mitochondrial neurotoxicity. Ann
Neurol. 2007;63:184–192.
311. Stracciari A, Gallassi R, Ciardulli C, et al. Neuropsychological and
EEG evaluation in exposure to trichloroethylene. J Neurol.
1985;232:120–122.
312. Roberts DM, Aoron CK. Clinical review. Management of acute
organophosphorus pesticide poisoning. Br Med J. 2007;334:
629–634.
313. Lockwood AL. Editorial comment: organophosphate pesticides
and public policy. Curr Opin Neurol. 2002;15:725–726.
314. Dassanyake T, Weerasinghe V, Dangahadeniya U, et al. Long-term
event-related potential changes following organophosphorus
insecticide poisoning. Clin Neurophysiol. 2008;199:144–150.
315. Okonek S, Rieger H. EEG Veränderungen bei
Alkylphosphatvergiftungen. Z EEG-EMG. 1975;6:19–27.
316. Lockwood AL. Exposure to environmental toxins. Curr Opin
Neurol. 1997;10:507–511.
317. Walder B, Tramer MR, Seeck M. Seizure-like phenomena and
propofol. Neurology. 2002;58:1327–1332.
Part VII Evoked Potentials and Event-Related EEG Phenomena
Event-Related Potentials: General
Aspects of Methodology
and Quantification
E
lectroencephalography as a general method for the
investigation of human brain function includes ways of
determining the reactions of the brain to a variety of
stimuli. Some of these reactions may be associated with clear-
cut changes in the EEG; some others, however, consist of
changes that are difficult to visualize. The research field dedi-
cated to the detection, quantification, and physiologic analysis
of those slight EEG changes that are related to particular events
has steadily grown as a field of great interest in the last decades.
These EEG changes may be treated globally under the common
term event-related potentials (ERPs); a subset of the ERPs is sen-
sory (visual, auditory, somatosensory) evoked potentials
(ERPs). This chapter focuses on some general aspects of ERPs.
Detailed descriptions of specific aspects of ERPs to stimuli of
different modalities are presented in separate chapters (visual in
Chapter 46, auditory in Chapter 47, and somatosensory in
Chapter 48), as well as general aspects of event-related
(des)ynchronization in Chapter 45. Specialized aspects of ERPs
of children and infants are presented in Chapter 49 and the use
of ERPs in the operating room is covered in Chapter 38.
Regan published an authoritative specialized textbook on
this topic in 1972 and a thorough revised form in 1989; other
early general sources include Storm van Leeuwen et al. (1), John
and Thatcher (2), Callaway et al. (3), Barber (4), Chiappa (5),
and Misulis and Fakhoury (6). Reviews of the applications of
computer methods to ERP detection and analysis are those of
Barrett (7), Gevins and Cutillo (8), and Regan (9). Insightful
accounts of steady-state (10) and transient evoked potentials
(11) are presented in the book edited by Nunez Neocortical
Dynamics and Human EEG Rhythms (10).
EVOKED-RELATED POTENTIAL
ANALYSIS: GENERAL ASPECTS
ERPs are usually defined in the time domain as the brain elec-
trical activity that is triggered by the occurrence of particular
events or stimuli. A basic problem of analysis is how to detect
ERP activity within the often much larger ongoing EEG or
background activity (i.e., the activity not related to the stimu-
lus). Dawson (12), a pioneer in the field, attempted first to solve
this problem by the photographic superposition on the screen
of an oscilloscope of a number of time-locked responses. This
method had the property of enhancing response in contrast to
CHAPTER
44
FERNANDO H. LOPES DA SILVA
the ongoing activity while providing, at the same time, an indi-
cation of response variability. It is, however, difficult to quantify
the results in this way. Since the introduction of the summation
method (13) and the development of digital computer tech-
niques (14–16), ERPs are obtained in special-purpose digital
computers and expressed as time-varying functions. In general
terms, there is no fundamental argument to justify the prefer-
ence for either time or frequency ERP analysis. Both reflect the
same reality. The choice of analytic domain, therefore, is deter-
mined by practical considerations. Nevertheless, a certain
model of the generation of ERPs is always implicit in the choice
of a method of analysis of these potentials.
BASIC MODELS OF ERPS
According to the most widely accepted model, ERPs are signals
generated by neural populations that are time-locked to the
stimulus; these signals would be summed to the ongoing EEG
activity. According to another model, however, ERPs are
assumed to result, at least partially, from a reorganization of the
ongoing activity. This view arose from the seminal work of
Sayers et al. (17) who analyzed segments of EEG recorded
before and immediately after the presentation of auditory stim-
uli. They carried out a Fourier analysis computing the real and
imaginary coefficients of the signal’s frequency components. In
this way they constructed amplitude and phase spectra, and
found that the brainstem-evoked potentials (BAEPs) to audi-
tory stimuli could be distinguished much better using the dis-
tributions of phase spectral values, rather than the distributions
of amplitude values. In a later study McClelland and Sayers (18)
showed further that the magnitude of the differences between
amplitude spectra of pre- and post-stimulus epochs decreased
progressively as stimulus intensity was reduced. Examining sep-
arately amplitude and phase spectra of the response ensemble,
these authors showed that the phase standard deviation for
responses at auditory threshold was generally significantly
smaller than for the corresponding harmonics obtained from
subthreshold responses. In contrast, the amplitude differences
between populations of threshold and subthreshold responses
were generally non-significant. These findings led the authors
to conclude that the use of BAEPs for auditory threshold deter-
mination should preferentially be based on ensemble phase
measures rather than amplitude measures.
923
924 Part VII ■ Evoked Potentials and Event-Related EEG Phenomena
These experimental results indicate that the brain response
to a stimulus may consist, at least partially, of a decrease of
phase variance of certain frequency components, that is, an
enhanced alignment of the phase components of ongoing EEG
or MEG signals, or phase resetting.
More recently Makeig et al. (19) demonstrated that ERPs
could be generated by stimulus-induced phase resetting of
ongoing EEG components. Mazaheri and Picton (20) showed,
by subtracting the spectra of the average ERP from the average
spectra of single trials, that ERPs can involve phase resetting of
the ongoing EEG rhythms. Nonetheless, Mäkinen et al. (21)
argued that a high degree of independence exists between
ongoing brain activity and auditory ERPs, and suggested to use
amplitude variance to distinguish whether or not phase reset-
ting is present in an ERP. An unambiguous assessment of phase
resetting, however, implies a measure of variance of phase val-
ues and cannot be based just on amplitude variance, as appro-
priately pointed out by Klimesch et al. (22).
This issue was addressed by Mazaheri and Jensen (23) in an
interesting way; they introduced a measure—the phase preser-
vation index—to compare quantitatively the phase of alpha
oscillations before and after a visual stimulus. The main finding
was that the alpha oscillations preserve their phase relationship
with respect to the phase before the stimulus, and their conclu-
sion was that different neuronal populations appear to generate
the ongoing oscillations and the ERPs, but this does not
imply that the neuronal populations are macroscopically
distinguishable.
An underlying problem, however, is the fact that in all these
cases of event-related responses the signal-to-noise ratio plays a
fundamental role. The original findings of Sayers et al. (17) that
the phase spectrum was more powerful than the amplitude
spectrum in distinguishing auditory ERPs were obtained at very
low stimulus intensities, that is, at the threshold for auditory
discrimination. This is not the case in many other experiments.
In a recent overview of this issue we (24) proposed that it is
likely that the relative amount of phase preservation from the
pre- to the post-stimulus epochs will depend on several factors;
at least four appear to be particularly relevant:
• the baseline level of the ongoing oscillations,
• the frequency band of the oscillations,
• the stimulus modality and its strength, and
• the specific cortical area, since EEG rhythmic activities
are not equally distributed over the cortex.
Moreover, it is important to note that a solution to this ques-
tion may not be resolved using only scalp EEG/MEG records, as
also stated by Mazaheri and Jensen (23) who propose that
intracranial recordings will be necessary, combining single cell
and local field potential recordings with EEG signals.
We may state from a theoretical point of view that it is not
probable that two independent neuronal populations, one
generating exclusively ongoing activity and another one
responsible for the evoked response, would exist side by side
as distinct entities in a given brain area. It is more likely that
the same neuronal elements are involved in the generation of
both types of signals but working in different modes. This
same idea was formulated by Shah et al. (25) who noted that
EEG ongoing activity and ERPs are generated by overlapping
neuronal elements. According to this model, ERPs may engage
a group of neurons, by way of two basic mechanisms: either by
enhancing the synchrony of increases, or decreases, of neu-
ronal firing rates and/or by synchronized membrane tran-
sients or oscillations. These two mechanisms are not mutually
exclusive. It follows that the former mechanism will lead to
changes of amplitude, but this will also be the case for the lat-
ter, since the activity of those neurons that work synchro-
nously in phase corresponds to larger amplitude at the level of
the population than that of neurons that display a random, or
weak, alignment of phases. This reinforces the conclusions
that amplitude variance is not sufficient to determine whether
phase resetting occurs in event-related responses. In any case,
the assessment of phase resetting needs the direct estimation
of phase variance of frequency components. This means that,
in the most general case, the analysis of event-related
EEG/MEG changes of activity should take into consideration
not only spectral amplitude, but also phase. In this respect, it
is also interesting to note that recently more attention is being
put on the use of the imaginary part of coherency, which in
fact constitutes the phase spectrum, to analyze interactions
between different EEG signals (26).
Notwithstanding the importance of phase in the generation
of ERPs in general, the wide use of time averaging of ERP
amplitude waveforms in practical ERP work is so overwhelm-
ing that this form of analysis is treated first in the following sec-
tion. After that we consider analyses in the frequency domain
and lastly analyses according to integrated time–frequency
approaches.
TIME AVERAGING OF ERPs
According to the classic view, ERP analysis is based on two basic
assumptions: (i) the electrical response evoked from the brain
is invariably delayed relative to the stimulus and (ii) the ongo-
ing activity is a stationary noise, the samples of which may or
may not be correlated. In this way, ERP can be considered a sig-
nal (s(k)) corrupted by additive noise (n(k)), the ongoing activ-
ity. (Assume that the signals are sampled, with k as the discrete
time variable.) ERP detection becomes, in this way, a question
of improving signal-to-noise ratio. In the more general and
simplified case one assumes a simple additive model (stochastic
variables are indicated by underlined symbols). The recorded
stochastic signal x(k) is given as a sum of two terms:
x(k) s(k) n(k) (44.1)
The expected value of n(k) is zero. The average of x(k) over N
realizations is then defined as:
1 N
x(k) a xi (k) (44.2)
Ni 1
The expected value of the average is given as:
1 N
E[x(k)] Ec a xii (k) d s(k) (44.3)
Ni 1
 Chapter 44 ■ Event-Related Potentials: General Aspects of Methodology and Quantification 925

[since E(n(k))] = 0. The variance of x(k) is as follows:
1 N 2
1 value of the EP.
var[x(k)] Ec a a n i (k) b d var[n(k)] (44.4)
Ni 1 N In principal components analysis, the first component is the
since s(k) is assumed to be invariant. Therefore, the signal-to-
noise ratio in terms of amplitude improves proportionally to
1 n . In more general terms, however, it should be assumed that
the signal s(k) is a stochastic signal, so that expression 44.1
should be reformulated as follows:
x(k) s(k) n(k)
The importance of this apparently small difference lies in the
fact that the model given by expression 44.5 implies that the EP
is not fully described by the mean value of x(k), but also by the
higher statistical moments, as, for instance, the variance. In
most practical cases, the latter model (expression 44.5) gives a
better account of reality. As pointed out by John et al. (27), the
former model (expression 44.1) is valid only for anesthetized
preparations and possibly for the short-latency components of
EP that mainly reflect sensory processes; this is, however, not
the case for long-latency components.
For a comprehensive study of EP, multivariate analysis meth-
ods may be used (28–31). In this case, one considers an EP to be
a multidimensional stochastic variable, that is, a vector x where
the elements are the values in subsequent time points:
x (x(1) p x(NS) ) T
Thus, the expression 44.5 can be rewritten as a vectorial sum:
x s n (44.7)
assuming that E(n ) 0 as before. Therefore:
E(x) E( s )
Nevertheless, the method can be very useful for data reduction
in EP research where one wants to go further than the mean
one accounting for the greatest variance; as many components
as one wishes can be determined in order to account for a cer-
tain percentage of the total variance. In EP analysis, it is usually
not necessary to go further than i = 4 or 5. In this way, a parsi-
monious description of EP can be achieved. Principal compo-
nents analysis, however, is no more than a transformation of the
(44.5) original data. In this form of analysis, no provision can be made
for variance components attributable only to the unreliability
of the observations (32). This and other shortcomings of prin-
cipal components analysis can be solved by way of the mathe-
matical model of factor analysis.
In factor analysis, each observation xi is represented in terms
of a linear function of common factor variables and of a signal
latent-specific variable that can be represented as follows (pro-
vided that the means of the observations are zero):
x1 l ijyj p l im ym ej
p (44.11)
xn l njyj l nm ym en
where yj is the jth common factor variable, ij is called the load-
ing of the ith observation on the jth common factor, and ei is
the ith specific factor variable. The covariances of the observa-
(44.6) tions xi are reflected only on the specific factors. Factor analysis
is based on a specific statistical model in which the number of
common factors is specified beforehand; an example of factor
analysis in EP research is shown in Figure 44.1. These methods
of analysis can be of great usefulness in order to provide a small
set of features that can be employed for the classification
and clustering of EP as discussed below (28,34–37). As an
(44.8)

The second-order moment of x is given by the covariance

matrix C; considering any two samples represented by k and 1,
one may write:
Covk1 E[ (x(k) E(x(k) ) ) (x(1) E(x(1) ) )] (44.9)
There are multivariate statistical methods that enable the struc-
ture of such signals as EPs to be represented as functions of a
small number of factors. A technique to achieve this form of
representation is the principal components method. The essence
of the principal components method is to account for the EP as
a linear combination of a small number of basis functions. It
can be proved (e.g., Ref. 32) that, if one chooses as basic func-
tions the eigenfunctions c i, of the covariance matrix, the fol-
lowing expression gives the best p-dimensional approximation
(in a least-squares sense) to the EP (P NS):
p
x E(x) a l 1c i (44.10) Figure 44.1 Average evoked potential (ERP) of one subject (parietal
i 1 leads; N = 512 epochs; epoch length, 765 msec) and the corresponding
In the expression, the coefficients l are uncorrelated and the seven factor loading plots. The reality of the first two factors in the
eigenfunctions c 1 are orthogonal. It should be noted, however, waveform of this example is questionable, because the factors also
that this form of description is only a statistical way of describ- showed loadings in the prestimulus period. (Adapted from Rebert CS.
ing EP; one cannot assume that the eigenfunctions necessarily Neuroelectric measures of lateral specialization in relation to perform-
correspond to independent physiologic generators (33). ance. Electroencephalogr Clin Neurophysiol. 1978;34:231–238.)
926 Part VII ■ Evoked Potentials and Event-Related EEG Phenomena

illustration of the usefulness of principal components and dis-

criminant analysis, the study of Donchin and Cohen (38) found
that the discriminant between EP recorded to task-relevant and
task-irrelevant stimuli was based essentially on time points at
300 msec.
Another form of data reduction of average EP involves
describing EP as a sum of a set of analytic functions, such as
damped sinusoids as thoroughly investigated by Freeman
(39,40). Later, this chapter discusses the possibility of describ-
ing EP as a sum of harmonically related sine waves (i.e., Fourier
components).

SPECIAL PROBLEMS IN ERP

TIME AVERAGING
ERPs in Strong Rhythmic Background EEG
One problem is whether the statistical properties of the back-
ground activity may have influence on the degree of improve-
ment regarding signal-to-noise ratio realized by the averaging
procedure. In detecting ERPs by classic time averaging, the
main objective, of course, is to increase the signal-to-noise ratio
so that the EEG background activity will be attenuated.
Equation 44.4 illustrates that the ratio between the variance of
the averaged signal and the variance of the noise is 1/N. This
holds for the case where the noise samples are uncorrelated.
Often, however, this is not the case, as, for instance, in the pres-
ence of a strong rhythmic background such as an alpha rhythm.
In such a case, subsequent samples of n(k) are not independent.
The degree of dependence of the samples of n(k) is given by the
autocorrelation function of the background activity (see for a
definition Chapter 54).
Assuming that 2 is the var[n(k)], the ERP variance
[var(x(k))] is not equal to [1/N(var(n(k)))], as in Equation 44.4,
but equal to [ 2/N(var(n(k)))] multiplied by a factor that
depends on the properties of the autocorrelation function of the
background activity (41). It is therefore important to emphasize
that rhythmicity of the background can influence the signal-to-
noise ratio. McGillem and Aunon (42), Steeger (43), and Steeger
and Reinhardt (44) described methods that can provide better
estimates of average ERP because they take into account the
statistical properties of the ongoing EEG activity.
A related problem is whether the stimulation should be peri-
odic or aperiodic; aperiodic stimulation can lead to reduced
ERP variance as shown theoretically by Ten Hoopen (45) and
experimentally by Arnal and Gerin (46). More specifically,
Ruchkin (47) has demonstrated that the variance of the averag-
ing estimate is closer to zero when the interstimuli intervals are
exponentially distributed (Fig. 44.2).
Single-Trial ERPs and Classification
In the last decades there have been several studies with the aim
of identifying single-trial ERPs and/or to discriminate
subgroups of ERPs among a series of trials. Initially most of
the approaches applied were based on cross-correlation of
EEG signals, including the ERP, and some template a priori
determined (42,48), and on maximum likelihood detectors
Figure 44.2 Average ERPs recorded between vertex and mastoid; stim-
ulus is a sinusoidal tone of 1000 Hz, 40 dB above 2104 bar, 600 msec.
A: The results of simple averaging are shown for a number of averag-
ing steps. B: The results obtained using an adaptive filter (based on a
maximum likelihood detector adapted to broadband colored noise) fol-
lowed by synchronization and averaging are shown. Note the improve-
ment in ERP discrimination. The plots are in relative vertical scale; t st
is the stimulus onset. (Courtesy of G.H. Steeger.)
(49). According to one of these methods (44), a matched filter
is created that is adapted optimally to the template (the aver-
age ERP) as well as to the actual ongoing activity, which is
estimated by its autocorrelation function (Woody’s adaptive
filter was only adapted optimally to the template and to white
noise). Each single-trial EEG segment is then passed through
the matched filter and the maximum output signal is found;
in this way an optimal average ERP can be obtained after cor-
rection for latency deviations. A method related to those
described above but also involving a classification process was
proposed by Pfurtscheller and Cooper (50) under the term
selective averaging; according to this technique, single-trial
ERPs are cross-correlated with different templates. The max-
ima of the cross-correlation functions are used to estimate the
amplitude and latency of particular components. Thereafter,
those trials that fall within certain amplitude and/or latency
limits are averaged. In this way, different types of selected
averages can be obtained that may correspond to different
subsystems and/or physiologic conditions. In cases where all
trials are averaged together, there is a considerable informa-
tion loss. The problem of measuring single-trial ERPs has
been extensively investigated by Coppola et al. (51). In this
study, equations have been proposed to estimate empirically
 Chapter 44 ■ Event-Related Potentials: General Aspects of Methodology and Quantification 927

S/N that may be useful in ERP clinical or psychophysiological Therefore, the averaging reduces the power spectrum of the
studies (for details see Ref. 51). noise term in the unaveraged signal Pnn (fi) by a factor 1/N.
The analysis of single-trial ERP is also important in deter- Moreover, assume that the spectrum of each realization has
mining whether a certain single-trial ERP belongs to a specific been made and that an average of all these spectra has been
ERP class. This problem of classification can be solved using computed: Pxx(fi ) ; thus, it can be written as follows:
multivariate statistical methods (34). In general terms, the
question involves: (i) describing ERP by a set of features (in the Pxx(fi ) Pss(fi ) Pnn (fi ) (44.13)
simplest case, the values of the sample points) and establishing
a feature profile, (ii) making a learning set where ERPs corre- Combining Equations 44.12 and 44.13 one can derive the fol-
sponding to a priori defined classes are pooled, and (iii) finding lowing expression:
a discriminant function (i.e., a decision rule) that must parti-
N 1
tion the space occupied by all objects (the ERP) so that unique Pss(fi ) Pxx(fi ) Pxx(fi ) (44.14)
N 1 N 1
patterns corresponding to different classes can be identified.
(iv) Finally, the class to which any new object (i.e., a single-trial Wiener (54) has shown that for a signal x(k) = s(k) n(k) with
ERP) belongs must be determined; this is accomplished by power spectra Pxx(fi), Pss(fi), and Pnn (fi), a linear filter can be
computing a discriminant score for each object. Thus, each sin- constructed that provides an optimal estimator of the signal
gle-trial ERP is classified in the subspace within which the dis- s(k); this linear filter has a transfer function h(k), or in the fre-
criminant score falls. The ERP features used in the procedure quency domain H(fi), which is given by the following expres-
can be simply the corresponding sample values or a reduced sion (55):
amount of data, such as can be obtained using factor or princi- Pss(fi )
pal component analysis. H(fi ) (44.15)
Pss(fi ) Pnn (fi )
The modern tendency is to use spatiotemporal filtering
methods for single-trial ERP component estimation, such Doyle (56) has shown, however, that one should use an expres-
as methods that make use of the spatial diversity of multi- sion other than 44.15 for defining the filter H(fi) for use on the
channel recordings and thus can extract signals in poor average ERP, based on N realizations:
signal-to-noise ratio conditions (52). Another example of Pss(fi )
such an integrated methodology has been proposed H(fi ) (44.16)
Pss(fi ) (1>N)Pnn (fi )
by Poolman et al. (53); this method consists of multiple pro-
cessing modules that can be applied as a whole, or in part, In this case, notice is taken of the situation after averaging,
including noise mitigation based on directed components because the averaging procedure reduces the power of the noise
analysis (DCA), source–space transformation utilizing a term by a factor 1/N as shown in Equation 44.12. Expression
finite difference model (FDM) of the human head, and 44.16 has been critically reviewed by Albrecht et al. (57), who
discrimination modules that interface with performance pointed out that in the denominator the term (1/N)Pnn(fi)
evaluation modules to generate classification statistics of should be replaced by a more general term (hN(fi) 2)Pnn(fi),
ERPs. This method performed very well in a task consisting where the function hN(fi) 2 reflects the attenuation of the noise
of discriminating targets from non-targets in a rapid serial due to averaging; this function depends not only on N, but also
visual presentation (RSVP). on the length of the realization and on the time interval
Other advanced methods in time–frequency domain are between the end of one realization and the following stimulus
currently used for single-trial ERP estimation combined with (for details see Ref. 57). Using Equation 44.16, the optimal filter
the analysis of the changes in ongoing EEG activities as dis- according to Doyle becomes:
cussed below.
[N>(N 1)] Pxx(fi ) [1>(N 1)] Pxx(fi )
H¿(fi ) (44.17)
Pxx(fi )
Improvement of Time-Averaging Efficiency
It has been suggested that a posteriori or Wiener (54) filtering, The application of filter H(fi) to the averaged signal x(k) can be
introduced by Walter (55) in the EEG field, can improve aver- written formally as follows:
aging efficiency. The following description uses the concepts of
x(k)*F 1 [H¿(fi )] (44.18)
power spectrum and related functions as explained in more
detail in Chapter 54. Assume that the additive model of where H(fi) is the impulse response of the Wiener filter accord-
Equation 44.5, in which the ERP, x(k), is considered the sum of ing to Doyle, the asterisk represents the convolution operation,
a signal s(k) and a noise term n(k), is valid, that s(k) and n(k) and the symbol F–1 indicates the inverse Fourier transforma-
are uncorrelated, and that both have means of zero. The power tion. However, in a theoretical study in which Wiener filtering
spectrum of the average signal x(k) obtained by averaging N was compared to ERP averaging, Albrecht et al. (57) demon-
realizations is given by the following expression: strated that the former yields a better estimate of s(k) than the
latter, although the difference between both decreases with an
1 increased number of realizations. (It can be seen from Equation
Pxx(fi ) Pss(fi ) Pnn (fi ) (44.12)
N 44.16 that as N becomes large, H(fi) tends to unity.) Moreover,
928 Part VII ■ Evoked Potentials and Event-Related EEG Phenomena
they also pointed out, in both cases, the importance of using
interstimuli intervals leading to the smallest error of the aver-
age (i.e., the intervals should be exponentially distributed), as
also shown by Ruchkin (47). Therefore, it appears that Wiener
filtering would be of practical interest in ERP analysis only in
those cases where a small number of realizations are available.
In fact, Strackee and Cerri (58) have shown in a theoretical
study that a posteriori (Wiener) filtering does not improve ERP
averages, although this technique does improve the signal-to-
noise ratio for individual realizations. Albrecht et al. (57) have
performed a number of simulations to study the influence of
interstimuli intervals and other factors on averaged or Wiener-
filtered ERP; when averaging only five realizations, the mean
square error of the estimate was, for the Wiener filter, about
76% that of the straightforward average ERP obtained using
random interstimuli intervals. If 20 realizations were averaged
using randomly distributed stimuli, the error for the Wiener fil-
ter was about 86% of that of the average ERP. The effectiveness
of a posteriori (Wiener) filtering in cases using few (5) stimuli,
but not larger sets (20), has also been demonstrated by Hartwell
and Erwin (59). De Weerd et al. (60,61) have described an adap-
tive time-varying filter capable of optimizing the estimation of
both low-frequency components of relatively long duration and
higher frequency components of short duration.
Variability of ERPs
It is highly recommended that the averaging procedure
should include some form of estimate of ERP variability and
the ongoing activity. A very simple solution is to compute
partial averages of ERPs (e.g., five sets of N realizations each)
and the corresponding means and standard deviations or, at
least, to compare graphically the different realizations. This
solution has the disadvantage of requiring long stimulation
sessions. Another useful procedure, which may be combined
with the former, is to perform a similar computation of par-
tial averages of EEG segments, but without the application of
the stimulus while averaging the same number of realiza-
tions. In this way one can estimate the mean and standard
deviation of the ongoing activity in the absence of the stimu-
lus. In any case, it is always best to include in the signal to be
averaged a period of time prior to the delivery of the stimu-
lus in order to get an estimate of the mean and variability of
the ongoing activity. An interesting method that can be com-
bined with the normal averaging procedure in order to esti-
mate the magnitude and structure of ongoing activity is the
plus–minus ( ) average proposed by Schimmel (62). The
( ) average is computed by alternate addition and subtrac-
tion of each realization and division by N (even number); the
deviation of each ERP about the average value and the ( )
average has the same statistical structure. The latter can be
implemented more easily by either manually or automatically
setting the averaging computer to “add” or “subtract” mode
or by alternating input polarity. In many instances, the main
limitation encountered in practice is the great difficulty of
recording a sufficiently large number of ERPs under equiva-
lent conditions. Therefore, the methods of improving signal-
to-noise ratio using a small number of trials (or even single
ones) are of great importance. An interesting addition to the
analytic tools that can be used for assessing the variability of
ERPs was recently proposed by Hu et al. (63) who developed
an unsupervised correlation-based clustering method to
quantify trial-to-trial variability of auditory evoked poten-
tials (AEPs).
ANALYSIS IN THE FREQUENCY DOMAIN
In the introductory remarks we discussed the benefit of an
analysis in the frequency domain of ERPs. This section consi-
ders, in particular, (i) those ERPs that are difficult to detect in
the time domain because they occur without a fixed phase or
time relation to the trigger and (ii) ERPs caused by continuous
stimuli that are preferably analyzed in the frequency domain.
ERPs consisting of changes in spectral intensity, for example, in
the alpha frequency band in relation to voluntary hand move-
ments (64,65) exemplify the first category. This issue is
described in detail in Chapter 45 and thus we just mention here
a few general aspects. This type of ERP can best be shown by
decomposing the EEG in frequency bands and averaging the
power within different bands in relation to the trigger. It is thus
necessary to use statistical methods of ERP evaluation to test
whether or not the changes occurred by chance.
Steady-state ERPs are elicited by continuous stimuli as, for
example, sine wave–modulated light or amplitude-modulated
tone. These ERPs may be described by measuring their ampli-
tudes and phases as functions of the frequency of the stimulus
as performed in Amsterdam’s school of van der Tweel (66)
and collaborators (41,67) and by Regan (9,68,69). If the phys-
iologic systems underlying the generation of the ERP were lin-
ear, the information from a study of the system by way of
transient or continuous stimuli would be the same. However,
considering that most of these systems have nonlinear proper-
ties, both types of stimuli can provide different information.
To construct an amplitude and phase plot of a steady-state
visual ERP, for example, it is necessary to stimulate succes-
sively with different frequencies. An alternative is to stimulate
with a light of a given mean luminosity, the intensity of which
is modulated by Gaussian noise. In this way, the stimulus con-
tains all frequencies at the same time. The ERP is thus
obtained by calculating the cross-correlation function
(Equation 56.10 in Chapter 56) or the cross-power spectrum
(Equation 56.11) between the stimulus and the corresponding
EEG (67,70,71); it can be shown that the result of the cross-
correlation in such a case is equivalent to the impulse
response of the linear components of the system (Fig. 44.3).
The system, however, may contain nonlinear elements; in such
a case, its characteristics may be analyzed using noise-
modulated light by computing higher order cross-spectra or
cross-correlation functions that can be defined similarly to
the bispectrum (see Equation 56.16). This question, however,
is too specialized to be considered in detail here (see review in
Refs. 9, 41, and 71). Steady-state stimuli particularly of the
visual modality (SSVEP) have been applied extensively in a
series of investigations by the groups of Nunez and Silberstein
(10,72–74) in both neurocognitive and clinical investigations
 Chapter 44 ■ Event-Related Potentials: General Aspects of Methodology and Quantification
Figure 44.3 A cross-correlation function between the noise-modulated
light stimulus field and the EEG. Note that in the cross-correlation
responses several components can be distinguished; the values of the
cross-correlation responses at negative times give a measure for the
variance of the response owing to the finite time length (4 minutes) of
the EEG record used for analysis. The responses were derived from two
electrodes placed on the midline (occipital region) in relation to a refer-
ence placed on the earlobe. Below the Fourier transformation are the
cross-correlation responses (i.e., the cross-power and phase spectra are
shown along with the normalized cross-power spectrum or coherence
function). From the slopes of the phase spectra, estimates of the time
delays of the responses can be obtained. It can be seen that up to about
20 Hz the phase lag increases more rapidly with frequency (i.e., the
time delays are longer) than for frequencies above 20 Hz. In the coher-
ence spectra there are two maxima; one is in the range 14 to 20 Hz and
the other is in the range of 30 to 45 Hz. The former corresponds to the
response components in the medium frequency range that have laten-
cies of 100 to 120 msec; the latter corresponds to the early components
of higher frequency content with latency of about 30 msec. (Adapted
from Spekreijse H, Estevez O, Reits D. Visual evoked potentials and the
physiological analysis of visual processes in man. In: Desmedt JE, ed.
Visual Evoked Potentials in Man. Oxford: Clarendon Press; 1976:3–15.)
of brain function, namely of memory and aging (75). In these
studies SSVEPs are used as a way to characterize preferred fre-
quencies of neocortical dynamic processes, especially because
of the excellent signal-to-noise ratios that can be obtained
using this form of stimulation, rather than with the aim of
investigating visual processes as such.
Besides these studies where the visual modality was stimu-
lated under steady-state conditions, also steady-state ampli-
tude- or frequency-modulated sound stimuli were used to
investigate functional properties of the auditory system in an
important series of studies (76,77). The corresponding ERPs
can be analyzed using frequency analysis. As described by
Picton et al. (78) the auditory steady-state ERP amplitude
increases as a function of modulation depth and/or sound
intensity, and the phase delay of the response increases as the
intensity or the carrier frequency decreases. Auditory steady-
state responses are generated at various levels of the auditory
929
nervous system, but cortical sources appear to contribute more
than brainstem neuronal sources to responses at lower
modulation frequencies. Several studies showed that these
steady-state ERPs can be used for an objective evaluation of
auditory thresholds, and also to analyze suprathreshold hear-
ing. These stimuli were shown to be useful for the study of not
only specific auditory functions, but also attentional processes
(79) and other neurocognitive processes (80,81).
TIME/ FREQUENCY ANALYSIS:
NEW APPROACHES
The findings that have been obtained in the course of the last
decades, regarding the relationships between ongoing EEG sig-
nals and ERPs, as discussed above, have led to the development
of new tools to analyze these responses. The main purpose is to
integrate time- and frequency-domain analyses. The basic con-
cept is that ERPs “do not emerge out a flat EEG baseline” (19),
but interfere in a more or less complex way with the ongoing
EEG activity. In this respect the neuronal sources contributing
to the ERP and to the ongoing EEG may share the same neu-
ronal networks, although these may work in different modes.
Among other observations the finding of Vijn et al. (82) who
showed that visual stimuli reduce the ongoing EEG amplitude
supports the notion that an ERP is not simply a response added
to the background EEG. This notion is also in line with the
experimental results of McClelland and Sayers (18), already
mentioned above, indicating that the brain response to a stim-
ulus may consist, at least partially, of a decrease of phase vari-
ance of certain frequency components, that is, an enhanced
alignment of some phase components of the ongoing
EEG/MEG activity. This phenomenon is commonly called
“phase resetting” of ongoing EEG frequency components.
According to this way of reasoning, ERPs can be considered to
result from a process incorporating phase resetting of specific
frequency components. In this manner, the notion of signal-to-
noise ratio in terms of power, or amplitude, does not provide all
the relevant information. Therefore, a method of detecting
ERPs based on phase spectra evaluation may be more fruitful
than simple time averaging. This issue was more recently inves-
tigated by Makeig et al. (19,83), reinforcing the concept that
phase resetting of neuronal oscillations, as originally hypothe-
sized by Sayers et al. (17) and also by Basar (84), can contribute
to the generation of ERPs.
A practical approach uses independent component analysis
(ICA) to decompose the EEG activity into the sum of distinct
temporally independent component processes (see also
Chapter 54) (85–87). In contrast to principal component analy-
sis, ICA does not require the components to be orthogonal. In
summary this novel approach consists of several steps. First, the
signal is decomposed into a number of independent compo-
nents (ICA); second, the spectral power of the components
(ERSP), and, third, the inter-trial phase coherence (ITC) are
computed. Applying ITC the phase-locked frequency compo-
nents of the ERPs can be identified. It should be noted that ITC
is similar to the “phase-averaging” measure proposed by Tallon-
Baudry et al. (88) to quantify phase-locking of EEG oscillatory
930 Part VII ■ Evoked Potentials and Event-Related EEG Phenomena

Figure 44.4 Maximally independent EEG components (ICA) compatible with a single cortical source.
Scalp maps, event-related spectra perturbation (ERSP) and inter-trial phase coherence (ITC) images,
and ERPs for two typical independent EEG processes during a task consisting of visual letter encoding
(green in images indicates p. 0.001). Below are the best fitting equivalent current dipole (ECD) models
(“r.v.” is the residual scalp map variance across the 69 scalp electrodes). A: Frontal midline theta com-
ponent with radial ECD. Spectral power increases shown in the ERSP accompany weak phase locking
on ITC and slow ERP effects (mV as at Fz). B: Mu-rhythm component with tangential ECD in right
somatomotor cortex. Alpha- and beta-band (ERSP) power changes, accompanied by partial phase lock-
ing (ITC) contributing to the component ERP. (Adapted from Makeig S, Debener S, Onton J, et al.
Mining event-related brain dynamics. Trends Cogn Sci. 2004;8(5):204–210.) (See color insert)

bursts with respect to a certain stimulus. The latter method
used a decomposition of the EEG signals using a family of
wavelets that can provide a time–frequency representation of
the energy of the signal (TF energy) (Fig. 44.4).
One aspect that merits being underlined is that these
novel tools may be used in combination with each other to
solve specific problems. In this context the study of Jung et
al. (89) demonstrated how ICA applied to single-trial multi-
channel EEG data from ERP experiments can be used com-
bined with spatial filters to identify components arising
from distinct or overlapping brain or extra-brain sources.
These authors created a new visualization tool, the “ERP
image,” in order to display single-trial variations in the
amplitudes and latencies of ERPs. This approach can be used
to discriminate ERPs associated with specific behavioral or
physiologic variables.
Time–frequency analysis of ERPs can be performed using
different kinds of basic functions. One kind of interesting func-
tions in this respect is the time–frequency dictionary of Gabor
functions, that is, sine functions modulated by Gaussians as
proposed by Sieluzycki et al. (90), who applied a multivariate
version of the matching pursuit algorithm (MMP) to perform
an iterative search for auditory MEG ERP waveforms, common
to all trials with the same time of occurrence, frequency, phase,
and time width, but varying amplitude in a study of habitua-
tion to auditory stimuli.
ERP TOPOLOGY
In separate chapters the topology of special ERP types is dis-
cussed. This section presents only some of the general aspects
concerning the topologic analysis of ERP. Different forms of
ERP involve activity within more or less well-circumscribed
areas of the cortical surface and even subcortical areas, as in the
case of the far-field ERP. Therefore, at scalp level, the ensuing
electrical activity can be recorded over a relatively large region
due to the presence of volume conduction. In this sense, ERP
recording using two electrodes gives a very limited image of the
real underlying neural activity. It is in some way an equivalent
of an ERP in the time domain based on the amplitude value at
two times samples only. Nevertheless, valuable ERP research
can be carried out using only one bipolar deviation if it is
placed on the appropriate brain region. To estimate the localiza-
tion of the neural systems responsible for ERP generation, how-
ever, it is essential to sample the scalp surface appropriately. As
discussed in Chapter 54, sampling of the EEG in space must
take into consideration spatial frequencies present in the ERP.
To determine these frequencies, it is necessary to have a priori
 Chapter 44 ■ Event-Related Potentials: General Aspects of Methodology and Quantification
knowledge of the spatial characteristics of the ERP in question.
This can be achieved by multichannel recordings with small
interelectrode distances. In this way, Vaughan (91), Vaughan et
al. (92), and Vaughan and Ritter (93) mapped the scalp for
several ERP types. In these studies, derivations against a com-
mon reference were used; the presence of significant activity in
the reference (chin or nose) and artifacts (muscular and eye
movements) was carefully assessed. It should be noted that, in
this type of studies, the pitfalls of simple averaging must be
taken into consideration. In other words, the fact that the EEG
background activity may differ substantially with scalp location
should be considered when processing ERPs recorded at differ-
ent sites; as a rule, the number of realizations for an identical
S/N improvement should not be the same at all recording sites.
Vaughan (91) was able to make isopotential maps for different
time samples of several forms of ERP and to relate the spatial
distributions with results of a computer model.
The early studies of Rémond and collaborators (94,95) and
of Lehmann (96,97) also provide estimates of spatial gradients
characteristic of some ERP types. The general problem of spa-
tial properties of EEG signals is discussed in more detail in
Chapters 54 and 55. In addition to the early studies mentioned
above, the new methodologic approaches have contributed to a
much finer analysis of processes of sensory information in dif-
ferent modalities, namely in auditory processing (98–100), in
the analysis of the visual parallel pathways that lead to the stri-
ate (V1) and prestriate cortex in the human brain (101), in the
functional localization of face-evoked MEG and EEG responses
using each modality separately, and by their “fusion” under a
common generative model (102) and of the somatosensory
system (103,104) including pain-related fields (105,106).
Methodologic aspects of MEG and EEG analysis of event-
related events were reviewed by Kakigi et al. (107) and by
Michel et al. (108). The relevance of spatial analysis of ERPs
based on high-density EEG, or MEG, recordings must be
emphasized (109). Significant advances were obtained regard-
ing the integration of EEG and MEG evoked potentials or fields
using dynamic causal modeling (110–112). This modeling
approach combined with advanced time–frequency analysis is
most promising in yielding new insights in the neural processes
underlying the generation of ERPs.
CONCLUDING REMARKS
Since the early phase of ERPs studies successive waves of
methodologic advances have taken place. This field has been
always a fruitful area for collaborative investigations between
biomedical engineers, particularly signal analysts, statisticians,
and physicists on the one hand, and neurophysiologists, neurol-
ogists, and psychologists on the other. The last decade has seen
an impressive increase of the number of analytic tools within
this field and of theoretical models with the emergence of novel
concepts about the generation and nature of ERPs. In this way
a plethora of methods are now available but the practical clini-
cal neurophysiologist may have difficulty in finding his way
within this multitude of challenging possibilities. The relative
value of different methods with respect to performance in ana-
lyzing ERPs cannot, in general, be assessed since most published
931
studies present the authors’ own methods implemented in
their “home” environment, which does not allow a comparative
and objective evaluation of the performance of different
methodologies applied to the same data. A few studies have
included comparative analysis of different methods, such as the
demonstration by Iyer and Zouridakis (113) that the applica-
tion of an iterative ICA procedure to the analysis of single-trial
auditory N100 component can yield significantly improved
results compared to plain averaging and wavelet denoising, and
that of Wang et al. (114) who demonstrated that wavelet-based
approaches yield better results than other methods including
the Wiener filter, least mean square (LMS), and recursive least
squares (RLS), especially based on translation-invariant
wavelets. In this field an important step forward was the intro-
duction of open source toolboxes that can be downloaded from
an Internet site. One of these is the “EEGLAB” toolbox for the
analysis of single-trial EEG dynamics and ERPs developed by
Delorme and Makeig (115) based on a MATLAB environment,
and the “ERPWAVELAB” toolbox developed by Morup et al.
(2007) for multichannel time–frequency analysis of event-
related activity of EEG and MEG data and that is an extension
of “EEGLAB,” which is required for its performance. “EEGLAB”
development is now centered at the Swartz Center for
Computational Neuroscience of the Institute for Neural
Computation at the University of California San Diego (UCSD)
in collaboration with the CNRS CERCO laboratory in France.
In summary there is a need for critical and open discussions
among users with experience in the field of the application of
these methodologies presented here, and for comprehensive
comparative studies of different methodologies, that may lead
to useful advances in the practical application of ERPs in neu-
rophysiologic investigations. Furthermore, the combination of
more sophisticated models of the process of generation of ERPs
with advanced analytic methods is likely to lead to major
insights in our understanding of brain functional dynamics.
REFERENCES
1. Storm van Leeuwen W, Lopes da Silva FH, Kamp A. Evoked
responses. Part A. In: Buser P, ed., Rémond A, ed.-in-chief.
Handbook of Electroencephalography and Clinical Neurophysiology.
Vol 8. Amsterdam: Elsevier; 1975.
2. John ER, Thatcher RW. Functional Neuroscience/Neurometrics.
Vol II. Hillsdale, NJ: Lawrence Erlbaum; 1977.
3. Callaway E, Tueting P, Koslow SH, eds. Event-related Brain
Potentials in Man. New York: Academic Press; 1978.
4. Barber C, ed. Evoked Potentials. Lancaster: MTP Press; 1980.
5. Chiappa KH, ed. Evoked Potentials in Clinical Medicine. 3rd ed.
Lippincott-Raven Press; 1997:700pp.
6. Misulis KE, Fakhoury T. Spehlmann’s Evoked Potential Primer. 3rd
ed. Butterworth-Heinemann; 2001:184pp.
7. Barrett G. Analytic techniques in the estimation of evoked poten-
tials. In: Lopes da Silva FH, Storm van Leeuwen W, Rémond A, eds.
Clinical Applications of Computer Analysis of EEG and Other
Neurophysiological Signals. Handbook of Electroencephalography
and Clinical Neurophysiology. Vol 2. Amsterdam: Elsevier;
1986:311–334 [new series].
8. Gevins AS, Cutillo BA. Signals of cognition. In: Lopes da Silva FH,
Storm van Leeuwen W, Rémond A, eds. Clinical Applications of
Computer Analysis of EEG and Other Neurophysiological Signals.
932 Part VII ■ Evoked Potentials and Event-Related EEG Phenomena
Handbook of Electroencephalography and Clinical Neurophysiology.
Vol 2. Amsterdam: Elsevier; 1986:335–384 [new series].
9. Regan D. Human Brain Electrophysiology: Evoked Potentials and
Evoked Magnetic Fields in Science and Medicine. New York: Elsevier
Science; 1989:672pp.
10. Silberstein RB. Neuroelectric measures of mind. In: Nunez PL, ed.
Neocortical Dynamics and Human EEG Rhythms. Oxford
University Press; 1995:591–627.
11. Gevins AS, Cutillo BA. Neuroelectric measures of mind. In: Nunez
PL, ed. Neocortical Dynamics and Human EEG Rhythms. Oxford
University Press; 1995:304–338.
12. Dawson GD. A summation technique for detecting small signals in
a large irregular background. J Physiol. 1951;115:2–3.
13. Dawson GD. A summation technique for the detection of small
evoked potentials. Electroencephalogr Clin Neurophysiol. 1954;6:
153–154.
14. Barlow JS. An electronic method for detecting evoked responses of
the brain and for reproducing their average waveforms.
Electroencephalogr Clin Neurophysiol. 1957;9(2):340–343.
15. Brazier MAB. Some uses of computers in experimental neurology.
Exp Neurol. 1960;2:123–143.
16. Clark WA. Average response computer (ARC1). Prog Rep Res Lab
Electron MIT. 1958;114–117.
17. Sayers P, Beagley HA, Hanshall WR. The mechanisms of auditory
evoked EEG responses. Nature. 1974;247:481–483.
18. McClelland RJ, Sayers BM. Towards fully objective evoked
response audiometry. Br J Audiol. 1983;17(4):263–270.
19. Makeig S, Debener S, Onton J, et al. Mining event-related brain
dynamics. Trends Cogn Sci. 2004;8(5):204–210.
20. Mazaheri A, Picton TW. EEG spectral dynamics during discrimi-
nation of auditory and visual targets. Brain Res Cogn Brain Res.
2005;24(1):81–96.
21. Mäkinen V, Tiitinen H, May P. Auditory event related responses are
generated independently of ongoing brain activity. Neuroimage.
2005;24:961–1068.
22. Klimesch W, Hanslmayr S, Sauseng P, et al. Distinguishing the
evoked response from phase reset: a comment to Makinen et al.
Neuroimage. 2006;29(3):808–811.
23. Mazaheri A, Jensen O. Posterior {alpha} activity is not phase-reset
by visual stimuli. Proc Natl Acad Sci U S A. 2006;103(8):2948–2952.
24. Lopes da Silva FH. Event-related neural activities: what about
phase? Prog Brain Res. 2006;159:3–17.
25. Shah AS, Bressler SL, Knuth KH, et al. Neural dynamics and the
fundamental mechanisms of event-related brain potentials. Cereb
Cortex. 2004;14:476–483.
26. Nolte G, Bai O, Wheaton L, et al. Identifying true brain interaction
from EEG data using the imaginary part of coherency. Clin
Neurophysiol. 2004;115:2292–2307.
27. John ER, Ruchkin DS, Vidal JJ. Measurement of event-related
potentials. In: Callaway E, Tueting P, Koslow SH, eds. Event-
related Brain Potentials in Man. New York: Academic Press; 1978:
93–138.
28. Donchin E. A multivariate approach to the analysis of average
evoked potentials. IEEE Trans Biomed Eng. 1966;13:131–139.
29. John ER, Ruchkin DS, Villegas J. Signal analysis and behavioral
correlates of evoked potentials configurations in cats. Ann N Y
Acad Sci. 1964;112:362–420.
30. Ruchkin DS, Villegas J, John ER. An analysis of average evoked
potentials making use of least mean square techniques. Ann N Y
Acad Sci. 1964;15:799–826.
31. Streeter DN, Raviv J. Research on Advanced Computer Methods for
Biological Data Processing. Rept. ASTIA, Doc. AD 637452.
Springfield, VA: Defense Documentation Center, Defense Supply
Agency, Clearing House for Federal Scientific and Technical
Information; 1965.
32. Morrison DF. Multivariate Statistical Methods. New York:
McGraw-Hill; 1967.
33. van Rotterdam, 1970.
34. Donchin E. Discriminant analysis in average evoked response
studies: the study of single trial data. Electroencephalogr Clin
Neurophysiol. 1969;27:311–314.
35. Donchin E, Herning RI. A simulation study of the efficacy of step-
wise discriminant analysis in the detection and comparison of
event related potentials. Electroencephalogr Clin Neurophysiol.
1975;38:51–68.
36. John ER, Walker P, Cawood D, et al. Factor analysis of evoked
potentials. Electroencephalogr Clin Neurophysiol. 1973;33:33–34.
37. Rebert CS. Neuroelectric measures of lateral specialization in rela-
tion to performance. Electroencephalogr Clin Neurophysiol Suppl.
1978;34:231–238.
38. Donchin E, Cohen L. Average evoked potentials and intramodality
selective attention. Electroencephalogr Clin Neurophysiol.
1967;22:537–546.
39. Freeman WJ. Use of digital adaptive filters for measuring pre-
pyriform evoked potentials from cats. Exp Neurol. 1964;
10:475–492.
40. Freeman WJ. Mass Action in the Nervous System. New York:
Academic Press; 1975.
41. Spekreijse H, Estevez O, Reits D. Visual evoked potentials and the
physiological analysis of visual processes in man. In: Desmedt JE,
ed. Visual Evoked Potentials in Man. Oxford: Clarendon Press;
1976:3–15.
42. McGillem CD, Aunon JI. Measurement of signal components in
single visually evoked brain potentials. IEEE Trans Biomed Eng.
1977;24(3):232–241.
43. Steeger GH. Ein Bettrag zur Verbesserung der Messung akustisch
evozierter Potentiate variabler Latenz im ElektroenzERPhalogramm
des Menschen [thesis]. Universitat Erlangen Nürnberg; 1979. ^
44. Steeger GH, Reinhardt J. Ein Mittelwertrechner fŸr reizausgel s ste
neuroclektrische Signale varibler Latenz. Biomed Tech. 1977;
22:68–74.
45. Ten Hoopen M. Variance in average response computation: regu-
lar versus irregular stimulation. In: Rémond A, ed. Handbook of
Electroencephalography and Clinical Neurophysiology, Part A. Vol 8.
Amsterdam: Elsevier; 1975;151–158.
46. Arnal D, Gerin P. Étude du bruit résiduel des potentials évoqués
moyens. Electroencephalogr Clin Neurophysiol. 1969;27:315–321.
47. Ruchkin DS. An analysis of average response computations based
upon aperiodic stimuli. IEEE Trans Biomed Eng. 1965;12:87–94.
48. Woody CD. Characterization of an adaptive filter for the analysis
of variable latency neuroelectric signals. Med Biol Eng.
1967;5:539–553.
49. Helstrom CW. Statistical Theory of Signal Detection. 2nd ed.
Oxford: Pergamon Press; 1968.
50. Pfurtscheller G, Cooper R. Selective averaging of the intracerebral
click evoked responses in man: an improved method of measuring
latencies and amplitudes. Electroencephalogr Clin Neurophysiol.
1975;38:187–190.
51. Coppola R, Tabor R, Buchsbaum MS. Signal to noise ratio and
response variability measurements in single trial evoked poten-
tials. Electroencephalogr Clin Neurophysiol. 1978;44:214–222.
52. Li R, Keil A, Principe JC. Single-trial P300 estimation with a spa-
tiotemporal filtering method. J Neurosci Methods. 2009;177(2):
488–496.
 Chapter 44 ■ Event-Related Potentials: General Aspects of Methodology and Quantification
53. Poolman P, Frank RM, Luu P, et al. A single-trial analytic frame-
work for EEG analysis and its application to target detection and
classification. Neuroimage. 2008;42(2):787–798.
54. Wiener N. Extrapolation, Interpolation and Smoothing of
Stationary Time Series. New York: John Wiley & Sons; 1949.
55. Walter DO. A posteriori “Wiener filtering” of averaged evoked
responses. Electroencephalogr Clin Neurophysiol Suppl. 1968;27:
61–70.
56. Doyle DJ. Some comments on the use of Wiener-filtering for the
estimation of evoked potentials. Electroencephalogr Clin Neuro-
physiol. 1975;38:533–534.
57. Albrecht V, Lansky P, Indra M, et al. Wiener filtration versus aver-
aging of evoked responses. Biol Cybern. 1977;27:147–154.
58. Strackee J, Cerri SA. Some statistical aspects of digital Wiener fil-
933
75. Macpherson H, Pipingas A, Silberstein R. A steady state visually
evoked potential investigation of memory and ageing. Brain Cogn.
2009;69(3):571–579.
76. Rees A, Green GG, Kay RH. Steady-state evoked responses to sinu-
soidally amplitude-modulated sounds recorded in man. Hear Res.
1986;23(2):123–133.
77. Picton TW, Skinner CR, Champagne SC, et al. Potentials evoked by
the sinusoidal modulation of the amplitude or frequency of a tone.
J Acoust Soc Am. 1987;82(1):165–178.
78. Picton TW, John MS, Purcell DW, et al. Human auditory steady-
state responses: the effects of recording technique and state of
arousal. Anesth Analg. 2003;97(5):1396–1402.
79. Ross B, Picton TW, Herdman AT, et al. The effect of attention on
the auditory steady-state response. Neurol Clin Neuroph