You are on page 1of 4

SEGMENTATION,MEDIAN FILTERING AND MORPHOLOGICAL

OPERATION BASED BRAIN TUMOR DETECTION USING


BACKGROUND LIKELIHOOD MODELS AND NEURAL NETWORK
INTRODUCTION
DIGITAL IMAGE PROCESSING
Digital Image Processing (DIP) refers to processing of digitized image through
computers and finds varieties of applications in the most diverse areas of business, science and
technology. Any typical image processing application comprises of extraction of important
features of a given image, using which later description, interpretation, or understanding of an
image is provided. DIP techniques are used today in solving varieties of problems related to
medical, office and industrial automation, remote sensing, science, criminology, image
transmission and storage, astronomy, space meteorology, information technology, entertainment,
consumer electronics, printing, graphics arts and defense. Digital images are captured through
imaging devices and cover almost the entire frequency spectrum, ranging from gamma to radio
waves. The images generated by these devices include ultrasound, MRI, electron microscopy
etc.
PURPOSE OF IMAGE PROCESSING
The purpose of image processing is divided into 5 groups.
They are:
Visualization - Observe the objects that are not visible.
Image sharpening and restoration - To create a better image.
Image retrieval - Seek for the image of interest.
Measurement of pattern Measures various objects in an image.
Image Recognition Distinguish the objects in an image.

BRAIN TUMOR
A tumor is an abnormal growth of tissue and this is the most common solid tumors in children.
Approximately 1500 children in the US are diagnosed with a brain tumor each year. Brain
tumors, either malignant or benign, are tumors that originate in the cells of the brain. A benign
tumor does not contain cancer cells and usually, once removed, does not recur. Most benign
brain tumors have clear borders, meaning they do not invade surrounding tissue. These tumors
can, however, cause symptoms similar to cancerous tumors because of their size and location in

the brain.7 Malignant brain tumors contain cancer cells. Malignant brain tumors are usually fast
growing and invade surrounding tissue. Malignant brain tumors spread to other areas of the body
and may recur after treatment. Sometimes, brain tumors that are cancer are called malignant
because of their size and location , and the damage they can do to vital functions of the brain.
Brain tumors can occur at any age. Brain tumors that occur in infants and children are very
different from adult brain tumors, both in terms of the type of cells and the responsiveness to
treatment.

BLOCK DIAGRAM EXPLANATION


The major blocks of the proposed system are input, preprocessing, threshold segmentation
,feature extraction , postprocessing( morphological operation),edge detection,feature
training,neural network classification and simulation output. A MRI scan image is taken as an
input for the proposed system. In the preprocessing state , the noise factors in the form of salt
and pepper noise are removed. The features are then extracted followed by morphological
operations . After removal of noises by erosion and dilation techniques the features are trained.
Finally with the help of artificial neural networks the features are utilized to classify the stage of
the tumour .

NEURAL NETWORK
A neural network for handwriting recognition is defined by a set of input neurons which may be
activated by the pixels of an input image. After being weighted and transformed by
a function (determined by the network's designer), the activations of these neurons are then
passed on to other neurons. This process is repeated until finally, an output neuron is activated.
This determines which character was read.
neural networks have been used to solve a wide variety of tasks that are hard to solve using
ordinary rule-based programming, including computer vision and speech recognition.

NEURAL NETWORK CLASSIFICATION


A suitable artificial neural network classifier is designed in this paper to identify the different
grades of brain tumors. Artificial neural networks are composed of simple elements operated in
parallel. These elements are inspired from biological nervous system. Each element in a network
called neuron [4-5]. The sum of multiplication of weights and inputs plus bias at the node is
positive then only output elements fires. Fire means it discharges energy to next element.
Otherwise it doesnt fire.

The artificial neural network is an adaptive system [10-11]. Adaptive means system parameters
are changed during the operation. The system parameter is nothing but weights. Two layer feed
forward neural network is taken in this paper .The two layer feed forward neural network
consists of one input layer and one output layer and one hidden layer and one output. In the
hidden layer 10 nodes are taken .In the two layer feed forward network two log sigmoid transfer
function are used. The two layer feed forward network with two log sigmoid functions are more
widely used in classification, pattern recognition .It gives better results in these classification. If
the sum of multiplication of weights and input values are greater than log sigmoid function then
output value becomes 1 , otherwise the output value become 0

LIKELIHOOD MODELS
In statistics, a likelihood function (often simply the likelihood) is a function of the parameters of
a statistical model. Likelihood functions play a key role in statistical inference, especially
methods of estimating a parameter from a set of statistics.
The likelihood of a set of parameter values, , given outcomes x, is equal to the probability of
those observed outcomes given those parameter values, that is
.
The likelihood function is defined differently for discrete and continuous probability
distributions.

You might also like