23rd International Symposium on Transport Phenomena

Auckland, New Zealand

1922 November 2012

Modeling of bioheat equation for skin and a preliminary study on

a noninvasive diagnostic method for skin burn wounds

Shong-Leih Lee* and Yung-Hsiang Lu

Department of Power Mechanical Engineering
National Tsing Hua University, Hsinchu 30013, Taiwan

Abstract dimensionless thickness of hypodermis

dimensionless heat capacity
A three-dimensional vascular model is employed in the present dimensionless thickness of epidermis
study for a unit area of skin covering just one single pair of T reference temperature difference
terminal artery-vein vessels. The tissue temperature and the
dimensionless intensity of a thermal radiation
blood temperatures in artery and vein vessels are solved with a
multi-grid system. A strong countercurrent heat exchange dimensionless temperature
between the bloods inside artery and vein vessels is evidenced. mean temperature of tissue
The mean temperature of the tissue over the cross-section of the dimensionless thermal conductivity
unit skin area is evaluated. The resulting one-dimensional perfusion parameter
function is regarded as the normal tissue temperature valid for
density, ( kg m 3 )
large area of skin in view of the symmetric and periodic structure
of the paired artery-vein vessels in nature. A three-dimensional dimensionless time
bioheat equation then is formulated by superposing the effect of dimensionless temperature
skin burn wound on the normal skin temperature with and dimensionless normal skin temperature
without thermal radiation exposure. Finally, the bioheat pseudo-Biot number
equation is employed to simulate the process of active dynamic b
thermography (ADT) for grade assessment of skin burn wounds. blood perfusion, ( m 3 s 1m 3 )
The decaying factor of the skin surface temperature is found to subscript
decrease monotonically in the self-cooing stage after a thermal a artery
radiation heating. The boundary of grades IIa and IIb in a large a0 secondary artery
burn wound might be detected by tracking the peak of the b blood
gradient of decaying factor within 30 seconds after the thermal c core
radiation is turned off. d dermis
ed epidermis
Nomenclature v vein
Bi Biot number v0 secondary vein
c dimensionless radius of secondary vessel
1.Introduction
cp specific heat, ( J kg 1K 1 )
sweat evaporation rate, ( Wm 2 )
Ev The first attempt to quantitatively describe heat transfer in human
tissue with blood flow effect was presented by Pennes [1]. He
ev dimensionless sweat evaporation rate added a source term b b (c p )b (1 k ) (Ta T ) to the heat
h ambient heat transfer coefficient, ( Wm 2 K 1 ) conduction equation for tissue temperature by considering heat
k thermal conductivity, ( Wm 1K 1 ) transfer from blood to tissue, where Ta and T denote arterial
L reference length or thickness of dermis, (m) blood and tissue temperatures, while b , (c p ) b and b are
m b mass flow rate of blood, ( kg s1 ) density, specific heat and perfusion of blood, respectively. The
q dimensionless heat flux at z 0 equilibrium factor
Qm heat generation by metabolism, ( Wm3 ) Tv T
k (1)
Q tr intensity of thermal radiation, ( Wm2 ) Ta T
qm dimensionless heat generation by metabolism
is a prescribed constant in the range 0 k 1 throughout the
(r , z ) cylindrical coordinates tissue. Pennes [1] assigned k 0 by assuming thermal
rb dimensionless radius of a round burn wound equilibrium between venous blood and tissue ( Tv T ). This is
r0 dimensionless radius of the computational domain the well-known Pennes equation.
T temperature or tissue temperature, (K)
t time, (s) The Pennes equation has gained widespread acceptance ever
since it was published in 1948, although its validity has been
( x, y, z ) Cartesian coordinates
seriously questioned in many applications [2, 3]. One of the
Greek symbols major problems is the countercurrent heat exchange taking place
decaying factor between artery and vein in paired artery-vein vessels. The net
d thermal diffusivity, ( m 2 s 1 ) heat lost to the tissue from the vessel pairs was found to be the
predominant mode of bioheat transfer for vessels of 50 200 m
in diameter [4-6]. This behavior has been confirmed by the
numerical investigation for a branching countercurrent network
in Brinck and Werner [7]. To take this effect into account,
some nonvascular models were proposed to improve the Pennes
equation by altering the equilibrium factor such as the efficiency
function [3] and the correction coefficient [6].
The artery and vein vessels in skin are typically paired. Their
diameters are less than 200 m in general [6]. Hence, the
countercurrent heat exchange in the vessel pairs might be
significant. Henriques and Moritz [8] were the pioneers in the
study of skin burn injury. They found that a first degree burn
occurs when the skin is maintained at a temperature above
44C . Recently, a few mathematical models of heat transfer in
skin have been proposed for some thermal hazards including
contact burn, scald burn, flash fire, and thermal radiation
exposure [9-12]. However, these models are all based on the
conventional Pennes equation without considering the effect of
countercurrent heat exchange in vessel pairs.
Skin consists of three layers, namely epidermis, dermis, and Figure 1. A schematic vascular system of the skin.
hypodermis. The grades of skin burn wound are classified
according to the burn wound depth. Grades I, IIa, IIb, and III
correspond to destruction of epidermis, shallow dermis, deep
equation then is employed to simulate the ADT process on burn
dermis, and full dermis thickness, respectively. The burn
wounds as in the experiment by Renkielska et al. [14]. A new
wounds of grades I and IIa would heal spontaneously within 3
parameter is proposed to estimate the boundary of grades IIa and
weeks of the burn, whereas both grades IIb and III need surgery.
IIb in large burn wounds from the different responses of the
In clinical practice even an inexperienced doctor has no difficulty
normal and burned skins.
in distinguishing grades I and III burn wounds. However,
differentiation between grade IIa and IIb is still problematic.
Accurate prognosis is only 50-70% in clinical evaluation based 2. Three-dimensional vascular model
on visual inspection [13]. 2.1 Governing equation
To assess the burn wound depth more accurately, Renkielska et al. Figure 1 illustrates a schematic vascular system of the skin [6, 7].
[14] developed a noninvasive diagnostic method termed as active The blood vessels (including artery and vein) beneath the muscle
dynamic thermography (ADT) for relatively large burn areas. are known as the primary vessels. The blood circulates between
They conducted an experiment on young domestic pigs (each the primary vessels and the cutaneous vessels by separate riser
weighted approximately 20 kg) in view of the high degree of vessels. The blood enters the secondary artery in the bottom of
functional and structural similarity of pig skin to human skin [15]. the dermis with a temperature Ta 0 approximately the same as
The wounds were inflicted by an aluminum rod which was
applied to the skin at controlled temperature and time. One day that in the primary artery. Next, the blood rises to the top of the
after the burn the ADT experiment was performed with a short dermis by the terminal artery, and then flows to the terminal vein
optical excitation (halogen lamps of 1000W). The optical through a capillary bed (not shown in Fig. 1). Subsequently, the
blood descends to the secondary vein by the terminal vein. The
heating resulted in a surface temperature rise of about 2.5C
blood temperature in the secondary vein Tv 0 is slightly smaller
followed by a self-cooling stage. The skin surface temperature
was assumed to decrease exponentially with a constant time than Ta 0 . There is no blood vessel in both epidermis and
constant during the self-cooling stage. The time constant was hypodermis.
calculated for each of the burn wounds, while the wound depth
was determined by the method of histological analysis [16]. The terminal vessels are roughly 20-40 m in diameter. They
form a pair of countercurrent heat exchanger as mentioned earlier.
Based on the experimental results, Renkielska et al. [14] The spacing between the terminal vessel pairs is the typical
concluded that burn wound having a time constant longer than length of the capillary bed, 500-1000 m . The typical diameter
10.125 s would heal after 3 weeks of burn spontaneously,
otherwise it would be unhealed. This discrimination threshold of the secondary vessels is 50-100 m [6, 7]. In the present
is equivalent to a burn wound shallower or deeper than 60% of study, diameter and spacing of the terminal vessels are assumed
the dermis thickness at the measurement sits. However, of the to be 30 and 750 m , respectively, while the distance between
23 burn wounds investigated in their study, two wounds with the terminal artery and terminal vein is 30 m (60 m center-
similar burn depth (57.5% and 62.1%) had significantly different to-center). The diameter of the secondary vessels is 75 m .
time constants (18.125s and 8.125s). By contrast, two time The average thicknesses of the epidermis, the dermis, and the
constants of essentially the same value (10.200 s and 10.023 s) hypodermis are, respectively, 75, 1500, and 10000 m [12].
corresponded to burn wounds of healed and unhealed with totally
different burn depths (22.0% and 67.8%). Hence, the time Figure 2 shows a simple 3D model for a unit skin area that
constant suggested by Renkielska et al. [14] is not necessarily an contains just one single pair of terminal artery and vein. The
efficient parameter for assessment of burn wounds grades. dimensionless coordinates ( x, y, z ) is normalized with the
thickness of the dermis L 1500 m such that the
The purpose of the present work is to formulate a bioheat
dimensionless thicknesses of the hypodermis and the epidermis
equation with a superposition technique for normal and burned
skins in particular. The effect of countercurrent heat exchange are 6.667 and 0.05 , respectively. Due to symmetry,
in artery-vein vessel pairs is taken into account. The bioheat the computational domain b x b , 0 y b , 0 z 1
 where / n represents the normal heat flux on the tissue
side of the vessel surface. The perfusion parameter is
proportional to m b which is the mass flow rate of blood inside
the blood vessels.

2.2 Boundary conditions

The hypodermis possesses a large thickness and a low thermal
conductivity. It provides a good insulation for the muscle under
the skin. This implies that the temperature inside the dermis is
essentially maintained by the secondary vessels. For simplicity,
the secondary artery and vein are simplified as a flat surface of
width 2c ( c y c ) on the bottom of the dermis ( z 0 ) as
illustrated in Fig. 2. This gives rise to the boundary condition
on the surface z 0

( x, y, 0)
0 for y c
z

( x , y , 0) b 0 for 0 y c (5)
Figure 2. A 3D vascular model for one single pair of terminal where b 0 ( a 0 v 0 ) / 2 is the average temperature of the
artery and vein.
secondary artery and vein. On the skin surface ( z 1 ), the
convective boundary condition is
is employed, where b 0.25 . The radius of the secondary ( x, y, 1 )
vessel is c 0.025 . All of the thermophysical properties in the Bi ( x, y, 1 ) ev 0 (6a)
z
tissue are assumed constant.
h L L Ev
After imposing the assumptions and introducing the Bi , ev (6b)
dimensionless transformation, ked ked T

d t T T Qm L2 where h is the ambient heat transfer coefficient, and Ev is

, , qm (2) the sweat evaporation rate. The boundary conditions in the y-
L2 T k d T
coordinate assume the form
the energy equation inside the tissue of the dermis and the
( x, 0, z ) ( x, b, z )
hypodermis is expressible as 0, 0 (7)
y y

( ) ( ) ( ) qm (3) due to symmetry, while the periodic boundary condition is
x x y y z z
imposed in the x-coordinate
where d and k d are, respectively, the thermal diffusivity (b, y, z ) (b, y, z )
and thermal conductivity of the dermis. Q m is the heat
generation by metabolism. The dermis and the epidermis are ( x 2b, y, z ) if x b
( x, y, z ) (8)
represented by 0 z 1 and 1 z 1 , respectively. The ( x 2b, y, z ) if x b
thermal conductivity is a step function. Its value is unity in
the dermis, and jumps to another constant k ed / k d in the The temperatures on the surfaces of the terminal artery and vein
are, respectively,
epidermis. Similarly, the specific heat jumps from unity in
the dermis to ( c p )ed /( c p ) d in the epidermis. The ( x, y , z ) a ( z ) , ( x, y , z ) v ( z ) (9)
reference temperature difference is assigned as T Ta0 T The initial condition for the artery blood temperature (4a) is
with T being the ambient temperature.
a ( 0) a 0 (10)
Both blood temperatures inside the terminal artery and vein are
Owing to the very tiny vessels, the blood temperature inside the
assumed one-dimensional. They are governed by the total heat
capillary bed is essentially the same as the surrounding tissue at
transfer across the wall of the blood vessels, i.e.
z 1 . Hence, one assigns the initial condition
d a

dz
(
a n
) ds (4a) v (1) (1) (11)

d v
(x, y, z) dx dy
1

dz v
(
n
) ds (4b) ( z)
ad ad
(12)

m b (c p ) b for the blood temperature in the terminal vein (4b), where ad is

(4c) the cross-section area of the dermis.
kd L

Figure 3. (a) Isotherms on the plane y 0 . (b) Isotherms on the
plane z 0.3 .
Figure 4. The mean tissue temperature ( z ) and blood
temperatures in artery a (z ) and in vein v (z ) at 0.1
and 0.2 .
2.3 Numerical method and the results
In the present study, a uniform fine grid system is employed in
the vicinity of the terminal vessels ( 0.05 x 0.05 and
0 y 0.05 ),
xi 0.05 (i 1) x , y j ( j 1) y (13)
where x y 0.001 , i 1, 2, 3, , 101, and j 1, 2, 3, ,
51 after a series of grid tests. To conserve computational effects,
the other coarse grid system with a grid mesh of 5x 5y
x i 0.25 5 (i 1) x , y j 5 ( j 1) y (14)
where i 1, 2, 3, , 101, and j 1, 2, 3, , 51, is used for the
remainder of the physical domain ( 0.25 x 0.25 and
0 y 0.25 ). The temperature of the outer boundary of the
fine grid system at x 0.05 and y 0.05 is provided by the
solution of the coarse grid system. Similarly, the temperature of
the inner boundary of the coarse grid system at x 0.045 and
y 0.045 is determined by the solution from the fine grid
system. Both gird systems have the same gird size in z-
coordinate,
z k (k 1)z (15)
where z 0.01105, k 1, 2, 3, , 96. The interface of the
dermis and the epidermis ( z 1 ) is located at the middle of the
two successive grid points z 91 and z 92 .
The thermophysical properties of skin and blood from the
literature [9, 12] are employed in the present computation, i.e.
k d 0.37 , k ed 0.21 , k hd 0.16 ,
d 1200 , ed 1200 , b 1100 ,
( c p ) d 3200, ( c p )ed 3580 , ( c p )b 3300 (16)
where the units of thermal conductivity, density, and specific heat
are W m 1 K 1 , kg m 3 , J kg 1K 1 , respectively. The
2
sweat evaporation rate is assumed Ev 10 W m , while the
metabolism heat is negligible in skin ( Qm 0 ) [12]. Under
normal condition the blood perfusion rate is
b 0.024 m3 s 1m 3 [12]. It is equivalent to
8 1
m b 2.110 kg s or 0.125 in the present
configuration. The other two dimensionless parameters in Eq.
(6b), the sweat evaporation and the Biot number, would be
ev 0.006803 if the reference temperature difference is
T 10.5C , and Bi 0.05 if the ambient heat transfer
coefficient is h 7 W m 2 K 1 [12].
Based on these parameters, the steady-state solution of the
problem is solved with the conventional central difference
scheme and the SIS solver [17]. The resulting isotherms on the
planes y 0 and z 0.3 are shown in Fig. 3(a) and 3(b),
respectively. In Fig. 3(a) the increments of the isotherms
employed in dermis and epidermis are 0.002 and 0.001,
respectively. From Fig. 3(a), one sees that the blood
temperature in the terminal artery decreases from a (0) 1 to
a (1) 0.959 while that in the terminal vein increases from
v (1) 0.9583 to v (0) 0.994 . This implied a strong
countercurrent heat exchange in the paired artery-vein vessels
through the gap between them as observable from Fig. 3(b).
Figure 4 reveals the mean temperature ( z ) over the cross-
section of the dermis (12) at two representative perfusion
parameters 0.1 and 0.2 . The perfusion parameter
after exercise could be twice as larger as in rest. It is very
interesting to note that the mean temperature of the dermis ( z )
is essentially independent of the perfusion rate.
Figure 5 shows the effect of the Biot number on the mean
temperature of the tissue ( z ) under normal condition
( 0.125 ). The Biot number is seen to have a significant
influence on the tissue temperature. Nevertheless, the mean
temperature of the tissue ( z ) is always a linear function of z
in both dermis and epidermis. The epidermis has a smaller
thermal conductivity ( k ed / k d 0.5676 ) and thus a larger
temperature gradient. For convenience, the mean temperature
of the tissue ( z ) is correlated by

( z) ( z ) (17)

where
q
( z) (1) ed (1 z)( ) for 0 z 1
ed

q
( z) (1) (1 z)( ) for 1 z 1 Figure 5. Influence of Biot number on the mean tissue
ed temperature ( z ) at 0.125.
1 q e
(1) ( )( ) v f1
Bi ed Bi

q Bi f 2 ev by noting that the heat flux q is maintained by the secondary

vessels (see Eq. (18)). In the skin burn wound area ( 0 r rb ),
ed 1 Bi (0.8 ed )
the burned secondary vessels provide no heat energy. The heat
f1 0.01812 0.03995Bi 0.08439Bi2 0.08799Bi3 0.03359Bi4 flux at z 0 would come from the core temperature c (at
z ) beneath the hypodermis. Hence, the boundary
f 2 0.9797 0.2651Bi 0.1411Bi2 0.03627Bi3 (18)
condition is assumed
The correlation is valid for 0 Bi 1 and ev 0.006803 .
The particular function (z ) is the solution of the one- (r,0, ) k hd
( c (r,0, )) , (22)
dimensional problem z kd
d d d (0)
( )0, q , Boundary condition (22) can be regarded as an insulation surface
dz dz dz because the pseudo-Biot number is very small ( 0.065 ) in
d (1 ) general. On the skin surface ( z 1 ), the convective
Bi (1 ) (ev Bi f1 ) 0 (19) boundary condition is employed
dz

3. Burned skin with thermal radiation exposure d

Bi (ev Bi f1 ) 0 (23a)
dz
Consider a round burn wound of radius rb . The bioheat
equation for the skin tissue in and around the burn wound can be L Qtr
(23b)
written as ked T

where Q tr is the intensity of a thermal radiation from some heat

( ) ( ) (20)
r r r r z z source such as halogen lamps. It is noted that Eq. (23a) has
been corrected by adding the term Bi f1 to match Eq. (19)
where ( r , z ) is a dimensionless cylindrical coordinate system
when the mean temperature is used instead of the local
normalized with reference length L. The definition of the mean
temperature for the tissue. Finally, the boundary conditions
temperature of the skin tissue (r, z, ) is the same as that in at r 0 and r r0 are assigned as
the previous section. The formulation is performed in the
region of 0 r r0 and 0 z 1 that might cover
(0, z, ) (r0 , z, )
thousands of terminal vessel pairs. The skin is normal in the 0 , 0 (24)
annulus rb r r0 where the boundary condition should be r r

The axially-symmetrical transient problem (20)-(24) should

(r,0, )
q (21) reduce to the one-dimensional steady problem (19) in the absence
z
 Figure 7. Variation of skin surface temperature after thermal
Figure 6. Isotherms of (r, z, ) for the case of rb 5 , radiation exposure.
Bi 0.05 , c 1.02 , and 0 .
Finally, let the solution revealed in Fig. 6 be the initial condition.
As suggested by Renkielska et al. [14] a thermal radiation of
of burn wound ( rb 0 ) and thermal radiation exposure ( 0 ). intensity Qtr 441Wm 2 is applied on the skin surface
Therefore, one gets the governing equation
(covering the burn wound) for 30.42s to raise the skin surface
temperature by about 2.5C , then turn off the halogen lamps

( ) ( ) (25) and leave the skin cooling-off gradually. The thermal radiation
r r r r z z in dimensionless form is

and the associated boundary conditions 0.3 1.303 0

( ) (28)
0 0
(0, z, ) (r0 , z , )
0, 0
r r
based on the reference time L2 / d 23.4s . The time step
(r , 1 , ) employed in the computation is 0.05 (or 1.17s ). The
Bi (r , 1 , ) ( ) 0
z resulting skin surface temperature is shown in Fig. 7. The burn
wound area ( 0 r rb ) is seen to have a larger temperature
(r ,0, ) decaying factor
0 for rb r r0
z
/
(29)
(r ,0, )
(r ,0, z ) q ( c (0)) for 0 r rb (26)
z

by substituting the superposition

( r , z , ) ( r , z , ) ( z ) (27)

into Eqs. (20)-(24). The unsteady term on the left-hand-side of

Eq. (25) arises from the time-dependent thermal radiation ( ) .
It appears that the solution of Eqs. (25) and (26) is (r , z, ) 0
for normal skin without thermal radiation exposure ( rb 0 and
0 ). Figure 6 shows the resulting isotherms of the solution
(r, z, ) for the case of rb 5 , Bi 0.05 , c 1.02 , and
0 , while the computation domain is truncated at r0 50 .
The burn wound has a low temperature as compared to the
normal skin due to the burned secondary vessels. The minimum
temperature occurring at the center point of the burn wound is
(0, 1 ) 0.76 on the skin surface. It corresponds to 34C
if the reference temperature difference and the ambient Figure 8. Decaying factor of skin surface temperature after
temperature are T 10.5C and T 26C , respectively. thermal radiation exposure.

Figure 9. Variation of / r with time after thermal
radiation exposure for Bi 0.05 and rb 5 .
Figure 10. The ratio rp / rb as a function of time for various
sizes of burn wound.
as compared to the normal skin (see Fig. 8). This finding is
consistent with the experiment [14]. However, the decaying
factor is not a constant. It seems to decrease monotonically
with time (see Fig. 8). Hence, it is not relevant to grade the
burn wounds with the value of (or the time constant
employed in [14]). In the present study, the gradient of the
decaying factor / r is proposed for the grade assessment
of skin burn wounds. Figure 9(a) and 9(b) show the distribution
of / r in the periods of 0 t 30.42s and
30.42s t 152.10s for the case of Bi 0.05 and rb 5 .
From Fig. 9(a) the gradient of the decaying factor is seen to have
a sharp peak at r rp near the boundary of the burn wound
( r p rb ). As time elapses the peak moves left gradually, while
the second peak emerges at t 143s (see Fig. 9(b)). For
convenience, the ratio r p / rb as a function of time is presented
in Fig. 10 for various sizes of burn wound. Figure 10 indicates
that r p is closed to rb at the very beginning of the cooling
stage. Therefore, the burn wound area ( 0 r rb ) might be
detected by tracking the first peak of the decaying factor gradient
of the skin surface temperature ( r rp ) within 30 seconds after
the halogen lamps are turned off.
4. Conclusion
In the present study, a one-dimensional function is obtained for
the tissue temperature of normal skin based on a three-
dimensional solution of a vascular model. Next, a three-
dimensional bioheat equation is formulated by superposing the
effect of skin burn wound on the normal skin temperature.
Finally, the proposed bioheat equation is employed to simulate
the method of active dynamic thermography for grade assessment
of skin burn wounds. Based on the numerical results, the
following conclusions are drawn.
(a) A strong countercurrent heat exchange between the bloods
inside the terminal artery and vein vessels is evidenced.
(b) The three-dimensional bioheat equation proposed in the
present study is able to handle a large area of skin burn wound.
(c) The decaying factor of the skin surface temperature decreases
monotonically in the self-cooing stage after a thermal radiation
heating.
(d) The boundary of grades IIa and IIb in a large skin burn wound
area might be detected by tracking the peak of the decaying
factor gradient of the skin surface temperature within 30 seconds
after a thermal radiation heating.
Acknowledgments
The authors wish to express their appreciation to the National
Science Council of Taiwan for the financial support of this work
through the contract NSC 100-2221-E-007-085.
Figures and Tables
[1] Pennes, H.H., Analysis of tissue and arterial blood
temperatures in the rising human forearm, J. of Applied
Physiology 1 ,1948, 93-122.
[2] Wissler, E.H., Pennes 1948 paper revisited, J. Applied
Physiology 85 ,1998, 35-41.
[3] Brinck, H. and Werner, J., Efficiency function:
improvement of classic bioheat approach, J. Applied
Physiology 77 ,1994, 1617-1622.
[4] Weinbaum, S. and Jiji, L.M., A new simplified bioheat
equation for the effect of flow on local average tissue
temperature, ASME J. Biomechanical Engineering
107 ,1985, 131-139.
[5] Baish, J.W., Ayyaswanty, P.S. and Foster, K.R., Heat
transport mechanism in vascular tissue: a model comparison,
ASME J. Biomechanical Engineering 108 ,1986, 324-331.
[6] Weinbaum, S., Xu, L.X., Zhu, L. and Ekpene, A., A new
fundamental bioheat equation for muscle tissue: part 1
blood perfusion term, ASME J. Biomechanical Engineering
119 ,1997, 278-288.
[7] Brinck, H. and Werner, J., Estimation of the thermal effect
of blood flow in a branching countercurrent network using a
three-dimensional vascular model, ASME J. Biomechanical
Engineering 116 ,1994, 324-330.
[8] Henriques, F.C. and Moritz, A.R., Studies of thermal injury:
the conduction of heat to and through skin and the
temperature therein. A theoretical and experimental
investigation, American J. of Pathology 23 ,1947, 531-549.
[9] Torvi, D.A. and Dale, J.D., A finite element model of skin
subjected to a flash fire, ASME J. Biomechanical
Engineering 116 ,1994, 250-255.
[10] Autrique, L. and Lormel, C., Numerical design of
experiment for sensitivity analysisApplication to skin
burn injury prediction, IEEE Trans. Biomechanical
Engineering 55 ,2008, 1279-1290.
[11] Hirata, A. and Fujiwara, O., Modeling time variation of
blood temperature in a bioheat equation and its application
to temperature analysis due to RF exposure, Physics in
Medicine and Biology 54 ,2009, N189-N196.
[12] Ng, E.Y.K., Tan, H.M. and Ooi, E.H., Boundary element
method with bioheat equation for skin burn injury, Burns
35 ,2009, 883-893.
[13] Heimbach, D., Engrave, L., Grube, B. and Marvin, J., Bern
depth: a review, World J. Surgery 16 ,1992, 10-15.
[14] Renkielska, A., Nowakowski, A., Kaczmarek, M. and
Ruminski, J., Burn depth evaluation based on active
dynamic IR thermal imaginga primarily study, Burns
32 ,2006, 867-875.
[15] Meyer, W., Schwarz, R. and Neurand, K., The skin of
domestic mammals as a model for human skin, with special
reference to domestic pig, Curr. Probl. Dermatol. 7 ,1978,
39-52.
[16] Singer, A.J., Berruti, L., Thode, H.C. and McClain, S.A.,
Standardized burn model using a multiparametric histologic
analysis of burn depth, Acad. Emerg. Med. 7 ,2000, 1-6.
[17] Lee, S.L., A strongly-implicit solver for two-dimensional
elliptic differential equations, Numerical Heat Transfer B
16 ,1989, 161-178.