Journal of Thermal Biology 38 (2013) 107125

Contents lists available at SciVerse ScienceDirect

Journal of Thermal Biology

journal homepage: www.elsevier.com/locate/jtherbio

Review

Conventional and newly developed bioheat transport models in vascularized

tissues: A review
Arka Bhowmik a, Rupesh Singh b, Ramjee Repaka a, Subhash C. Mishra b,n
a
School of Mechanical Materials and Energy Engineering, Indian Institute of Technology Ropar, Rupnagar, Punjab 140001, India
b
Department of Mechanical Engineering, Indian Institute of Technology Guwahati, Guwahati 781309, India

a r t i c l e i n f o abstract

Article history: Heat transfer in a biological system is a complex process and its analysis is difcult. Heterogeneous
Received 22 October 2012 vascular architecture, blood ow in the complex network of arteries and veins, varying metabolic heat
Accepted 20 December 2012 generation rates and dependence of tissue properties on its physiological condition contribute to this
Available online 3 January 2013
complexity. The understanding of heat transfer in human body is important for better insight of
Keywords: thermoregulatory mechanism and physiological conditions. Its understanding is also important for
Bioheat equation accurate prediction of thermal transport and temperature distribution during biomedical applications.
Blood perfusion model During the last three decades, many attempts have been made by researchers to model the complex
Countercurrent model thermal behavior of the human body. These models, viz., blood perfusion, countercurrent, thermal
Dual-phase-lag model
phase-lag, porous-media, perturbation, radiation, etc. have their corresponding strengths and limita-
Radiative model
tions. Along with their biomedical applications, this article reviews various contextual issues associated
with these models. After brief discussion of early bioheat models, the newly developed bioheat models
are discussed in detail. Dependence of these models on biological properties, viz., thermophysical and
optical properties are also discussed.
& 2013 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Heat generation and transfer in human body. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
3. Early bioheat models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.1. The blood perfusion based bioheat models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.2. The countercurrent bioheat models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3.3. The statistical bioheat model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4. Recently developed bioheat models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.1. The dual-phase-lag bioheat model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.2. The porous media based bioheat model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.3. The discrete vascular bioheat model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.4. The Deng and Liu perturbation bioheat model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.5. The newly developed countercurrent vascular model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.6. The Shrivastava and Vaughan perfusion based bioheat model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.7. The radiation based bioheat models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.7.1. The surface radiation phenomenon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.7.2. The volumetric radiative model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.8. The combined continuum and radiative model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5. Thermophysical and optical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

n
Corresponding author. Tel.: 91 36125 82660; fax: 91 3612 690762.
E-mail address: scm_iitg@yahoo.com (S.C. Mishra).

0306-4565/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jtherbio.2012.12.003
108 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
1. Introduction
Human body is a thermal system. It generates heat by various
biochemical reactions at cellular level, and maintains its tem-
perature within a short-range using various thermoregulatory
mechanisms. The overall thermal behavior of the human body is
the combined effect of heat generation and heat transfer at
various scales, viz., microscale (molecular or cellular level),
mesoscale (tissue and organ level) and macroscale (thermal
interaction of the complete body with environment). Under-
standing the thermal behavior at these length scales is essential
in many applications that include therapeutic, diagnosis, bio-
chemical, thermal, environmental and preservation sciences. The
therapeutic application includes hyperthermia treatment
(Overgaard, 1977; Yue et al., 2011), brain hypothermia resuscitation
(Varon and Acosta, 2008), cryosurgery (Gage and Baus, 2002;
Takeda et al., 2009), laser surgery (Van-Hillegersberg, 1997), laser
coagulation (Heisterkamp et al., 1999) and estimation of burn
injury (Despa et al., 2005). The non-invasive thermal (Barnes,
1968) and optical techniques (Wnek and Bowlin, 2005) to identify
the structural and functional abnormalities of tissues at mild
temperature range come under diagnostic applications. The bio-
chemical engineering applications pertains to the study of thermo-
dynamics of protein folding (Baker and Agard, 1994; Ruell et al.,
2009) and DNA folding (SantaLucia and Hicks, 2004), heat transfer
for rapid cycle DNA amplication (Wittwer et al., 1990), heat
inactivation to avoid thermal denaturation of macro-molecules
(Lindahl, 1967), thermodynamics of normal and abnormal cellular
metabolism (Toussaint and Schneider, 1998), etc. Similarly, thermal
engineering applications include study of human thermoregulation
(Gisol and Wegner, 1984; Gonzalez et al., 2010) and comfort
systems (Dear and Brager, 1998; McLeod et al., 2009). The study of
biological heat transfer in extreme environments (i.e., in space,
ocean, cryonics and elevated conditions) belong to the environ-
mental science applications. Bio-preservation and food preservation
come under the preservation science applications.
During the last three decades most of the studies in the eld of
bioheat transfer have been limited to mesoscale or tissue level.
The heat transfer phenomena in deep body and close to the outer
skin are quite complex. These have been reviewed by many
authors (Chato, 1981; Baish et al., 1986; Charny, 1992; Arkin
et al., 1994; Rubinsky, 1999;Khaled and Vafai, 2003). These
reviews are limited to mathematical modeling of early continuum
and vascular bioheat models. Specically, most of the works are
based on simplied continuum and vascular models.
The nature of heat transfer depends on multiple factors such as
active/inactive body conditions, environmental conditions, thermo-
physical and optical properties of tissues and bio-uids, embedded
vascular geometry and patterns in tissues, physiological and patho-
physiological changes, etc. A literature that discusses various
biological heat transfer models in vascularized tissues, viz. con-
tinuum, vascular and radiative models along with their applications
in biomedical engineering is scarce. Thus, the present article is an
attempt to provide an insight on vascular, radiative and recently
introduced models, viz., porous-media, perturbation, dual-phase-
lag models apart from much discussed conventional continuum and
vascular models.
The article starts with a brief description of heat generation and
heat transfer mechanism in the human body. Early and newly
developed bioheat models are reviewed next. The dependence of
these models on thermophysical properties, viz., density, specic
heat, thermal conductivity and blood perfusion rate, and optical
properties, viz., absorption coefcient, scattering coefcient and
anisotropic scattering phase function along with associated changes
in these properties with physiological conditions are discussed at
the end.
2. Heat generation and transfer in human body
Since early days of medical practice, any irregular thermal
deviations observed in the human body are indicative of abnor-
mal physiological condition. These conditions can be identied by
proper understanding of heat generation and heat transfer
mechanisms in the human body. Like any other thermal system,
thermal interaction of human body is due to conduction, convec-
tion and radiation.
Fundamentally, all living systems need continuous supply of
energy for various processes such as biosynthesis of molecules,
transport of signals in the form of molecules and ions across the
cellular boundaries, and to perform mechanical work for cellular
movement and muscle contraction. The energy required for
various mechanical and chemical activities are derived from
oxidation of food and light. The human digestive system breaks
down the food into smaller particles by various physical and
chemical means, which are absorbed as individual molecules by
intestinal walls. Further, blood vessels distribute the correspond-
ing molecules throughout the body, to be taken up by cells,
according to the need such as carbohydrate molecules for energy,
amino acid molecules for building proteins and lipid molecules
for building membranes and hormones.
The carbohydrate molecules are consumed by cells as simple
sugar, viz., glucose, fructose or galactose. Similarly, lipid mole-
cules are consumed by cells as fatty acids and glycerol. The
glucose and fatty acid molecules are oxidized to carbon dioxide
and water by various catabolic reactions in the cells, which
release useful energy. Some part of this useful energy is used to
reform ATP molecules which can be further used as an energy
capsule for anabolic reactions such as biosynthesis of glucose,
lipids, DNA and various proteins. The remaining part of the useful
energy is converted into work. These ATP molecules are reformed
within the mitochondria in huge amount by combined effort of
citric acid cycle and electron transport chain. Heat is generated
within the cell during catabolic reactions as well as during
anabolic reactions. The detailed description of these biochemical
events which will give actual molecular reactions can be obtained
in the literature of biochemistry (Stryer, 1988).
The metabolic heat generated in the tissues, due to various
cellular activities under normal and abnormal conditions, is either
lost to environment and/or transferred to other locations by
combination of several mechanisms, viz. conduction, convection
and radiation. For a body to be in thermal equilibrium, the rate of
heat gain and the rate of heat dissipation should be the same.
Otherwise, the excess heat leads to a change in the local tempera-
ture of the body. According to Gisol and Wegner (1984), the heat
diffusion and heat convection are not sufcient to release the basic
metabolic rate to surroundings. Radiation is equally a potential
bioheat transfer mechanism for thermoregulation. Moreover, the
ability of the human body to maintain the core temperature
(37 1C) for proper functioning and survival is only possible if
heat diffusion, convection and radiation act simultaneously.
The blood ow through complex vascular network of arteries,
veins and capillaries of different sizes regulates the local tem-
perature of tissues deep inside the body as well as close to the
skin. Blood ow through vessels acts as local thermal regulator
during hypothermia, hyperthermia and changing environmental
conditions. The thermal interaction between blood vessels and
tissues depends on various thermophysical and transport proper-
ties such as density and specic heat of tissue and bio-uids,
thermal conductance between tissues of vessel walls and body
parts, thermal conductivity of blood and vessel tissues, volumetric
perfusion rates and various convective properties. Kutz (2002)
and Xu et al. (2009) have provided detailed description of bioheat
transfer and biothermomechnics.
 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
The consideration of thermal radiation in bioheat transfer is
equally important. Owing to absorption, emission and scattering,
transport of thermal radiation in human body is a volumetric
phenomenon. The intensity in a particular direction while passing
through the participating biological tissue, viz., blood, blood vessels
and tissues, is either subjected to attenuation by absorption and
out-scattering or augmentation by emission and in-scattering from
other directions. The interaction of radiation with biological media
depends on optical properties of the tissues and bio-uids, and thus
consideration of various optical properties is important for correct
analysis of heat transfer in a biological system.
3. Early bioheat models
Bioheat models provide quantitative analysis of complex
biological heat transfer phenomenon. These models play an
important role in understanding various heat transfer phenomena
in the human body, and are classied as continuum and vascular
models. The continuum models are simplied form of bioheat
equations. In these models, the effect of blood ow is averaged
over the control volume. The blood ow through each individual
vessel is ignored. On the other hand, the vascular models are
complex forms of bioheat equations which accounts for blood
ow in each individual vessel. This enables the prediction of
temperature distributions along the directions of interest. These
models have remained useful for numerous applications till date.
In the following section, briey discusses some of the early
continuum and vascular bioheat models developed before 1995.
3.1. The blood perfusion based bioheat models
In mid 20th century, Pennes (1948) developed a bioheat
transfer equation based on the experimental analysis of human
forearm. Pennes bioheat equation is a modied form of the
transient heat conduction equation including the effects of blood
ow and metabolic heat generation rate. It is one of the oldest and
simplest blood perfusion models, which have been used for
numerous biological and medical applications. The model is
widely used to estimate the heat transport as well as temperature
variation within biological tissues and organs in various thera-
peutic applications, viz., hyperthermia (Roemer and Cetas, 1984),
brain hypothermia (Zhu and Diao, 2001), tissue ablation (Jaunich
et al., 2008), thermal therapy (Okajima et al., 2009; Gupta et al.,
2010), cryosurgery (Rossi and Rabin, 2007), characterization of
tumor (Das et al., 2012), etc. Pennes assumed that the heat
transfer occurs only in the porous capillary beds. His assumption
is based on the fact that the microscopic capillaries distributed
blood to various tissues where the transfer of nutrients, oxygen
and waste products takes place. As a result, the densely distrib-
uted capillaries provide larger area for heat exchange between
blood and tissues. The temperature of venous blood leaving
tissues and capillaries depends on the extent of heat transfer
needed to achieve thermal equilibrium. Thus, Pennes also intro-
0
duced a thermal equilibration parameter k to correlate tempera-
tures of the arterial blood and the venous blood and tissue.
Pennes model, originally developed for the measurement of
thermal changes in the resting human forearm, is given as
   
@T t 0 000
rt ct rUkt rT t 1k ob cb T a T t  qm 1 the solid tissue surrounding it decreases by an exponential factor of
@t
where t is the time and T is the temperature, and with subscripts
t, b, a and m standing for tissue, blood, arterial blood and
metabolism, respectively, r is the density, c is the specic heat,
k is the thermal conductivity, o is the blood perfusion rate per
0
unit tissue volume, k is the thermal equilibration parameter
109
000
which is either 0 or 1 and qm is the rate of volumetric heat
generation rate due to metabolism.
Over the years, this model has been criticized, reinvestigated
and modied by many investigators because of the various
assumptions made by Pennes. Nevertheless, many important
investigations in bioheat transfer are based on the mathematical
analysis of this model.
In 1974, Wulff developed an energy conservation equation and
identied four fundamental shortcomings in Pennes model
(Wulff, 1974). He pointed out that combining the local effects
(energy storage, heat diffusion and metabolic heat generation
rate) with global effects (convection due to blood perfusion) in a
control system in Pennes model is not appropriate. Secondly, he
identied that in the Pennes equation, it is also wrong to consider
three temperatures, viz., tissue temperature, inlet and outlet
blood stream temperatures, simultaneously at the same point in
the tissue. Thirdly, Pennes model failed to incorporate the fact
that all intra- and extra- cellular bio-uids are in motion and are
capable to convect heat in any direction. Lastly, the heat transfer
between stationary tissue and bio-uids should be modeled based
on temperature difference between these two media, rather than
between blood stream temperatures. Basically, Wulff introduced
the convection heat transfer term and he suggested that the blood
ow contribution must be modeled by a directional convection
term rather than the scalar perfusion term. The Wulffs energy
conservation equation is given as
 
@T t  
! !
rt ct rUkt rT t  rb cb v h UrT b rb v h DHf rc 2
@t
Z
1  
rb hb !
vh rb hb v dO 3
4p
O
where DHf is the specic enthalpy of metabolic reaction, c is the
!
extent of reaction, v h is the Darcy velocity or local mean
apparent blood velocity, v is the actual blood velocity in the
capillary, O 4p is the spherical solid angle and hb is the specic
enthalpy of blood consisting of both sensible enthalpy and
enthalpy of metabolic reaction. Further, Wulff suggested that
thermal interaction between the blood and the tissue is quiet
efcient in porous media which maintains equilibrium with
surrounding tissues. As a result, he considered that temperature
of the blood is equivalent to tissue temperature Tt. Besides that,
!
he assumed metabolic reaction term rb v h DHf rc is equivalent
to volumetric heat generation. Altogether, Wulffs bioheat model
emphasized more on the directional nature of heat convection
due to blood perfusion. The observed limitation of this equation is
that the tissue and the blood volume are needed to be in thermal
equilibrium, which is not the case everywhere.
Consequently, in 1980 Chen and Holmes (1980) suggested that
the perfused bloodtissue continuum must be modeled as two
separate sub-continuum based on length scale analysis. First one is
the tissue sub-continuum which includes scales not larger than one
millimeter (such as tissues and smaller blood vessels) and the second
one is the vessel sub-continuum which includes larger vessels. Based
on the analysis of thermal equilibration length, their investigation led
to an important concept of thermally signicant blood vessels. They
dened the thermal equilibration length as the length of the blood
vessel over which the temperature difference between the blood and
its initial value. Later analysis of the equilibration length on the basis
of geometrical parameters and ow data suggested that the diameter
of the vessel close to 175 mm have a comparable thermal equilibra-
tion length with the actual vessel length. Based on thermal equilibra-
tion length, they classied the blood vessels into thermally signicant
vessels and thermally insignicant vessels. Their observation revealed
110 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
that thermal equilibrium between blood and tissue is achieved
mostly in medium sized blood vessels (i.e. diameter 50300 mm)
instead of smaller blood vessels (i.e. less than 50 mm diameter as in
capillaries). Thus, the tissue sub-continuum model given by Chen and
Holmes is written as
 
@T t    !
rt ct rU kef f rT t onb rb cb T na r in T t rb cb v UrT t q00m
@t
4
kef f kt kp
5 Arterial flow : 4m
!
where r in is blood inlet position under consideration, v is blood
velocity, kef f is effective thermal conductivity and kp is the thermal
conductivity due to blood perfusion. Chen and Holmes model is thus
an extension of Pennes model. It includes more precise terms like
directionality effect and convective heat transfer effects. The second
term on the right hand side of Eq. (4) has an appearance of blood
perfusion term in Pennes model, with few changes in the assump-
tions. In the present model, the arterial temperature is not the core
temperature, rather T na r in is the temperature of arterial blood in the
largest artery included in the continuum model. Also, the blood
temperature of pre-capillary arterioles and venules are in thermal
equilibrium with surrounding solid tissue temperature Tt. Moreover,
onb is the blood perfusion rate for blood vessels with a diameter
smaller than the arterial blood vessel.
The third term on the right hand side in Eq. (4) is a typical
expression for convection through a porous media which repre-
sents capillary ow. For vascular sub-continuum model, to
capture heat transfer due to ow through individual blood vessels
to surrounding tissues, Chen and Holmes suggested that larger
vessel should be analyzed on individual basis.
3.2. The countercurrent bioheat models
The blood perfusion based bioheat models discussed above
missed one important feature that is the heat transfer due to
juxtaposed arteryvein pairs during the counter ow. Mitchell
and Myers (1968) are the rst one to suggest a bioheat model
describing the countercurrent heat exchange between arteries
and veins of the limbs, and with environment for the extremity of
an animal. They described that the thermal energy balance
between outer skin and environment is achieved by counter-
current heat exchange between juxtaposed blood vessels, and due
to energy ow to and from the environment by convection,
radiation and evaporation. They pointed out that the limbs are
signicant contributor for thermoregulation and are maintained
at steady temperature at different environmental conditions by
thermal energy exchange between counter-current blood ow
and metabolic heat. The mathematical model proposed by
Mitchell and Myers is based on energy conservation principle
for steady-state vascular ow system consisting of juxtaposed
arteryvein pair, which includes six prior assumptions (Mitchell
and Myers, 1968). Their model is only applicable for simplied
vessel patterns where ow velocity is insignicant. Moreover, the
assumption of constant blood ow rate for countercurrent vessel
has restricted their model to be applied for other larger main
supplying vessels.
Consequently, Keller and Seiler in 1971 suggested a model for
peripheral heat transfer in subcutaneous fat region for human
body (Keller and Seiler, 1971). They divided their region of
interest, i.e., subcutaneous region into constant temperature core
and peripheral region with the assumption that heat transfer is
signicant in the direction normal to the skin. Their model
included the effect of heat conduction in the tissue and blood
perfusion in the capillaries apart from the countercurrent heat
exchange between the blood vessels, viz., artery and vein pair as
reported by Mitchell and Myers (1968). Their model coupled an
energy conservation equation for the vascularized tissue (Eq. (6))
with the arterial and venous ow energy equations. These
equations are given as
!
2
d T  
Vascularized tissue : kt
000
hA000 cb ob T a T  hA000 T v T  qm 0
dx2
6
2 3
Zx  
dT a
_ a 0  ob dx5cb hA000 T a T  0 7
dx
0
2 3
Zx  
dT v  
Venous flow : 4m
_ v 0  ob dx5cb hA000 cb ob TT v  0
dx
0
8
where x is the distance along the direction normal to skin surface,
h is the average heat transfer coefcient from vessels to surround-
ing tissues, A000 is the average area for heat transfer per unit volume,
ob is the blood perfusion rate, m_ is the blood ow rate, d is the
thickness of tissue layer, subscripts 0, c and t represents inlet
condition, constant core temperature and tissue, respectively.
Later, based on the anatomical study of vascular cast for rabbit
thigh, Weinbaum et al. (1984a) suggested a detailed three-layer
theoretical model for tissuevessel thermal interaction due to coun-
tercurrent blood ow. They developed three models for the three
layers (Weinbaum et al., 1984b). This was followed by ultrastructural
anatomical study and high spatial resolution temperature measure-
ment (Weinbaum et al., 1984a). These are: (i) deep tissue layer, (ii)
the intermediate layer and (iii) the cutaneous layer. The deep tissue
layer was modeled as isolated countercurrent large vessels in tapered
tissue cylinder. The intermediate layer was modeled as counter-
current transverse terminal vessel pair. The cutaneous layer was
modeled as single or pair of vessels in the plexus near the skin. The
purpose of their model is to formulate quantitative equations based
on physiological parameters such as vessel number density, size, ow
velocity and direction of arterial-venous arcade in peripheral tissues
for local energy exchange in vascularized tissues. Observations and
theoretical study by Weinbaum et al. (1984a) revealed that the heat
exchange mechanism between the paired vessels in deep tissue layer
is mainly due to countercurrent ow. That is, the heat is exchanged
both through normal conduction and capillary bleed-off within the
tissue space between the paired vessels in the counter ow. Their
experimental observation revealed that the local tissue temperature
uctuates around a mean temperature whose length scale is small
compared to the distance between the centers of arteryvein pair.
Further, they assumed that the small diameter transverse vessels in
the intermediate layer are thermally insignicant vessels. They
suggested that the radial as well as normal heat conduction and the
capillary perfusion of blood between transverse terminal arteries and
veins are signicant. According to them, the blood circulation through
supercial arteries and veins in the deeper cutaneous plexus releases
most of the heat in the cutaneous layer. The complete solution to the
three-layer model is achieved by coupling the three distinct equations
and the common boundary conditions of adjacent layers. More details
on the theoretical models and analytical solution for the coupled
three-layer boundary value problems can be obtained from
Weinbaum et al. (1984b). Their observation during temperature
measurements and the results of detailed three vascular models
revealed that temperature differences is mainly due to incomplete
countercurrent energy exchange in the thermally signicant vessels.
Later, in 1985 Weinbaum and Jiji realized the difculties of
applying their three-layer vascular model to practical situations
and suggested a simplied bloodtissue continuum model
(Weinbaum and Jiji, 1985). This model includes the effect of
 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
blood ow and vascular architecture on heat transfer, which is
based on anatomical observations and measurements as dis-
cussed earlier. They suggested that vascularized tissue behaves
as an anisotropic heat transfer medium and the isotropic blood
perfusion term in the existing bioheat equations is negligible
compared to energy exchange due to countercurrent blood ow.
Further, they suggested that complete heat exchange is only
possible if the heat leaving from the thermally signicant arteries
is transferred to its counter ow blood through thermally sig-
nicant veins and vice versa without releasing to surrounding
tissues. Presence of any tissue temperature gradient along the
vessel axis due to normal and diseased conditions leads to
different amounts of heat transfer from arteries and veins to
tissues in a direction perpendicular to the axis. Thus, they
assumed that most of the heat conducted away through the wall
of arteries is conducted in through the wall of its paired veins and
vice-versa. The mathematical model suggested by WeinbaumJiji
considers an effective thermal conductivity which incorporates
the effect of blood ow as well as vascular architecture. This is
given as
 
 
 model the diffusion of angiogenic factors in real tissues to
  @T t @   @T t p2 ak2b Pe @
rc kij ef f  nali Pe ngPe li
@t @xi @xj 4kx @xi
where kij ef f is effective thermal conductivity which is a function
of the local Peclet number (Pe), blood vessels geometrical loca-
tions and architecture. The second term on the right hand side of
Eq. (9) is due to convective tapering effect that results from the
spatial variation of vessel diameter and Reynolds number of the
countercurrent vessels. The third term on right hand side of
Eq. (9) is due to directed capillary perfusion between the counter-
current artery and vein. The effective thermal conductivity kij ef f
in Eq. (9) is given by
" #
2
  np2 a2 kb li lj Pe2
kij ef f k dij 2
10
4xk
where rc is the average bloodtissue heat capacity, k is the
thermal conductivity, kij is the thermal conductivity
 of anisotropic
 common boundary condition at the vessel walls. The steady-
material, Tt is the mean tissue temperature T a T v =2 , a is the
artery radius, v is the average blood velocity in countercurrent
artery or vein, n is the number density of vessel pairs crossing
surface of control volume per unit area, Pe is the Peclet number
 
PrRe 2rb cb av=kb , g is the volumetric blood bleed-off rate per
unit vessel surface area leaving arterial ow or entering venous
ow, li and lj are direction cosines, dij is the Kronecker delta, x is
the position vector, subscript i represents coordinate, i.e. (i1, 2,
3), subscript j is the summation index, x is the conduction shape
1
factor p=cosh ls=2as, l is the spacing between centers of
neighboring vessels and s is the axial direction of vessel. Further,
the convective term on the right hand side of Eq. (9) vanishes
entirely if the countercurrent vessels were straight since @li =@xj
would be zero and the model reduces to a single differential
equation for the mean tissue temperature. This model was under
criticism because of the simplifying assumptions such as (i) the
tissue temperature is the average of the arterial and venous
temperatures and (ii) the heat transfer between countercurrent
vessels depends on (Ta Tv) with no inuence of tissue tempera-
ture. Moreover, this model demands detailed anatomical data
compared to other models which are commonly not available and
thus require prior knowledge of the anatomy.
3.3. The statistical bioheat model
Baish in 1994 discussed the associated problems with the
available continuum and vascular models, and he proposed a
111
statistical model for vascularized tissue (Baish, 1994). He sug-
gested that vascular model which explicitly considers a few
vessels, is not appropriate because in reality, in the vicinity of
the vessels considered, there are other signicant vessels of about
the same dimensions. On the other hand, the continuum model
that provides averaged information of the collective blood ow to
predict the tissue temperature fails to accurately assess the
convective effect. Baish pointed that the problem with explicit
vascular model and collective continuum model is that the blood
vessels are arbitrarily dened in the vascularized tissue. Due to
the difculty of exact modeling of complex vascularized tissue,
Baish (1994) suggested a steady-state heat transport model for all
large vessels in the tissues which combined the convection model
of blood ow with the conduction model of extravascular tissues.
The most important consideration of the model is that the
temperature eld in vascularized tissue is determined by statis-
tical interpretation. This model includes both deterministic and
statistical approaches, and has three key elements:
(a) A numerical algorithm (i.e. Gottleib algorithm) is used to
2 @T t
q00m generate a complete vascular tree that is geometrically
v @xj
similar to a real vascular tree. In the beginning of numerical
9 simulation, he used a vesseltissue growth algorithm that
models vesseltissue volume in a realistic way and accounts
for the size, orientation, number density, and branching
characteristics of blood vessels. Also, it models growth of
different kinds of tissue cells in the vicinity of vessels.
Actually, the growth algorithm models vascularized tissue
based on details of initial vessel tree and grid of tissue cells,
the threshold distance and the rate of growth per generation.
More details about this algorithm are available in Baish
(1994).
(b) The rate of heat ow into each vessel, temperature at every
point along the vessel and temperature at any point in the
tissue are determined by both analytical and numerical
approaches. The model combines the heat transfer in tissue
sub-volume with that of the blood sub-volume through their
state thermal energy equation for the tissue sub-volume and
the vessel sub-volume are given as
rUkt rT t r St r 0 11
rUkb rT b rrb cb vrUrT b r Sb r 0 12
where r is the position relative to the center of the organ, St
and Sb are the source terms for tissue and blood, respectively
representing the rate of volumetric heat generation. Consid-
ering the mixing cup blood temperature T b x, the energy
equation for vessel the sub-volume is further modied as
dT b x
_ i cb
m q0b x 13
dx
q0b x pNukb T rxT b x 14
where m _ i is the mass ow rate of blood, x represents coordinate
along the axis of a vessel, q0b is the rate of heat ow per unit
length in a vessel, T rx is the temperature of the vessel wall
and Nu is the Nusselt number. Further, the conservation of mass
and energy at vessel junction are given as
X
m_i m_j 15
j
X
_ i T bi
m _ j T bj
m 16
j
112 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125

where j represents daughter of vessel i, T bi is the temperature of experimentally established by researchers that biological tissue
blood leaving the junction and T bj is the temperature of blood with inner structural non-homogeneity exhibits non-Fourier heat
moving towards the junction from daughter vessels. The mass conduction behavior characterized by temperature oscillation
transport across the vessel wall is considered negligible, and a (Roemer et al., 1985) and wave-like behavior (Mitra et al., 1995)
heat sink that convects heat away from the tissue sub-volume with nite speed of propagation. In other words, there will be a
lies along the centerline of the vessel. At each location along the time lag between cause and effect, if a thermal disturbance is
vessel, the heat ow per unit length into the vessel is matched imposed on the tissue. The generalized form of Fouriers law of heat
with the strength of the heat sink. The analytical solution of conduction equation in tissue is given as
Eq. (11) obtained using Greens function for an innite domain is qt r,t kt rT t r,t 18
given as
Z Z where qt is the tissue heat ux vector due to conduction. The
TrT ref Sr 0 Gr,r 0 dV q0b r 0 Gr,r 0 dx 17 thermal lag behavior in mathematical form can be given by either
r0 A V organ r0 A tree
(i) CV constitutive equation named after Cattaneo and Vernotte
  (Cattaneo, 1958; Vernotte, 1958) or (ii) dual-phase-lag hyperbolic
where G(r,r0 ) is the Greens function 1=4pkt 9rr 0 9 and r0 is conduction equation. In fact, the CV constitutive relation and
the new position. Further, the rate of heat generation S is Fouriers law are special cases of dual-phase-lag model. The dual-
considered the same in both sub-volumes. For numerical evalua- phase-lag (Tzou, 1995a, 1995b) constitutive relation is used for
tion, the energy equations for tissue and blood sub-volumes are describing heat conduction involving high heat ux and strong
discretized with the assumption that the rate of heat absorbed by unsteadiness, where conduction ux is changing at micro time as
the blood sub-volume of each vessel segment q0bi is the same with well as space dimensions.
that of the tissue sub-volume. Thus, from Eq. (13), the tempera-  
qt r,t tq kt rT t r,t ty 19
ture distribution along the axial direction x of each vessel
segment can be determined. Further, q0bi is determined at the where tq and ty are phase-lags arising due to thermal inertia and
midpoint of the vessel segments, i.e., x Li =2 using Eq. (14) micro-structural interaction, respectively. In Eq. (19), ty represents
which is further used to match the rate of heat ow from tissue phase-lag of temperature gradient observed at the micro scale heat
sub-volume at the vessel wall. Baish (1994) provided further transfer and tq is the phase-lag of heat ux which is the relaxation
details on the numerical approach to simulate heat transfer along time due to fast transient effects or small scale effect
and normal to the vessel axis. of heat transport in time. Both tq and ty are material dependent.
(c) By sampling the calculated temperatures at many points in The thermal relaxation time tq for most solid varies from
the tissue, statistical measurement is performed to predict the 1010 1014 s, and can be as high as 102 s for some biological
probability density function of temperature at a particular materials (Wang et al., 2008). Using the rst order Taylor series
point. This is also done to predict the uncertainty in the expansion, the non-Fourier dual-phase-lag conduction equation
temperature due to the random proximity of large number of (Eq. (19)) is written as
blood vessels. These predictions are based on the statistical

@q r,t @
information of the vascular geometry and blood ow. qt r,t tq t kt rT t r,t ty rT t r,t 20
@t @t
In spite of being computationally extensive technique, Baish From Eq. (20) and Eq. (1), the resulting dual-phase-lag con-
statistical model has many advantages over the other models. tinuum bioheat model is written as
Baish model permits modeling of realistic vascularized tissue by !  
@2 T t 2 2 @T t
statistical information of the vascular geometry, blood ow and tq rt ct kt r T t ty kt r ob rb cb T t
the use of growth algorithm without prior knowledge of the @t 2 @t
vessel architecture.   @T t
 rt ct ob rb cb tq
@t
@q000
ob rb cb T a qm tq m
000
21
4. Recently developed bioheat models @t
If any external heat source or sink is considered in the Pennes
With advancement in technology, it is now possible to simu- bioheat equation (Eq. (1)), the dual-phase-lag bioheat equation
late realistic and application oriented processes, and perform (Eq. (21)) is further modied as
measurements on living tissues. With availability of these tech- !  
@2 T t 2 2 @T t
nologies and with need for accurate diagnostic and treatment tq rt ct kt r T t t k
y t r ob rb cb T t
procedures, researchers have proposed models which form the @t 2 @t
basis of modern biomedical applications such as hyperthermia   @T t
 rt ct ob rb cb tq
treatment using laser, laser induced interstitial thermotherapy, @t  00 
etc. These new models are reviewed in this section. @qm @qext
ob rb cb T a q00m qext tq 22
@t @t

4.1. The dual-phase-lag bioheat model
The Fouriers law of heat conduction used in all continuum
models discussed above assumes innite speed of heat wave
propagation. In other words, if a perturbation in the thermal state
is made, the corresponding change will be felt at everywhere at
same instant, no matter how far it is located. Although, Fouriers
law is applicable in many real life problems, it gives erroneous
results with material with non-homogenous inner structure as in
the case of biological tissues. It also does not hold true for problems
of rapid heating or cooling (Banerjee et al., 2005). It has been
where qext is the external heat source/sink and other symbols have
same meaning as mentioned before. As suggested by Xu et al.
(2008), the dual-phase-lag model can be further tuned by taking
second or higher order terms in the Taylor series expansion.
4.2. The porous media based bioheat model
In the year 1998, Roeztel and Xuan (1998) developed a two-
equation bioheat model that considers the effect of heat transfer in
porous media. They modeled vascular region and extravascular
region (tissue and interstitial blood) separately without considering
 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
local thermal equilibrium between the two media. In order to
account for the effect of thermal dispersion of blood through the
tissue due to porosity, they introduced an equivalent effective
thermal conductivity in the energy equations of blood and tissue.
They described that the thermal dispersion accounts for the effect of
blood perfusion in micro-vascular capillary beds. Further, their
model introduced interfacial convective heat transfer term instead
of perfusion term. This was done to consider heat transfer due to
blood bleed-off from interfacial vascular walls. With above con-
sideration, the mathematical formulation for blood and tissue are
given as
   
erb cb @T@tb ! b
v b :rT b r: kef f rT b hbt T t T b 23
   
@T t t 000
1e rt ct r: kef f rT htb T t T b 1e qm
@t t
!
where e is the porosity, v b is the blood velocity, T is the volume
averaged temperature, kef f is the effective thermal conductivity
tensor, h is the local volumetric/interfacial convective heat transfer
coefcient. In Eqs. (23) and (24), subscripts and superscripts t and b
stand for tissue and blood, respectively. For isotropic conduction,
b t
kef f and kef f are related as
b t
kef f ekb and kef f 1ekt
Basically, the two-phase models given by Roeztel and Xuan
(1998) are coupled locally by the interfacial convective heat
transfer coefcients. In case of multilayered tissue, the two-
phase models are coupled globally with the common boundary
conditions between adjacent layers. Further, the two-equation
model is extended to three-equation model in order to account
for the effect of countercurrent arteryvein pairs.
Later, Shih et al. (2002) realized that the interfacial convective
heat transfer is not similar to blood perfusion term, and they
introduced blood perfusion term instead of interfacial convective
heat transfer term in the two-equation model. They investigated the
thermal dose distribution in the living tissues during thermal
therapy of abnormal porous tissues and introduced volumetric heat
source term in the two energy equations. The model proposed by
Shih et al. combined the directional effect of blood ow, tempera-
113
have the same meaning as described earlier. Thus, the volume
averaged combined energy conservation equation is achieved by
rearranging Eqs. (26) and (27) with additional homogenous porous
property for micro-vascular porous bed
     
@T b @T t !
erb cb 1ert ct erb cb v UrT b erUkb rT b
@t @t
 
1erU kef f rT b 1eob cb T t T a 1eSt eSb 28
where e is the porosity and it is the ratio of blood sub-domain volume
to the total volume of two sub-domains. Further, Shih et al. suggested
that, based on the two sub-domain models, the thermal interaction
between blood vessels with tissue for three different cases can be
studied. These are: (i) if thermal equilibrium does not persist between
blood and tissue volume then the governing equations (Eqs. (26) and
24 (27)) must be solved separately, (ii) if the domain includes larger
blood vessels embedded in tissues, then the temperature of blood Tb
is constant and is equal to the core body temperature. In this case, the
overall energy equation (Eq. (28)) is reduced to a single equation (Eq.
(26)) where T b T a and, (iii) thermally insignicant vessels
embedded in tissues cause the tissue to quickly reach the blood
temperature, i.e., T t T b . As a result, the perfusion term will
disappear from both the sub-domain equations.
Recently, Nakayama and Kuwahara (2008) formulated a gen-
25
eralized two-equation bioheat models for vascular space (artery,
vein and capillaries) and extravascular space (tissue and inter-
stitial blood) without the consideration of local thermal equili-
brium. They subdivided the blood lled porous vascularized
tissue into the blood sub-volume and the tissue sub-volume.
According to them, the ltration process of blood from vascular
space to extravascular space through capillaries exceeds the
reabsorption process of excess blood ow back to vascular space
using lymphatic system. Thus, to account for the blood exchange
process between vascular and extravascular space, they incorpo-
rated blood perfusion term within the two sub-volume equations.
On the other hand, the dispersion and the interfacial heat transfer
terms are modeled based on conventional porous medium theory.
According to Nakayama and Kuwahara (2008), energy equations
for blood sub-volume and tissue sub-volume are given as
!
@hT ib @ 
b @ @hT ib @hT ib
ture distribution due to effective thermal conductivity and scalar erb cb rb cb u hT i ekb eK disjk
@t @xj j @xj @xj @xk
blood perfusion along with porous medium property for the micro- h i h i
vascular tumor bed. The assumption of porous tumor media is quiet b t b t
ab hb hT i hT i rb cb ob hT i hT i 29
accurate because of the fact that the abnormal tumors consist of
denser vascular pattern in the ow direction, and can be regarded as h i
@hT it @ @hT it
porous structure consisting of tumor tissues fully lled with blood. 1ert ct 1ekt ab hb hT ib hT it
@t @xj @xj
As a result, the governing equations for micro-vascular structure h i
b t
consists of two separate sub-domain models, and are coupled rb cb ob hT i hT i 1eSm 30
together by common boundary conditions. The model by Shih
et al. is quiet similar to that of Xuan and Roetzel two-equation where uj is velocity vector, e is local porosity which is the ratio of
model, and combines the energy equation of blood and tissue in blood sub-volume to total volume, k is the thermal conductivity,
order to account for the local thermal equilibrium. The solid sub- Kjk is the permeability tensor, ab is the specic surface area, hb is
domain model for the tissue is similar to Pennes model, which 
i ob is the blood
the interfacial convection heat transfer coefcient,
includes effective thermal conductivity instead of inherent thermal perfusion rate, with y as the dummy variable, y is the intrinsic


conductivity of the tissue and a source term for thermal storage average, y is the normal volume average. Superscripts and
instead of metabolic heat generation. This model is given as subscripts b and t stand for blood and tissue, and subscript dis is
  for dispersion.
@T t  
rt ct rU kef f rT t ob cb T t T b St 26 The second term on the left hand side of Eq. (29) represents
@t the macroscopic convection term due to blood ow in vessels, and
The equation for energy transfer for liquid sub-domain is given as the four terms on the right hand side of Eq. (29) represent
  macroscopic conduction, thermal dispersion, interfacial convec-
@T b !
rb cb v UrT b rUkb rT b ob cb T t T b Sb 27 tive and blood perfusion, respectively. Whereas, the four terms on
@t
the right hand side of Eq. (30) pertain to macroscopic conduction,
where St and Sb are the source terms for tissue and blood, respectively interfacial convective, blood perfusion term and metabolic source,
indicating absorbed power deposition density due to thermal dosage, respectively. Further, Nakayama and Kuwahara (2008) introduced
kef f is the enhanced thermal conductivity of the tissue (kt lob), a set of equations, viz., continuity equation and Darcy law apart
l is the constant coefcient (0.002 kg/m3s), and all other terms from the above two equations to attain the close form of governing
114 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
equations. These set of equations can be used for both temperature
and velocity computations. Subsequently, they extended the two-
equation model to three-equation model to incorporate the effect of
heat transfer in closely spaced countercurrent arteryvein pair
similar to Roetzel and Xuan model (as discussed earlier). Their
two-equation model is considered as general form of bioheat
transfer in porous medium and can be reduced to some of the
previously described perfusion based bioheat models. Similarly, the
three equation model can be reduced to some of the previously
mentioned countercurrent bioheat models
4.3. The discrete vascular bioheat model
The concept of discrete vascular model was initially developed by
Lagendijk in the year 1982 with the goal to identify the temperature
distribution around un-equilibrated large blood vessels in heated
tissues during hyperthermia treatment (Lagendijk, 1982). His obser-
vation revealed that the blood ow in larger vessels within the
heated tissue causes local non-uniformity in the temperature dis-
tribution. Later, Lagendijk et al. (1984) developed a three-
dimensional discrete vascular algorithm (DIVA) for time-dependent
temperature distributions in the vascularized tissue having straight
large vessels embedded in the tissue. In 1991, Mooibroek and
Lagendijk (1991) extended that work to much more versatile
three-dimensional algorithms by incorporating complex bending
and branching feature of vesicular geometry in the model. Later,
the completed form of DIVA model was developed by Lagendijik
et al. in 1999 which added the effect of countercurrent blood ow
(Van-Leeuwen et al., 1997) and ow through segmented real vessel
trees (Kotte et al., 1999).
The DIVA model estimates the three-dimensional temperature
distribution by logically combining the blood perfusion continuum
bioheat model for tissue volume and energy ux equation for the
ow of blood in large vessel through a common boundary condition
in the same global space. The three-dimensional coordinates in
global space is used for both positioning the elements and specify-
ing the extent of elements in the same frame of reference. These
elements in global space are either tissue sink set (also known as
voxel) or vessel segments. Voxels are discretized tissues having
regular cuboid shape and the vessel segments are three-
dimensional curved tubes embedded into certain voxels. Further,
vessel segments are discretized into sub-elements known as buck-
ets having similar nodal distance as in the tissue voxel. These
buckets can be visualized as cylindrical elements inside the vessel
segments, i.e., the vessel segments fully contain cylindrical buckets.
Moreover, every bucket is associated with one mixing cup tem-
perature, i.e., blood of each bucket is thoroughly mixed to a uniform
temperature. All these buckets form a temperature prole inside
the vessel segment. This temperature prole is shifted along the
axis of vessel segment to estimate the heat transfer due to ow of
blood in the vessel. The heat ow from the surrounding tissues to
each bucket in vessel segment embedded inside the tissue voxel
can be estimated by dividing that particular tissue voxel into
boundary elements surrounding the vessel segment. These bound-
ary elements are essentially the intermediate sets that bring heat
ow balance between the outside sink set, i.e., tissue voxel with
buckets inside vessel segment. The calculation of energy transfer
between smaller blood vessels and tissue is described by the
simplied perfusion model for sink set as
@T t
rt ct rUkt rT t q000
m Sqp 31
@t
where T t is the volume averaged tissue temperature, q000
m is the rate
of heat generation per unit volume due to metabolic heat, S
describes the source due to power deposition per unit volume
caused by heating and qp is the amount of heat withdrawn due to
the local blood ow, i.e., heat sink term.
The convective heat transport between ow of blood within
large vessel segments and vessel walls are modeled separately,
since it cannot be described by bioheat Eq. (31) applied for tissue
voxel. The vessel segment embedded in certain voxel is repre-
sented as cylindrical tube with radius r ves surrounded by coaxial
tissue cylinder representing vessel wall having radius R. Thus,
with the assumption of rotational symmetry around vessel axis,
the energy equation in cylindrical coordinate for vessel wall is
given as
 
@T 1 @ kt @T t
rt ct t @r
S 32
@t r @r
The radially outward heat ux qrves from the blood to the inner
vessel walls can be expressed as
Nukb
qrves  T t r ves T mix  33
2r ves
where Nu is the Nusselt number, kb is the thermal conductivity of
blood and T mix is the mixing cup blood temperature. The mixing
cup blood temperature T mix for each bucket in the vessel seg-
ments is given as
Z rves
T mix vm pr 2ves 2prvrTrdr 34
r0
where Tr is the radial temperature of blood, vr is the radial
velocity of blood and vm is the mean velocity of blood
R ves
r 0 2prvrdr. Thus, an analytical solution for heat ux
through vessel wall can be determined by solving Eqs. (32) and
(33), where qrves is function of T t R, Nu, R, r ves and is independent
of T t r ves . Using the analytical solution, a mean heat ux can be
estimated between bucket and neighboring intermediate tissue
sets. The temperature of intermediate set is similar to the vessel
wall T t R since radial distance R from vessel axis and intermedi-
ate set is the same, i.e., the circumferential temperature around
the vessel wall is a mean constant temperature and is common.
The intermediate tissue set balances the heat ow from the
surrounding sink set to the vessel buckets. Further, the heat ow
in the vessel related to the mass ow of the blood within the
vessel segment is simulated by shifting and re-sampling the
bucket temperature prole. The shifting of temperature proles
is achieved by assuming fully developed condition with negligible
thermal instability along the ow direction by a distance
d vm Dt in the time interval Dt. During shifting, blood
exchanges heat with the surrounding vessel boundary with the
known vessel wall temperature (Lagendijk et al., 1984) leading to
a new prole. The bucket temperature prole is estimated by
linear interpolation between two sample temperatures also
known as re-sampling.
The description of DIVA model with the analytical solution can
be obtained in detail in Kotte et al. (1996), and the description of
heat ow in countercurrent geometry is given by Van-Leeuwen
et al. (1997). The DIVA model demands more information about
the vasculature in the region of interest, and keeps track of all the
blood temperatures to estimate the tissue temperature. Due to
which, the DIVA requires high computational power.
4.4. The Deng and Liu perturbation bioheat model
In order to characterize the effect of temperature uctuations
and pulsating blood perfusion in the living tissues, in 2001, Deng
and Liu introduced uctuation terms in conventional Pennes
bioheat model (Deng and Liu, 2001). They proposed that the
blood perfusion should be treated as a transient term rather than
a constant term due to the pulsating blood ow in irregular blood
 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125 115

vessels and due to external perturbation. As a result of pulsating 4.5. The newly developed countercurrent vascular model
blood ow, the temperature in living tissues also appears uctu-
ating. This has been reported by many investigators. They In the recent past, Shrivastava and Roemer (2006) considered a
introduced a perturbation bioheat model that is capable of giving generic countercurrent vesicular network model similar to the
quantitative relation between temperature uctuation, pulsating conjugate model of Huang et al. (1996) in the vascularized tissue.
blood perfusion, and arterial blood temperature and heat genera- They readdressed some of the conceptual challenges with the
tion in living tissues. Their model incorporates small perturba- consideration of the thermally signicant blood vessels in bioheat
tions to arterial blood temperature, tissue temperature, blood transfer models as discussed earlier by many investigators (Chen
perfusion term and metabolic heat generation term by adding a and Holmes, 1980; Kotte et al., 1996; Chato, 1980; Lemons et al.,
transient uctuating term with the mean terms in traditional 1987; Baish, 1994; Huang et al., 1996). The concept of thermally
Pennes model. Deng and Liu bioheat model is given as signicant vessels was initially introduced to quantify the vessels
  which are thermally important in the vascularized tissue and to
@ T~ t T 0t h  i   h 
reduce the computational expenses for modeling each vessel in
rt ct rU kt r T~ t T 0t o ~ b o0b cb T~ a T 0a
@t  i  the tissue (Chen and Holmes, 1980). The idea was to categorize

 T~ t T 0t q~ m q0m 35 the blood vessels into thermally signicant and insignicant
blood vessels based on the thermal equilibration length analysis
where symbol  represents mean parameters and using the geometrical parameter (e.g., vessel diameter) and
represents transient uctuating parameters. All other symbols ow data.
have same meaning as discussed earlier. Further, considering the The thermal equilibration length is a characteristic length over
time averaging technique, Eq. (35) is written as which inlet blood vessel temperature equilibrates with the
surrounding tissue temperature, and if the thermal equilibration
@T~ t T 0t
rt ct rUkt rT~ t T 0t  length is comparable to the actual vessel length, the vessels are
@t
 considered as thermally signicant (Chen and Holmes; 1980).
o~ b o0b cb T~ a T 0a T~ t T 0t  q~ m q0m 36 Subsequently, Chato (1980), Lemons et al. (1987) and Baish
where the symbol represents time averaged quantity. Sub- (1994) stated that the existence of local thermal equilibrium
stituting the statistical relations for small uctuating quantities condition between tissue and vessel outlet leads to thermally
given as insignicant smaller diameter vessels branched from the larger
diameter vessels. This is mainly due to the fact that the smaller
T 0t 0, T 0a 0, o0b 0, q0m 0 37 and the larger diameter vessels stay in thermal equilibrium
in the time averaged form, Eq. (36) is reduced to a simplied form throughout their length and are not contributing to the overall
given as heat transfer rate. As a result, the condition of local thermal non-
equilibrium at the vessel outlet is important for thermally
@T~ t h i h i
rt ct rU kt rT~ t ob T a T~ t cb o0b T 0t cb o0b T 0a qm 38 signicant vessels. Further, the quantication of thermally sig-
@t nicant vessel was done by various numerical investigations
Now subtracting Eq. (38) from Eq. (35) leads to the perturba- (Baish, 1994; Huang et al., 1996). According to them, if the total
tion bioheat equation and is given as thermal energy exchanged by vessels is comparable to that of the
h i energy exchanged by the complete vessel network embedded in
@T 0 
rt ct t rU kt rT 0t o0b cb T a T~ t cb o0b T 0t cb o0b T 0a the tissue, then those highly contributing vessels are regarded as
@t
thermally signicant vessels. Later, Shrivastava and Roemer
cb o0b T 0t cb ob T 0t cb o0b T 0a cb ob T 0a q0m 39
(2006) suggested that (a) the concept of vessel diameter and
Further, Eqs. (38) and (39) are simplied by the order of equilibration length, (b) local thermal non-equilibrium/equili-
magnitude analysis, which yield brium conditions of vessels and (c) the criterion of total heat
h i h i transfer rate are inappropriate to categorize signicance of
@T~ t
rt ct rU kt rT~ t ob cb T a T~ t qm 40 different diameter vessels in bioheat transfer models. They
@t
pointed out some very important factors which have been over-
@T 0t  h i looked by many investigators while considering the concept of
rt ct rU kt rT 0t o0b cb T a T~ t cb ob T 0t cb ob T 0a q0m 41 thermally signicant/insignicant vessels in the vascularized
@t
tissue. These are: (a) the increase in heat transfer in the form of
The mean and the uctuating forms of convective boundary
diffusion and convection across the surface of smaller diameter
conditions, adiabatic body core conditions and initial conditions
blood vessels due to the increase in the total surface area (as
are given as
smaller vessels are numerous), (b) the effect of blood vessel
@T~ t   @T 0 spatial distribution in the tissue and the effect of source term
kt hf T f T~ t and kt t hf T 0t , x0 42
@x @x on local thermal equilibrium/non-equilibrium condition and
(c) the effect of uniform and non-uniform heating and cooling
@T~ t @T 0 of vascularized tissues in numerous biomedical applications.
kt 0 and kt t 0, xL 43
@x @x In view of the above, they have modeled 3-D generic vascular
geometry in counterow arrangement having vessel diameter
T~ t T t 0,x, and T0t 0,x 0, t0 44
ranging from 101000 mm within the tissue. The aim of this
where x 0 is for skin surface, x L is for body core, hf is the heat model is to determine: (a) the absolute heat transfer between
convection coefcient, T f is the environmental temperature and tissues and different diameter vessels, (b) the thermal signicance
T f 0,x is the initial condition. The study by Deng and Liu (2001) of these vessels, (c) the sensitivity of network variables on the
indicates that the small perturbation in the blood perfusion thermal contributions and (d) the effect of hyperthermia and
causes signicant temperature uctuation in the living tissues. hypothermia conditions. The mathematical model comprises of
Their model is quiet similar to the time averaged Reynolds two energy equations, viz., conduction and convection equations,
equation in uid dynamics, and serves as a foundation model to and a condition to satisfy the heat transfer between the
quantify temperature uctuation behavior in living tissues. vessel walls and the surrounding tissues. The three dimensional
116 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125

steady-state conduction equation has been used to determine the where in the case of no other vessel in the adjacent nodes, the
temperature distribution in the tissue and is given as thermal resistance is given as

rUK t rT t Q 000t 0 45 Ddi rv  4
Rth di Rth di 
K t Ddj Ddk i a j a k NuK b pDsv
Eq. (45) has been discretized using second order nite differ- where i 1,2; j 1,2,3 and k 1,2,3 51
ence scheme. It is given as
and in the case of another vessel on the adjacent of rst vessel at

T x Dx,y,zT x,y,z T xDx,y,zT x,y,z
Rth x Rth x

T x,y Dy,zT x,y,z T x,yDy,zT x,y,z
Rth y Rth y

T x,y,z DzT x,y,z T x,y,zDzT x,y,z
Rth z Rth z
Q 000
t DxDyDz 0
where Q 000
t is the applied volumetric power of the tissue, Rth is the
thermal resistance, Dx, Dy and Dz are discretizations in x, y and z
coordinates, subscript t is for tissue, subscript th is for thermal,
and  within the parenthesis represent the positive and
negative directions, respectively.
The thermal resistance incorporates the effect of existence/
non-existence of adjacent vessel node to the tissue node within
the tissue conduction equation (Eq. (46)). For the case of non-
adjacent vessel to the tissue node, Rth di is given as
 vascular network model is necessary to have a realistic prediction. In
Ddi 
Rth di Rth di 
K t Ddj Ddk i a j a k
where i 1,2,3; j 1,2,3 and k 1,2,3
and, for the case of vessel at the next node having radius r 0v
adjacent to the tissue node, the thermal resistance is given as
 ow rate and size of vessel in vascularized tissue during hyperther-
  Ddi r0v  4
Rth di =
K t Ddj Ddk i a j a k NuK b pDs0v
where i 1,2,3; j 1,2,3 and k 1,2,3
where Dd is the nodal spacing, i, j, k are the orthogonal directions,
Nu is the Nusselt number, K b is the thermal conductivity of blood,
Ds0v is the discretization of vessel present at the next node from
the node in consideration, subscripts b and v represent blood and
vessel, respectively.
The blood vessel network has been modeled using one-
dimensional steady-state convective energy equation, and is
given as
    000
mcp b rT b NuK b p T v,w T b Q b pr 2v
where r v is the blood vessel diameter, subscript w is for wall. Eq.
(49) has been discretized using rst-order nite difference
scheme and is solved for both arteries and veins in counterow.
The condition to satisfy the heat transfer between the vessel walls
and surrounding tissues has been obtained by conservation of
energy between tissue adjacent to the vessel walls and by
considering that at the vesseltissue interface, the tissue tem-
perature is equal to the vessel temperature, which is given as (say
blood is owing in z direction)
some branching corners, the thermal resistance is given as
  Dd r r0  4 4
Rth di = K tiDdjvDd v 
k iajak
NuK b pDsv NuK b pDs0v
where i 1,2; j 1,2,3 and k 1,2,3 52
where Dsv is the vessel discretization.
Further, Shrivastava and Roemer (2006) have suggested a special
relationship to satisfy the conservation of mass and energy at each
arterial and venous branching node. The extensive parametric study
46
performed by Shrivastava and Roemer (2006) lead to the conclu-
sions: (a) the thermally signicant blood vessel is a strong function of
the inlet blood temperature, surrounding tissue temperature dis-
tribution and blood mass ow, (b) the absolute diameters and heat
transfer coefcients do not affect the thermal signicance of the
vessels much, and (c) both arteries and veins play important role in
the exchange of energy, with the arteries playing major role. These
results lead to the conclusion that, depending on the problem under
consideration the thermal signicance of vessels can only be
established on case-to-case basis, and the site and problem specic
the application of hyperthermia treatment earlier Huang et al. (1996)
considered a similar countercurrent vascular network model to
determine the temperature distribution and the role of counter-
47 current vessel network. Their model combined the convection
equation for vascular ow with the Pennes bioheat equation instead
of conduction equation for tissue. Recently, the same conjugate
model of Huang et al. (1996) has been used to predict the effect of
mia treatment (Huang et al., 2010).
Owing to the involved difculties and increased cost of computa-
48 tional modeling of heat transfer phenomenon in the vascularized
tissue, many investigators (Chen and Holmes, 1980; Lemons et al.,
1987; Chato, 1980; Baish and Ayyaswamy, 1986; Zhu et al., 1988,
1990; Zhu et al., 1996; Wu et al., 1993; Shrivastava and Roemer,
2004, 2005) developed conduction shape factors to approximate the
heat transfer phenomenon between complex shaped vessels
embedded in the tissue. These shape factors have proven to be
unique tool because it signicantly reduces the computational cost
by casting the multi-dimensional form of bioheat transfer problem
into a simple empirical form involving complex geometrical arrange-
ment (Shrivastava et al., 2005; Shrivastava and Roemer, 2005). In
49 principal, the conduction shape factors depend on geometry. As a
result, previous approximations for conduction shape factors were
limited to the geometrical constraint (Chen and Holmes, 1980;
Lemons et al., 1987; Chato, 1980; Baish and Ayyaswamy, 1986;
Zhu et al., 1988, 1990, 1996; Wu et al., 1993), and neglected the
effect of: (a) tissue source term which are omnipresent in the form of
local hyperthermia and hypothermia due to blood ow, metabolism
and external sources, and (b) unequal vesseltissue heat transfer
rates from the paired vessels due to unheated/heated tissues.
Recently, Shrivastava et al. considered the effect of source term on
the conduction shape factor for vesseltissue congurations

 (Shrivastava and Roemer, 2004, 2005) and the effect of unequal

T x Dx,y,zT v,w T xDx,y,zT v,w
vesseltissue heat transfer (Shrivastava et al., 2005) on the conduc-
Rth x Rth x
tion shape factors due to unheated/heated tissue (Shrivastava and


T x,y Dy,zT v,w T x,yDy,zT v,w Roemer, 2005). Their analysis was based on the analytical solution of

Rth y Rth y Poisson conduction shape factor for individual (Shrivastava and
 Roemer, 2004) and paired blood vessels (Shrivastava and Roemer,
NuK b p T v,w T b x,y,z 50 2005) within the nite tissue matrix.
 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125 117

Based on the parametric study of different geometrical and absorbed power deposition in W/m3. By integrating Eq. (53) over
thermal congurations of vessels and tissues, for paired vessels the tissue and blood sub-volumes separately and applying Greens
embedded in the tissue matrix as in the case of countercurrent heat divergence theorem to the rst term on the left and right hand
transfer, they have concluded the following. They are: (a) the sides gives a modied form
vesselvessel heat transfer rates are independent of the source R R   R h i
@T i ! !
term where as the heat transfer rate between the two vessels and V i rci @t dV V i T i rU rc v dV Ai ki rT i rci v i T i Un^ i dA
i
surrounding tissue is a strong function of source term (Shrivastava Z Z
et al., 2005), (b) in an unheated/heated tissue, the vesselvessel ki rT i Un^ ji dA Q 000
i dV, i j t,b and i aj 54
Poisson conduction shape factors are only dependent on geometry Aji Vi

and are independent of source term, vessel wall temperature and where n^ ji is the normal vector from j to i, Aji is the surface area
tissue boundary temperature. But, the vesseltissue Poisson con- between control volume i and j. By considering blood ow as
duction shape factor is a strong function of source term, vessel wall incompressible and all the properties constant, the second term
temperature and tissue boundary temperature (Shrivastava and on the left hand side of Eq. (54) becomes zero due to conservation
Roemer, 2005). (c) Caution should be taken while considering of mass giving a continuum form of the energy equation. Further,
vesseltissue Poisson conduction shape factor since the shape factor the volume averaged energy equation is obtained by normalizing
might become singular in heated/unheated tissue (Shrivastava and Eq. (54) by volume Vi and is given as
Roemer, 2005). Further, the effect of individual vessel embedded R R h i
1 @T i 1 !
V i V i rci @t dV V i Ai ki rT i rci v i T i Uni dA
within the tissue matrix for different geometrical and thermal ^
congurations on vesseltissue Poisson conduction shape factors Z Z
1 1
are studied by Shrivastava and Roemer (2004). Based on the ki rT i Un^ ji dA Q 000 dV, i j t,b and i aj
V i Aji V i Vi i
analytical study, the following conclusions were made:
(a) Poisson conduction shape factor based on average tissue matrix 55
temperature is more suitable instead of average tissue boundary where the rst term on the right hand side of Eq. (55) represents
temperature when the source term is present, and (b) the presence the energy gained by ith tissue (or blood) control volume from
of source term on Poisson conduction shape factor has negligible adjacent tissue (or blood) control volume. The second term on the
effect on the associated changes in the geometrical congurations, right hand side of Eq. (55) represents the energy gained from
viz., vessel size and eccentricities. The quantitative and comparative tissue-blood thermal interaction in the vascularized tissue
study by Shrivastava et al. (2005) on the heat transfer rates volume. Eq. (55) is further simplied by applying four funda-
between vesselvessel and vesseltissue within the conguration mental assumptions. These are: (i) the rst simplication is
of arbitrary located paired vessels innite, unheated/heated tissue similar to that proposed by other investigators (Chen and
domain with non-insulated boundary conditions lead to many Holmes, 1980; Weinbaum and Jiji, 1985) for conductive heat
important conclusions which were overlooked by previous transfer given as
researchers (Chato, 1980; Lemons et al., 1987; Baish and Z
1  
Ayyaswamy, 1986; Baish et al., 1986; Zhu et al., 1988, 1990, ki rT i n^ i dA C i1 rU ki rT i , i t,b 56
V i Ai
1996; Wu et al., 1993; Van-Leeuwen et al., 1997). These conclusions
are: (a) the vesselvessel and the vesseltissue heat transfer rates where Ci1 is the new constant term introduced in the equation to
depends on the temperature difference between the vessel walls keep the simplication generalized and T is the volume averaged
and the tissue boundary as well as on the geometrical conguration tissue (or blood) temperature. Also, they suggested that this
of the vessels, viz., radius of vessel/surface area and eccentricity, simplication is further needed to be veried for vascularized
(b)the vesseltissue heat transfer rate is strong function of the tissue both numerically and experimentally. (ii) The second
deposited power compared to vesselvessel heat transfer rate, simplication describes the heat transfer due to tissuevessel
(c) the vesselvessel and vesseltissue heat transfer rate are of thermal interaction and is given as
the same order in case of cold vessel with unheated tissue, whereas Z
1 UAji 
the heat transfer between vesselvessel demarcates compared to ki rT i n^ ji dA T j T i , i j t,b and i a j 57
V i Aji Vi
vesseltissue as the tissue is heated gradually, and (d) there is
always a net heat transfer between paired vessel and tissue, and the where U is the heat transfer coefcient based on volume averaged
heat transfer rates between two vessel and tissue are unequal. temperatures and are not similar to conventional heat transfer
Similar observations are pointed out by varying the thermal coefcient. They suggested that the new heat transfer coefcient
conditions of paired vessels and tissues, which are detailed in can be determined by proper quantication of Nusselt number
Shrivastava et al. (2005). More information on the various solutions and obtained data from Poisson conduction shape factor
and mathematical analysis can be obtained from Shrivastava et al. (Shrivastava and Vaughan, 2009). (iii) The third simplication
(2005) and Shrivastava and Roemer (2004, 2005). analyzes the relative importance of convective term over con-
ductive term for blood volume, and is given as
4.6. The Shrivastava and Vaughan perfusion based bioheat model Z h i Z
1 ! 1
rci v i T i n^ i dAb ki rT i n^ i dA, i b only 58
V i Ai V i Ai
The Shrivastava and Vaughan model is one of the recent
bioheat models (Shrivastava and Vaughan, 2009). Owing to its As suggested by Shrivastava and Vaughan, the above simpli-
continuum nature, this model is computationally less intensive cation is quiet appropriate since the convective term is larger
and is applicable for both unheated and heated tissues. For tissue than the conductive term for blood volume. (iv) The fourth
and blood sub-domains, the energy conservation equation with simplication refers to introducing perfusion term to convective
heated/unheated vascularized tissue is term for blood volume and is given as
  Z h i
@T i ! 1 ! 
rci v i UrT i rUki rT i Q 000
i , i t,b 53 rci v i T i n^ i dA C i2 rU Pci T i , i b only 59
@t V i Ai

where the subscript i is for tissue and blood sub-domain, the where Ci2 is the constant parameter to keep simplication
!
velocity of tissue v t is zero and Q 000 is the source term indicating generalized and P is the perfusion vector mass ux in kg/m2s
118 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
R !
and is 1=vb vb r v b dV. Similar to the rst assumption, they
suggested that, the fourth assumption needed to be veried by
both numerical and experimental investigations for vascularized
tissues. The nal form of bioheat equations for tissue sub-volume
and blood sub-volume after implementing the above simplica-
tions in Eq. (55) are given as
@T t    UAbt   spectral region of 2.5 mm15 mm (Barnes, 1968), which lies between
rt ct C t1 rU kt rT t T b T t Q 000
t 60
@t Vt
@T b  UAbt   the intensity of radiation is independent of direction and is given as
rb cb C b2 rU Pcb T b T t T b Q 000
b 61
@t Vb
where the parameters r and c are volume averaged quantities,
and Q 000 is the volumetric absorbed power deposition in the tissue
and blood volume. Shrivastava and Vaughan (2009) proposed
that, once all the parameters are available, these two models can
be coupled together to determine the volume averaged tempera-
ture distribution in the vascularized tissue during heated/
unheated physiological conditions instead of solving true point-
to-point temperature distribution in the vascular model. More
details on the identication of unknown parameters are available
in Shrivastava and Vaughan (2009). Further, they claimed that
this model requires sensitive study to determine the relative
importance of different parameters in predicting temperature of
specic vascularized organs.
Although, the Shrivastava and Vaughan model has satised the
required criteria as discussed earlier but it requires detailed
parameters based on knowledge of real organ and vasculature
geometry, physiological/patho-physiological conditions of the
organ and relative changes in the conditions due to heated/
unheated vascularized tissues. Furthermore, they have suggested
that the new constants which were introduced in this model
needs to be veried.
4.7. The radiation based bioheat models
Thermal radiation is a volumetric phenomenon, and its man-
ifestation is owing to the nite temperature of a body. The
average core body temperature of the human body is 310 K. Thus,
every component of the human body is the source of thermal
radiation. Further, the human tissue is an optically participating
medium, and externally imposed radiation penetrates inside the
human body, and in the process, while propagating, it undergoes
absorption, emission and scattering. The thermal radiation phe-
nomenon always has a signicant contribution to biological heat
transfer when compared to conduction and convection processes.
Since thermal radiation is possible with or without temperature
gradient for a body whose temperature is above the absolute zero
Kelvin. Present section of the article discusses about the radiative
transfer models that have evolved into an important non-invasive
tools for heat transfer analysis because of its paramount usage in
biomedical applications, viz. application of thermography in
diagnosis of subsurface cancer (Centigul and Herman, 2011;
Absalan et al., 2012), post-treatment monitoring (Costello et al.,
2012) and plethysmography (Bouzida et al., 2009), and applica-
tion of laserphotonics in diagnosis (Das et al., 2003) and therapy
(Jiao and Guo, 2009; Mohammed and Verhey, 2005).
4.7.1. The surface radiation phenomenon
Initial attempt to quantify human body heat transfer due to
thermal radiation was based on surface radiation phenomenon
(Barnes, 1968). Fundamentally, the surface radiation phenom-
enon is based on StefanBoltzmann law which provides a relation
between emissive power and temperature. It states that the total
energy emitted per unit surface area of a body per unit time
is directly proportional to the fourth power of the blackbody
temperature. Mathematically, it is given as
E esT 4t 62
where T t is the absolute temperature of the tissue, s 5:67  108
Wm2 K4 is the StefanBoltzmann constant, e is emissivity and it
reach unity for the blackbody. For the human skin, the emission is in
near infra-red region and thermal infra-red region of the electro-
magnetic spectrum. The emission from the blackbody is diffuse, i.e.,
sT 4t
Ib 63
p
where Ib is black body intensity. Since emissivity of the human skin
is close to unity (Steketee, 1973), the energy emitted by the human
skin is closely approximated to that of the blackbody.
4.7.2. The volumetric radiative model
Generally, all biological tissues are radiatively participating
and exhibit torturous photon transport within the tissue due to
associated in-homogeneities embedded in them. The transport of
light while propagating in any direction within a participating
bio-media undergoes attenuation due to absorption and
out-scattering, and augmentation owing to emission and in-
scattering. Thus, the general relation that governs the behavior
of radiative heat transfer within a radiatively participating bio-
media in any direction s^ about the elemental solid angle DO is
given by (Modest, 2003; Howell et al., 2011; Mishra et al., 2006):
  Z
1 @I dI ss  
bI ka Ib IF s^ , s^ i dOi 64
c @t ds 4p
4p
where I is the intensity which is a function of direction Oy, f and
wavelength l, y is the polar angle, f is the azimuthal angle, s is the
geometrical distance in the direction s^ , t is the time, c is the speed of
light in the medium cvac =n, cvac is the speed of light in vacuum, n
is refractive index, ka is the absorption coefcient, ss is the scattering
 
coefcient, b ka ss is the extinction coefcient, F s^ , s^ i is the
scattering phase function and subscript i is for incident direction. The
rst term on the right hand side of Eq. (64) explains attenuation due
to absorption and out-scattering, the second and third terms explain
augmentation in the direction s^ due to emission and in-scattering,
respectively. The above theory has been developed by neglecting
the polarization effect, by considering independent scattering
i: e:, c=l b1:0 and by assuming that the scatterers are far enough
to scatter independently (Modest, 2003).
In the case of short-pulse laser irradiation of biological tissues
during medical diagnosis (Das et al., 2003; Pal et al., 2006) and
thermal therapeutic techniques (Diederich, 2005; Whelan et al.,
2005), the external collimated intensity of laser Ic undergoes attenua-
tion while propagating within the medium. The attenuation of
collimated intensity of laser Ic within the medium, gives rise to the
diffuse intensity Id . Thus, in the participating medium, intensity I is
composed of two components, viz., collimated and diffuse: I Ic Id .
The attenuation of collimated intensity within the medium is
given as
1 dIc dIc
bIc 65
c dt ds
Substituting the expression for intensity I Ic Id as well as
attenuation of collimated intensity (Eq. (65)) in Eq. (64) results
into a modied radiative transfer equation given as
1 @Id dId
bId Sc Sd
c @t ds
bId St 66
 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125 119

where St Sc Sd is the total source term, Sc is the source term the solution of radiative transfer Eq. (66), the distribution of total
!
due to collimated intensity and Sd is the source term due to heat ux qt at each location r within the medium is determined as
diffuse intensity. The source term resulting from collimated      
! ! !
intensity Sc and diffuse intensity of radiation Sd are given as qt r ,t qc r ,t qd r ,t 73
Z
ss  
Sc Ic F s^ , s^ i dOi 67 where qc and qd are the heat ux due to collimated radiation and
4p
4p diffuse radiation respectively at each node. The heat ux due to
Z diffuse Id and collimated Ic intensity can be calculated at each
ss   !
Sd ka Ib Id F s^ , s^ i dOi 68 location r using
4p   Z  
4p ! !
qj r ,t Ij r , s^ ,t cosyi dOi 74
The diffuse intensity Id is obtained from the solution of 4p
resulting hyperbolic form of differential equation Eq. (66), which
is given in details by Mishra et al. (2006). The scattering of light in where subscript j is a variable representing either collimated, c or
biological medium are mostly in forward direction (Van-Gemert diffuse, d. The detailed description of the formulations for transient
 
et al., 1989), due to which the scattering phase functions F s^ , s^ i in radiative equation and various numerical techniques are given by
Eqs. (67) and (68) are approximated either using Legendre Mishra et al. (2006).
decomposition phase function (Modest, 2003) or Henyey In case of optical detection techniques, the heat ux at the
Greenstein phase function (Henyey and Greenstein, 1941). The boundaries serves as a non-invasive means to determine the
Legendre decomposition scattering phase function in terms of the conditions of biological tissues. These characteristic signals, viz.,
Legendre polynomial Pn is given as reectance and transmittance from tissue boundaries are found to
be a function of tissue optical properties and conguration of
  X
N  
F s^ , s^ i 1 An Pn s^ , s^ i 69 laser source. In order to demonstrate this, the characteristic
n1 signals/normalized heat uxes from human breast model
(Fig. 1) are analyzed when subjected to pulsating laser of
and the HenyeyGreenstein phase function in terms of the
Gaussian temporal prole. These normalized heat uxes, viz.,
Legendre polynomial Pn is given as
the reectance qnref r,t and the transmittance qntr r,t from the
  X
N   boundaries of human breast models are depicted in Figs. 24. The
F s^ , s^ i 1 2n 1 g n P n s^ , s^ i 70
n1
breast is approximated as 1-d planar inhomogeneous medium
and a longitudinal section of tissue are cut along the dorsoventral
where An is anisotropy coefcient, g is the asymmetry factor/ axis of breast while moving along the anteroposterior axis of
average cosine of phase function, N is the order of approximation, breast, as can be seen from Fig. 1. The anatomical details of breast
s^ and s^ i are the direction vectors. The coefcients An and g are tissues are obtained from the available literature (Sampson and
related to Mie-scattering coefcients an and bn through compli- Hillman, 2004; Highnam et al., 1998; Burch and Law, 1995). The
cated formulae given by Chu and Churchill (1955) and Modest optical properties of multilayered breast tissues (Bashkatov et al.,
(2003). 2011; Peters et al., 1990; Key et al., 1991) are considered within
The variation of collimated intensity Ic within a participating the therapeutic optical window, i.e., 500 nm, 800 nm and
medium subjected to a short-pulse laser of step and Gaussian 1300 nm wavelengths as shown in Table 1. The refractive indices
temporal prole is given by Muthukumaran and Mishra (2008, of the breast tissues are considered as 1.4. The top boundary of
2009) as the breast model has been irradiated using a laser source at an
   
Ic s^ ,t Ic,max s^ ,t expbsc angle (yc 01) from outward normal n. ^ The reectance and
" #
HfbctcNp 1T p sc g   transmittance signals are thus given as
 HfbctcN 1T s bct g f or 0 o t p o t  
p p c p   qd r w,N ,t
  qnref r w,N ,t   75
dyyc  d ffc 71 qc,max r w,N ,t

 s   
    2
t cc t c N p 1T p
Ic s^ ,t Ic,max s^ ,t expbsc  exp 4 tp ln2
 
dyyc  d ffc for 0 r t r2t c
where Ic,max is the collimated intensity at the irradiated boundary,
Np is the number of pulses, sc z=cosyc is the geometric distance
in the direction Oc , subscript c is for incident collimated ray
direction,t p is the pulse width, t c is the cut-off period ( 3tp), T p is
the time period of pulse train which is 2t p for step pulse and 6t p
for Gaussian pulse, d is the Dirac-delta function and H is the
Heaviside function. The Dirac-delta function d ensures the exis-
tence of collimated intensity in the direction Oc , and the Heavi-
side function H guarantees that the pulsating collimated intensity
persists for short-span t p within the medium at each location.
It is to be noted that the evaluation of source term St Sc Sd
in Eq. (66) requires the knowledge of intensity distribution Ic and
Id . With the variation of collimated intensity calculated from
Eqs. (71) and (72), the distribution of diffuse intensity Id is obtained
from the solution of Eq. (66) using implicit iteration technique
(Mishra et al., 2006). Once the intensities have been evaluated
within the medium about discrete angles at discrete nodes using
 
  q r w,S ,t qd r w,S ,t
qntr r w,S ,t c   76
qc,max r w,N ,t
72
where qc,max is the maximum heat ux due collimated ray at the
 
irradiated boundary r w,N , r w,N represents position of north wall,
r w:S represents position of south wall, subscript w, N and S
represents wall, north and south, respectively.
The temporal spectra of characteristic signals, viz., transmit-
tance and reectance from breast at three different wavelengths
l 500, 800 and 1100 nm are analyzed in Fig. 2. Owing to the
increased scattering at longer wavelength l 1100 nm, light
diffuses to a greater depth at longer wavelength compared to that
at shorter wavelength l 800 nm as can be seen from the
variation in magnitude of transmitted signals (Fig. 2(a)). It is
observed that at shorter wavelength, i.e., l 500 nm no signals
appear at the south boundary. This is mainly due to the fact that,
(a) the absorption of light at shorter wavelength is more com-
pared to longer wavelength as can be seen from Table 1, and
(b) breast is optically thick causing no/weak transmitted signals
at the south boundary. Whereas, the magnitude of reected
signals are detectable at shorter wavelength and exhibit
120 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
Z
Dorsoventral axis
X X
Anteroposterior
axis
Z
n Ic
Fig. 1. Schematic of breast model given as: (a) side view, (b) sectional front view (zoomed) and (c) 1-d planar longitudinal cut.
demarcation in peak magnitude compared to longer wavelength
as can be seen from Fig. 2(b).
The temporal variation of characteristic signals from breast for
different laser congurations, viz., pulse width t p and number of
 
pulses Np are analyzed in Figs. 3 and 4 at a particular wavelength
l 1100 nm. The variation of transmittance and reectance from
the human breast subjected to single and train of 5 laser pulses
having pulse width of 1ns are depicted in Fig. 3(a) and (b). Similarly,
the variation of transmittance and reectance from the human
breast subjected to single and train of 5 laser pulses for pulse width
of 100ps, 10ps and 1ps are depicted in Fig. 4(a) and (b), Fig. 4(c) and
(d), and Fig. 4(e) and (f), respectively. The common observation in
Figs. 3(a), 4(a), (c) and (e) is that the peak magnitude and temporal
spread of transmitted signals reduces as the pulse width decreases.
Similarly, the peak magnitude and temporal spread of reected
signals reduces with the decreasing pulse width as can be seen from
Figs. 3(b), 4(b), (d) and (f). This is due to the fact that the energy
input to the medium is more at longer pulse width t p compared to
shorter pulse widths. It can be observed from Fig. 3(a) and (b) that at
1 ns pulse width, the peak magnitude and temporal spreads of
characteristic signals exhibit similar behavior for 1 and 5 incident
pulses. The transmittance is highly attenuated owing to the
increased opacity in the direction of propagation of light. The
temporal spreads of single and 5 pulses are almost same as that of
the incident laser pulse as can be seen from Fig. 3(a). No cascading
effect has been observed. This is for the reason that the effect of one
incident pulse persist within the breast for 6 ns and the effect of
next pulse will be observed after 6 ns, whereas the laser wave takes
much less time1:4  31:1  103 m=3:0  108 m s1 0:145 ns to
reach the south boundary of the breast. In case of laser having pulse
width 100 ps (0.1 ns) (Fig. 4(a)), the effect of one incident pulse is
Z
Epidermis
Dermis
Y Y
Fat
Fibro-glandular
Z
0.1 mm
1.0 mm
5.0 mm
25 mm
over by 0.6 ns which is comparable to the time taken by laser to
reach the south boundary, i.e. 0.145 ns and hence the cascading
effect become visible from the next pulses. On the other hand, in
case of laser having pulse width of 10 ps (Fig. 4(c)) and 1 ps
(Fig. 4(e)), the effect of one pulse is over by 0.06 ns and 0.006 ns,
respectively. Whereas, the next pulse appears after 0.06 ns in case of
10 ps and after 0.006 ns in case of 1 ps which is quicker than the
time taken by the signals to reach the south boundary, i.e. 0.145 ns
and hence the observed change in trend of transmitted signals as
can be seen from Fig. 4(c) and (e).
In case of laser having pulse width of 1ns (Fig. 3b) and 100 ps
(Fig. 4b), no cascading effect is visible and the signals behave
almost the same as that of the incident pulse for single and
5 pulses. Since the reected signal reaches the north boundary
much earlier than the time for which the incident laser pulse
persist within the medium, i.e. 6 ns  5 30 ns for 5 pulses of
1 ns pulse width and 0:6 ns  5 3 ns for 5 pulses of 100 ps
pulse width. On the other hand, in case of laser having pulse
width of 10 ps (Fig. 4d) and 1 ps (Fig. 4f) the time of arrival of
reected signals at north boundary is comparable to the single
incident laser pulse, i.e. 0.06 ns for 10 ps and 0.006 ns for 1 ps,
and hence the cascading effect become visible from next pulses
after the effect of one incident laser pulse is over.
4.8. The combined continuum and radiative model
The integrated thermal effect due to heat diffusion, blood
perfusion, metabolic heat generation and thermal radiation can
be evaluated using combined form of continuum bioheat model
and radiative transfer model. Especially, the Pennes bioheat
model is successfully combined with the radiative model by some
 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125 121

1.2 1.6x10-8
tp = 1 ns, = 1100 nm
tp = 1 ns, Np = 1
1.4x10-8 Number of pulse
1.0 = 1100 nm Np = 1
= 800 nm 1.2x10-8 Np = 5
= 500 nm
0.8

Transmittance (q*tr)
Transmittance (q*tr)

1.0x10-8

0.6 8.0x10-8

X 10-8
6.0x10-9
0.4
4.0x10-9

X 10-25
0.2 2.0x10-9

0.0
0 5 10 15 20 25 30 35
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 Time, t (ns)
Time, t (ns) 0.9
0.6 tp = 1 ns, = 1100 nm
0.8
Number of pulse
tp = 1 ns, Np = 1 Np = 1
= 1100 nm 0.7
0.5 Np = 5
= 800 nm
0.6
= 500 nm
Reflectance (q*ref )

0.4 0.5
Reflectance (q*ref )

0.4
0.3
0.3

0.2 0.2

0.1
0.1
0.0
0 5 10 15 20 25 30 35
Time, t (ns)
0.0    
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 Fig. 3. Temporal variation of (a) transmitted qntr and (b) reected qnref signals
from the human breast subjected to single and train of 5 laser pulses having pulse
Time, t (ns) width of t p 1 ns.

Fig. 2. Temporal spectra of characteristic signals: (a) transmittance and

(b) reectance from human breast at three different l 500, 800 and 1100 nm 5. Thermophysical and optical properties
wavelengths.

Heat transfer in a biological system is a complex process as it

investigators (Kim and Guo, 2007; Feng et al., 2009; Jiao and Guo, involves many mechanisms such as conduction, convection,
2009) and is given as radiation, metabolism, evaporation, phase change, etc. The knowl-
  edge of tissue and bio-uid properties, viz., thermo-physical and
! h    i
@T t r ,t ! ! optical properties helps us to understand the thermal behavior as
rt ct rU qcond r ,t qrad r ,t
well as the various heat transfer phenomena in the biological
@t
 h  i  
0 ! ! medium. Moreover, the change in metabolic activities, composi-
1k ob cb T a T t r ,t q000
m r ,t 77
tion and growth of cells also changes the in-vivo thermophysical
and optical properties of the affected tissue compared to normal
where qcond is the heat ux due to conduction. The above equation
conditions. These thermophysical and optical properties are key
can further written as
parameters to identify any anomalies in the human body. This
 
! h  i   section discusses the thermophysical and optical properties of
@T t r ,t ! !
rt ct rU kt rUT t r ,t rUqrad r ,t biological medium for normal and abnormal conditions.
@t
h  i   The thermophysical properties of any substance is a function of
 0 ! !
1k ob cb T a T t r ,t q000
m r ,t 78 temperature and pressure, while in case of living beings, it also
varies with various physiological activities in normal and abnormal
where qrad is the heat ux due to radiation and can be evaluated conditions. Especially, thermal properties in biological system show
from Eq. (73). signicant change due to temperature, and play an important role in
122 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125

6.0x10-9 0.8
tp = 100 ps, = 1100 nm tp = 100 ps, = 1100 nm
Number of pulse 0.7 Number of pulse
5.0x10-9 Np = 1 Np = 1
0.6
Transmittance (q*tr) Np = 5 Np = 5

Reflectance (q*ref )
4.0x10-9 0.5

3.0x10-9 0.4

0.3
2.0x10-9
0.2
1.0x10-9 0.1

0.0 0.0
0.0 1.0 2.0 3.0 4.0 5.0 0.0 1.0 2.0 3.0 4.0
Time, t (ps) Time, t (ps)

2.5x10-9 0.44
tp= 10 ps, = 1100 nm 0.40 tp = 10 ps, = 1100 nm
Number of pulse Number of pulse
0.36
2.0x10-9 Np = 1 Np = 1
0.32
Transmittance (q*tr)

Np = 5

Reflectance (q*ref )
Np = 5
0.28
1.5x10-9
0.24
0.20
1.0x10-9 0.16
0.12
5.0x10-9 0.08
0.04
0.0 0.00
0 400 800 1200 1600 0 100 200 300 400 500
Time, t (ps) Time, t (ps)

3.5x10-10 0.16
tp = 1 ps, = 1100 nm tp = 1 ps, = 1100 nm
0.14
3.0x10-10 Number of pulse Number of pulse
Np= 1 0.12 Np = 1
2.5x10-10
Transmittance (q*tr)

Np = 5
Reflectance (q*ref )

Np = 5
0.10
2.0x10-10
0.08
1.5x10-10
0.06
1.0x10-10 0.04
5.0x10-11 0.02

0.0 0.00
0 200 400 600 800 1000 1200 1400 0 20 40 60 80
Time, t (ps) Time, t (ps)
   
Fig. 4. Temporal variation of transmittance qntr and reectance qnref from the human breast subjected to single and train of 5 laser pulses for (a) and (b) t p 100 ps,
(c) and (d) t p 10 ps, and (e) and (f) t p 1 ps.

Table 1
Optical properties of normal breast at wavelengths l 500, 800 and 1100 nm (Bashkatov et al., 2011; Peters et al., 1990; Key et al., 1991).

Tissues l 500 nm l 800 nm l 1100 nm

1 1
b r (mm 1
) ss r (mm ) g b r(mm 1
) ss r (mm ) g b r (mm  1) ss r (mm  1) g

Normal breast tissues

Epidermis 71.307 70.6 0.9 36.971 36.8 0.9 29.217 29.2 0.9
Dermis 46.536 46.2 0.9 22.622 22.5 0.9 17.646 17.6 0.9
Fat 30.7452 30.455 0.971 31.013 30.936 0.976 31.6433 31.513 0.977
Fibroglandular 18.695 15.0 0.95 13.527 12.5 0.94 12.247 12.175 0.9384

understanding these normal and abnormal phenomena. These transfer. Moreover, thermal characteristics of patho-physiological
normal and abnormal physiological conditions also change with conditions, especially in case of cancer, keep changing with time and
interior and exterior conditions, thus inuences the rate of heat are a function of degree of abnormalities. The properties that are of
 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
interest to bioheat transfer are specic heat, thermal conductivity,
thermal diffusivity, thermal effusivity and volumetric perfusion rate.
These properties if measured in-vivo can accurately explain the heat
transfer phenomenon. However, due to the difculty in dealing with
living objects, most of the properties are obtained from in-vitro
measurements.
The temperature dependence of tissue thermophysical proper-
ties was determined by Valvano and colleagues in a series of in-
vitro measurements of human and animal tissues (Valvano et al.,
1985; Valvano and Chitsabesan, 1987; Kharalkar et al., 2008).
Copper and Trezek (1971) determined the thermophysical prop-
erties of human tissues based on correlations with percentage of
water contents. They suggested various empirical relations for
thermal conductivity, density and specic heat based on mass
fractions of water, protein and fat contents observed in the tissue.
Later, Olsrud et al. (1998) performed in-vitro measurement of
thermal conductivity and the water content of uterine tissue by
using the same relation as that used by Copper and Trezek (1971)
to calculate density and specic heat. Further, they carried
statistical tests (unpaired t-test and MannWhitney U-test) to
ascertain whether there is any signicant difference in the
thermal conductivity of myometrial with and without endome-
trium tissue. The thermophysical properties of normal, un-
perfused human brain at different temperature ranges are given
by Black and Hastings (1998). More thermophysical properties of
human and animal tissues are detailed in Hateld and Pugh
(1951), Klug (1874), Cohen (1977), and Tanasawa and Tanishita
(1984). Further, Erdmann et al. (1999) suggested few relations for
temperature dependent blood perfusion in human muscle, fat and
tumor using a non-linear heat transfer model. These are given as
(   perfusion term and effective thermal conductivity. They also intro-
2
0:45 3:55exp  T45
12 T r45 1C
omuscle 79
4 T 4 45 1C
(   tissues. The models developed by Weinbaum et al. evolved into more
2
0:36 0:36exp  T45
12 T r45 1C
of at
0:72 T 4 45 1C
8 detailed description of vascular geometry. These models added more
< 0:833
> T o37 1C
otumor 0:833T374:8 =5:483E 3 37 ZT r 45 1C 81
>
: understanding of the vascular geometry in the tissues and organs. As
0:416 T 445 1C
More information about various techniques used by research-
ers to determine thermophysical properties are detailed in
Hateld (1953), Bowman et al. (1975), Balasubramaniam and
Bowman (1977), Chen et al. (1981), Shitzer and Eberhart (1985),
Duke (1991), and Welch and Gemert (2011).
Similarly, optical properties of tissues and bio-uids vary with
physiological and patho-physiological activities, as well as due to
changes in composition. These properties are function of time
specically due to growth of abnormalities. Primarily, emission is
a function of local energy contents of the tissue, and is different
for tissues with and without abnormalities. The absorption of
incident energy by tissue depends on intensity of incident radia-
tion and density of absorbers/chromophores in the tissue, and is
indirectly depends on the wavelength, i.e. light absorbed by tissue
in longer wavelength is low. On the other hand, scattering
appears due to localized non-uniformity, concentration of scatter-
ing particles, scattering cross-section of the scatterers, material of
the particle (i.e., complex index of refraction) as well as on the
wavelength. Overall, the scattering is due to three separate
phenomenon (Modest, 2003), viz., diffraction (photons never
come in contact with particles but there direction of propagation
changes), reection (photons are reected back from the surface
of particle) and refraction (photons that penetrate into particles
are partial absorbed and reemerged into different directions)
123
provided the dual nature of photon is considered. Thus, the
scattering effect can either augment or attenuate intensity in a
particular direction. Furthermore, scattering due to uctuation of
refractive index at the interface and deep inside tissue can be
wavelength dependent which can be elastic and non-elastic
(Welch and Gemert, 2011). But, vast majority of photons are
scattered without changing wavelength due to which the scatter-
ing effects of tissue have much weaker wavelength dependence
and are considered mostly elastic. The properties of interest to
tissue-optics are absorption coefcient, scattering coefcient,
average cosine of phase function, reduced scattering coefcient,
total attenuation and effective attenuation coefcient.
The inuence of optical properties to radiative heat transfer
and various techniques to measure optical properties are detailed
in Welch and Gemert (2011) and Cheong et al. (1990). More
details on the optical properties of normal and abnormal human
tissues are available in an excellent review on optical properties
of tissues by Cheong et al. (1990). The radiative outputs in terms
of transmittance and reectance will be different for various
normal and abnormal tissue conditions, and can be identied by
unusual patterns when compared to normal conditions. On the
other hand, these optical parameters are obtained by converting
radiative outputs, viz., reected and transmitted rays into useful
data that characterizes light tissue interaction.
6. Conclusions
The concept of perfusion based model was rst introduced by
Pennes for vascularized tissues. Later, Chen and Holmes brought
modications to this model by including more precise description of
duced the concept of thermal equilibration length. Subsequently,
Weinbaum et al. introduced detailed three-layer models followed by
an equivalent one-equation continuum model for peripheral muscle
80 realistic model because of the consideration of countercurrent heat
exchange between closely spaced arteryvein pairs along with the
complexities to the biological heat transfer, but at the same time
brought the attention of researchers for the need to have better
a result, both the DIVA and the statistical models focused more on the
accurate modeling of blood vessels in vascularized tissue. In the
recent past, Mitra et al. brought the concept of phase lag in thermal
conduction, and Nakayama and Kuwahara emphasized the consid-
eration of porous media theory to make the bioheat transfer analysis
more appropriate. Subsequently, Shrivastava and Vaughan intro-
duced a new two-equation continuum model different from the
much discussed conventional bioheat models. Apart from the bioheat
transfer mechanisms, viz., conduction, perfusion and convection, the
thermal radiation also plays a vital role in maintaining the thermal
synergy in all living entities. Moreover, the use of laser in therapeutic
applications has warranted the study of radiative models in bioheat
transfer. To this effect, the combined continuum and radiative models
have been discussed for better understanding of temperature dis-
tribution in biological media during thermal therapy. Last but not the
least, the signicance of biological properties and their dependence
on physiological and patho-physiological factors have been discussed.
Therefore, from the standpoint of involved complexities in biological
systems, the present review on bioheat transfer serves as an
important foundation for many future analytical and computational
investigations. These areas include, but not limited to, biological
thermo-mechanics/neurophysiology of extreme conditions (i.e.,
during pain, relief, burns, thermal therapy and mechanical loading)
as well as extreme environments (i.e., in space, oceans and cryonics),
biomaterials, nanotechnology, biophotons, biophotonics, etc.
124 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
Reference
Absalan, H., SalmanOgli, A., Rostami, R., Maghoul, A., 2012. Simulation and
investigation of quantum dot effects as internal heat-generator source in
breast tumor site. J. Therm. Biol. 37, 490495.
Arkin, H., Xu, L.X., Holmes, K.R., 1994. Recent developments in modeling heat
transfer in blood perfused tissues. IEEE Trans. Biomed. Eng. 163, 97107.
Baish, J.W., Ayyaswamy, P.S., 1986. Small-scale temperature uctuations in
perfused tissue during local hyperthermia. J. Biomech. Eng. 108, 246250.
Baish, J.W., 1994. Formulation of a statistical model of heat transfer in perfused
tissue. J. Biomech. Eng. 116, 512527.
Baish, J.W., Ayyaswamy, P.S., Foster, K.R., 1986. Heat transport mechanism in
vascular tissues: a model comparision. J. Biomech. Eng. 108, 324331.
Baker, D., Agard, D.A., 1994. Kinetics versus thermodynamics in protein folding.
Biochemistry 33, 75057509.
Balasubramaniam, T.A., Bowman, H.F., 1977. Thermal conductivity and thermal
diffusivity of biomaterials: a simultaneous measurement technique. J. Bio-
mech. Eng. 99, 148154.
Banerjee, A., Ogale, A.A., Das, C., Mitra, K., Subramanian, C., 2005. Temperature
distribution in different materials due to short pulse laser irradiation. Heat
Transfer Eng. 26, 4149.
Barnes, R.B., 1968. Diagnostic thermography. Appl. Opt. 7, 16731686.
Bashkatov, A.N., Genina, E.A., Tuchin, V.V., 2011. Optical properties of skin,
subcutaneous, and muscle tissues: a review. J. Innov. Opt. Health Sci. 4, 938.
Black, J., Hastings, G., 1998. Handbook of Biomaterial Properties. Chapman and
Hall, Thomson Science, London.
Bouzida, N., Bendada, A., Maldague, X.P., 2009. Visualization of body thermo-
regulation by infrared imaging. J. Therm. Biol. 34, 120126.
Burch, A., Law, J., 1995. A method for estimating compressed breast thickness
during mammography. Br. J. Radiol. 68, 394399.
Bowman, H.F., Cravalho, E.G., Woods, M., 1975. Theory, measurement and applica-
tion of thermal properties of biomaterials. Annu. Rev. Biophys. 4, 4380.
Cattaneo, C., 1958. A form of heat conduction equation which eliminates the
paradox of instantaneous propagation. C. R. Acad. Sci. 247, 431.
Centigul, M.P., Herman, C., 2011. Quantication of the thermal signature of a
melanoma lesion. Int. J. Therm. Sci. 50, 421431.
Charny, C.K., 1992. Mathematical models of bioheat transfer. In: Cho, Y.S. (Ed.),
Advances in Heat Transfer. Academic press, Boston, pp. 91155.
Chato, J.C., 1980. Heat transfer to blood vessels. J. Biomech. Eng. 102, 110118.
Chato, J.C., 1981. Reections of the history of heat and mass transfer in
bioengineering. J. Biomech. Eng. 103, 97101.
Chen, M.M., Holmes, K.R., 1980. Microvascular contribution in tissue heat transfer.
Ann. N.Y. Acad. Sci. 335, 137142.
Chen, M.M., Holmes, K.R., Rupinskas, V., 1981. Pulse-decay method for measuring
the thermal conductivity of living tissues. J. Biomech. Eng. 103, 253260.
Cheong, W.F., Prahl, S.A., Welch, A.J., 1990. A review of the optical properties of
biological tissues. IEEE J. Quantum Electron. 26, 21662185.
Chu, C.M., Churchill, S.W., 1955. Representation of the angular distribution of
radiation scattered by a spherical particle. J. Opt. Soc. Am. 45, 958962.
Cohen, M.L., 1977. Measurement of the thermal properties of human skina
review. J. Invest. Dermatol. 69, 333338.
Copper, T.E., Trezek, G.J., 1971. Correlation of thermal properties of some human
tissue with water content. Aerosp. Med. 42, 2427.
Costello, J.T., Mclnerney, C.D., Bleakley, C.M., Selfe, J., Donnelly, A.E., 2012. The use
of thermal imaging in assessing skin temperature following cryotherapy: a
review. J. Therm. Biol. 37, 103110.
Das, C., Trivedi, A., Mitra., K., Vo-dinh, T., 2003. Experimental and numerical
analysis of short-pulse laser interaction with tissue phantoms containing
inhomogeneities. Appl. Opt. 42, 1619.
Das, K., Singh, R., Mishra., S.C., 2012. Numerical analysis for determination of the
presence of a tumor and estimation of its size and location in a tissue. J. Therm.
Biol. 38, 3240.
Dear, R.D., Brager, G.S., 1998. Developing an adaptive model of thermal comfort
and preference. ASHRAE Trans. 104, 145167.
Deng, Z., Liu, J., 2001. Blood perfusion based model for characterizing the
temperature uctuation in living tissues. Phys. A 300, 521530.
Despa, F., Orgill, D.P., Neuwalder, J., Lee, R.C., 2005. The relative thermal stability of
tissue macromolecules and cellular structure in burn injury. Burn 31, 568577.
Diederich, C.J., 2005. Thermal ablation and high-temperature thermal therapy:
overview of technology and clinical implementation. Int. J. Hyperther. 21,
745753.
Duke, F.A., 1991. Physical Properties of Tissue: A Comprehensive Reference Book.
Academic Press, London.
Erdmann, B., Lang, J., Seebass, M., 1999. Impact of nonlinear heat transfer on
temperature control in regional hyperthermia. IEEE Trans. Biomed. Eng. 46,
11291138.
Feng, Y., Fuentes, D., Hawkins, A., Bass, J.M., Rylander, N.M., 2009. Optimization
and real-time control for laser treatment of heterogeneous soft tissues.
Comput. Methods Appl. Mech. Eng. 198, 17421750.
Gage, A., Baus, J., 2002. Mechanisms of tissue injury in cryosurgery. Cryobiology
37, 171186.
Gisol, C.V., Wegner, C.B., 1984. Temperature regulation during exercise: old
concepts, new ideas. Exercise Sport Sci. Rev. 12, 339372.
Gonzalez, R.R., Halford, C., Keach, E.M., 2010. Environmental and physiological
simulation of heat stroke: a case study analysis and validation. J. Therm. Biol.
35, 441449.
Gupta, P.K., Singh, J., Rai, K.N., 2010. Numerical simulation for heat transfer in
tissues during thermal therapy. J. Therm. Biol. 35, 295301.
Hateld, H.S., 1953. An apparatus for measurement of the thermal conductivity of
biological tissue. J. Sci. Instrum. 30, 460461.
Hateld, H.S., Pugh, L., 1951. Thermal conductivity of human fat and muscle.
Nature 168, 918919.
Heisterkamp, J., Van-Hillegersberg, R., Ijzermans, J., 1999. Interstitial laser coagu-
lation for hepatic tumors. Br. J. Surg. 86, 293304.
Henyey, L.G., Greenstein, J.L., 1941. Diffuse radiation in the galaxy. Astrophys. J. 93
83, 7083.
Highnam, R.P., Bardy, J.M., Shepstone, B.J., 1998. Estimation of compressed breast
thickness during mammography. Br. J. Radiol. 71, 646653.
Howell, J.R., Siegel, R., Menguc, M.P., 2011. Thermal Radiation Heat Transferfth
edn CRC Press, London.
Huang, H.W., Chen, Z.P., Roemer, R.B., 1996. A countercurrent vascular network
model of heat transfer in tissues. J. Biomech. Eng. 118, 120129.
Huang, H.W., Shih, T.C., Liauh, C.T., 2010. Predicting effects of blood ow rate and
size of vessels in a vasculature on hyperthermia treatments using computer
simulation. Biomed. Eng. Online 9, 118.
Jaunich, M., Raje, S., Kim, K., Mitra, K., Guo, Z., 2008. Bio-heat transfer analysis
during short pulse laser irradiation of tissues. Int. J. Heat Mass Transfer. 51,
55115521.
Jiao, J., Guo, Z., 2009. Thermal interaction of short-pulsed laser focused beams with
skin tissues. Phys. Med. Biol. 54, 42254441.
Keller, K.H., Seiler, L., 1971. An analysis of peripheral heat transfer in man. J. Appl.
Physiol. 30, 779786.
Key, H., Davis, E.R., Jackson, P.C., Wells, P.N.T., 1991. Optical attenuation char-
acteristics of breast tissues at visible and near-infrared wavelengths. Phys.
Med. Biol. 36, 579590.
Kharalkar, N.M., Hayes, L.J., Valvano, J.W., 2008. Pulse-power integrated-decay
technique for the measurement of thermal conductivity. Meas. Sci. Technol.
19, 75104.
Kim, K., Guo, Z., 2007. Multi-time-scale heat transfer modeling of turbid tissues
exposed to short-pulsed irradiations. Comput. Methods Programs Biomed. 86,
112123.
Khaled, A.R.A, Vafai, K., 2003. The role of porous media in modeling ow and heat
transfer in biological tissues. Int. J. Heat Mass Transfer 46, 49895003.
Klug, F., 1874. Research on the conduction of heat in the skin. Z. Biol. 10, 73.
Kotte, A., Van-Leeuwen, G., Bree, J.D., Koijk, J.V., Crezee, H., Lagendijk, J., 1996.
A description of discrete vessel segments in thermal modelling of tissues.
Phys. Med. Biol. 41, 865884.
Kotte, A., Van-Leeuwen, G., Lagendijk, J., 1999. Modelling the thermal impact of a
discrete vessel tree. Phys. Med. Biol. 44, 5774.
Kutz, M., 2002. Standard Handbook of Biomedical Engineering and Design.
McGraw-Hill Professional, New York.
Lagendijk, J., 1982. The inuence of blood ow in large vessels on the temperature
distribution in hyperthermia. Phys. Med. Biol. 27, 1723.
Lagendijk, J., Schellekens, M., Schipper, J., Linden, P., 1984. A three-dimensional
description of heating patterns in vascularised tissues during hyperthermic
treatment. Phys. Med. Biol. 29, 495507.
Lemons, D.E., Chien, S., Crawshaw, L.I., Weinbaum, S., Jiji, L.M., 1987. Signicance of
vessel size and type in vascular heat transfer. Am. J. Physiol. 253, R128R135.
Lindahl, T., 1967. Irreversible heat inactivation of transfer ribonucleic acids. J. Biol.
Chem. 242, 19702173.
McLeod, K.J., Lesperance, L.M., Laramee, C., 2009. Inuence of indoor cooling on
heat balance and body weight gain in Americans over recent decades. J. Therm.
Biol. 34, 385390.
Mishra, S.C., Chungh, P., Kumar, P., Mitra, K., 2006. Development and comparison of
the DTM, the DOM and the FVM formulations for the short-pulse laser transport
through a participating medium. Int. J. Heat Mass Transfer 49, 18201832.
Mitra, K., Kumar, S., Vedavarz, A., Moallemi, M.K., 1995. Experimental evidence of
hyperbolic heat conduction in processed meat. ASME J. Heat Transfer 117,
568573.
Mitchell, J.W., Myers, G.E., 1968. An analytical model of the counter-current heat
exchange phenomena. Biophys. J. 8, 897911.
Modest, M.F, 2003. Radiative Heat Transfer, Second edn. Academic Press, New York.
Mohammed, Y., Verhey, J.F., 2005. A nite element method to simulate laser
interstitial thermo therapy in anatomical inhomogeneous regions. Biomed.
Eng. Online 4, 116.
Mooibroek, J., Lagendijk, J., 1991. A fast and simple algorithm for the calculation of
convective heat transfer by large vessels in three-dimensional inhomogeneous
tissues. IEEE Trans. Biomed. Eng. 38, 490501.
Muthukumaran, R., Mishra, S.C., 2009. Analysis of the transport of a train of short-
pulse radiation of Gaussian temporal prole through a 2-D participating
medium. Heat Transfer Eng. 30, 11971207.
Muthukumaran, R., Mishra, S.C., 2008. Thermal signatures of a localized inhomo-
geneity in a 2-D participating medium subjected to an ultra-fast step-pulse
laser wave. J. Quant. Spectrosc. Radiat. Transfer 109, 705726.
Nakayama, A., Kuwahara, F., 2008. A general bioheat transfer model based on the
theory of porous media. Int. J. Heat Mass Transfer 51, 31903199.
Okajima, J., Maruyama, S., Takeda, H., Komiya, A., 2009. Dimensionless solutions
and general characteristics of bioheat transfer during thermal therapy. J.
Therm. Biol. 34, 377384.
 A. Bhowmik et al. / Journal of Thermal Biology 38 (2013) 107125
Olsrud, J., Friberg, B., Ahlgren, M., Persson, B.R., 1998. Thermal conductivity of
uterine tissue in vitro. Phys. Med. Biol. 43, 23972406.
Overgaard, J., 1977. Effect of hyperthermia on malignant cells in vivo: a review and
a hypothesis. Cancer 39, 26372646.
Pal, G., Basu, S., Mitra, K., Vo-Dinh, T., 2006. Time-resolved optical tomography
using short-pulse laser for tumor detection. Appl. Opt. 45, 62706282.
Peters, V.G., Wyman, D.R., Patterson, M.S., Frank, G.L., 1990. Optical properties of
normal and diseased human breast tissues in the visible and near infrared.
Phys. Med. Biol. 35, 13171334.
Pennes, H.H., 1948. Analysis of tissue and arterial blood ow temperatures in the
resting forearm. J. Appl. Physiol. 1, 93122.
Roeztel, W., Xuan, Y., 1998. Transient response of the human limb to an external
stimulus. Int. J. Heat Mass Transfer 41, 229239.
Roemer, R.B., Cetas, T.C., 1984. Application of bioheat transfer simulations in
hyperthermia. Cancer Res. 44, 4788s4798s.
Roemer, R.B., Oleson, J.R., Cetas, T.C., 1985. Oscillatory temperature response to
constant power applied to canine muscle. Am. J. Physiol. 249, R153R158.
Rossi, M.R., Rabin, Y., 2007. Experimental verication of numerical simulations of
cryosurgery with application to computerized planning. Phys. Med. Biol. 52,
45534567.
Rubinsky, B., 1999. Heat transfer in biomedical engineering and biotechnology.
In: Proceedings of tht 5th ASME/JSME Joint Thermal Engineering Conference,
(keynote article).
Ruell, P.A., Thompson, M.W., Hoffman, K.M., 2009. Heat shock proteins as an aid in
the treatment and diagnosis of heat stroke. J. Therm. Biol. 34, 17.
Sampson, D.D., Hillman, T.R., 2004. Optical coherence tomography. In: Palumbo,
G., Pratesi, R. (Eds.), Lasers and Current Optical Techniques in Biology. Royal
Society of Chemistry, Cambridge, pp. 549554.
SantaLucia, J., Hicks, D., 2004. The thermodynamics of DNA structural motifs.
Annu. Rev. Biophys. Biomol. Struct. 33, 415440.
Shih, T.C., Kou, H.S., Liauh, C.T., Lin, W.L., 2002. Thermal models of bioheat transfer
equations in living tissue and thermal dose equivalence due to hyperthermia.
Biomed. Eng. Appl. Basis Commun. 14, 4150.
Shitzer, A., Eberhart, R.C., 1985. Heat Transfer in Medicine and Biology. Plenum,
New York.
Shrivastava, D., Roemer, R.B., 2004. A general analytical derivation of a new, source
term dependent 2-D Poisson conduction shape factors. Int. J. Heat Mass
Transfer 47, 42934300.
Shrivastava, D., McKay, B., Roemer, R.B., 2005. An analytical study of heat transfer
in nite tissue with two blood vessels and uniform Dirichlet boundary
conditions. J. Heat Transfer 127, 179188.
Shrivastava, D., Roemer, R.B., 2005. An analytical study of Poisson conduction
shape factor for two thermally signicant vessels in nite, heated tissue. Phys.
Med. Biol. 50, 36273641.
Shrivastava, D., Roemer, R.B., 2006. Readdressing the issue of thermally signicant
blood vessels using a countercurrent vessel network. J. Biomech. Eng. 128,
210216.
Shrivastava, D., Vaughan, J.T., 2009. A generic bioheat transfer thermal model for a
perfused tissue. J. Biomech. Eng. 131, 112.
Steketee, J., 1973. Spectral emissivity of skin and pericardium. Phys. Med. Biol. 18,
686694.
Stryer, L., 1988. Biochemistrythird edn W.H. Freeman and Company, New York.
Takeda, H., Maruyama, S., Okajima, J., Alba, S., Komiya, A., 2009. Development and
estimation of a novel cryoprobe utilizing the peltier effect for precise and safe
cryosurgery. Cryobiology 59, 275284.
Tzou, D.Y., 1995a. A unied eld approach for heat conduction from macro- to
micro-scales. ASME J. Heat Transfer 117, 816.
Tzou, D.Y., 1995b. Experimental support for the lagging behaviour in heat
propagation. J. Thermophys. Heat Transfer 9, 686693.
Tanasawa, I., Tanishita, K., 1984. Genuine and pseudo-thermophysical properties
of biological media. Int. J. Thermophys. 5, 301311.
125
Toussaint, O., Schneider, E.D., 1998. The thermodynamics and evolution of
complexity in biological systems. Comp. Biochem. Physiol. Mol. Integrative
Physiol. 120, 39.
Valvano, J.W., Chitsabesan, B., 1987. Thermal conductivity and diffusivity of
arterial wall and atherosclerotic plaque. Lasers Life Sci. 1, 219229.
Valvano, J.W., Cochran, J.R., Diller, K.R., 1985. Thermal conductivity and diffusivity
of biomaterials measured with self-heated thermistors. Int. J. Thermophys. 6,
301311.
Van-Gemert, M.J.C., Jacques, S.L., Sterenborg, H.J.C.M., Star, W.M., 1989. Skin optics.
IEEE Trans. Biomed. Eng. 36, 11461153.
Van-Hillegersberg, R., 1997. Fundamentals of laser surgery. Eur. J. Surg. 1, 312.
Van-Leeuwen, G., Kotte, A., Crezee, J., Lagendijk, J., 1997. Tests of the geometrical
description of blood vessels in a thermal model using counter-current
geometries. Phys. Med. Biol. 42, 15151532.
Varon, J., Acosta, P., 2008. Therapeutic hypothermia: past, present and future.
Chest J. 133, 12671274.
Vernotte, P., 1958. Les paradoxes de la thorie continue de lquation de la chaleur. C.
R. Acad. Sci. 246, 3154.
Wang, L., Zhou, X., Wei, X., 2008. Heat Conduction, Mathematical Models and
Analytical Solutions, rst edn. Springer  Verlag LLC, New York.
Weinbaum, S., Jiji, L., Lemons, D.E., 1984a. Theory and experiment for the effect of
vascular microstructure on surface tissue heat transferpart I: anatomical
foundation and model conceptualization. J. Biomech. Eng. 106, 321330.
Weinbaum, S., Jiji, L., Lemons, D.E., 1984b. Theory and experiment for the effect of
vascular microstructure on surface tissue heat transferpart II: model
formulation and solution. J. Biomech. Eng. 106, 331341.
Weinbaum, S., Jiji, L., 1985. A New simplied bioheat equation for the effect of
blood ow on local average tissue temperature. J. Biomech. Eng. 107, 131141.
Welch, A.J., Gemert, M.J.C., 2011. Optical and Thermal Response of Laser-irradiated
Tissue, second edn. Springer, New York.
Whelan, W.M., Davidson, S.R.H., Chin, L.C.L., Vitkin, I.A., 2005. A novel strategy for
monitoring laser thermal therapy based on changes in optothermal properties
of heated tissues. Int. J. Thermophys. 26, 233241.
Wittwer, C.T., Fillmore, G.C., Garling, D.J., 1990. Minimizing the time required for
DNA amplication by efcient heat transfer to small samples. Anal. Biochem.
186, 328331.
Wnek, G.E., Bowlin, G.L., 2005. Encyclopedia of Biomaterials and Biomedical
Engineering. Marcel Dekker Inc., New York.
Wulff, W., 1974. The energy conservation equation for living tissue. IEEE Trans.
Biomed. Eng. 21, 494497.
Wu, Y.L., Weinbaum, S., Jiji, L.M., 1993. A new analytic technique for 3-D heat
transfer from a cylinder with two or more axially interacting eccentrically
embedded vessels with application to countercurrent blood ow. Int. J. Heat
Mass Transfer 36, 10731083.
Xu, F., Lu, T., Seffen, K.A., 2008. Dual-phase-lag model of skin bioheat transfer.
Proc. Int. Conf. Biomed. Eng. 1, 505511.
Xu, F., Lu, T.J., Seffen, K.A., Ng, E.Y.K., 2009. Mathematical modeling of skin bioheat
transfer. Appl. Mech. Rev. 62, 05080105080135.
Yue, K., Zheng, S., Luo, Y., Zhang, X., Tang, J., 2011. Determination of the 3D
temperature distribution during ferromagnetic hyperthermia under the inu-
ence of blood ow. J. Therm. Biol. 36, 498506.
Zhu, L., Diao, C., 2001. Theoretical simulation of temperature distribution in the
brain during mild hypothermia treatment for brain injury. Med. Biol. Eng.
Comput. 39, 681687.
Zhu, L., Lemons, D.E., Weinbaum, S., 1996. Microvascular thermal equilibration in
rat cremaster muscle. J. Biomech. Eng. 24, 109123.
Zhu, M., Weinbaum, S., Jiji, L.M., 1990. Heat exchange between unequal counter-
current vessels asymmetrically embedded in a cylinder with surface convec-
tion. Int. J. Heat Mass Transfer 33, 22752283.
Zhu, M., Weinbaum, S., Jiji, L.M., Lemons, D.E., 1988. On the generalization of the
WeinbaumJiji bioheat equation to microvessels of unequal size; the relation
between the near eld and local average tissue temperaturesJ. Biomech. Eng.
110, 7481.