Journal of Thermal Biology 58 (2016) 8090

Contents lists available at ScienceDirect

Journal of Thermal Biology

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

Thermal analysis of induced damage to the healthy cell during RFA of

breast tumor
Sundeep Singh a, Arka Bhowmik b, Ramjee Repaka a,n
a
Department of Mechanical Engineering, Indian Institute of Technology Ropar, Rupnagar, Punjab 140001, India
b
Mechanical and Aerospace Engineering Dept., University of California, 420 Westwood Plaza, Engineering IV, Los Angeles, CA 90095, USA

art ic l e i nf o a b s t r a c t

Article history: Effective pre-clinical computational modeling strategies have been demonstrated in this article to enable
Received 27 July 2015 risk free clinical application of radiofrequency ablation (RFA) of breast tumor. The present study
Received in revised form (a) determines various optimal regulating parameters required for RFA of tumor and (b) introduces an
6 January 2016
essential clinical monitoring scheme to minimize the extent of damage to the healthy cell during RFA of
Accepted 7 April 2016
tumor. The therapeutic capabilities offered by RFA of breast tumor, viz., the rise in local temperature and
Available online 20 April 2016
induced thermal damage have been predicted by integrating the bioheat transfer model, the electric eld
Keywords: distribution model and the thermal damage model. The mathematical model has been validated with the
Pre-clinical modeling experimental results available in the literature. The results revealed that, the effective damage of tumor
Radiofrequency ablation
volume sparing healthy tissue essentially depends on the voltage, the exposure time, the local heat
Breast cancer
distribution, the tumor stage and the electrode geometric conguration. It has been conrmed that, the
Finite element analysis
Bioheat equation assessment of damage front can accurately determine the extent of damage as compared to the thermal
front. The study further evaluates the damaged healthy and tumor volumes due to RFA of different stages
of breast cancer. The assessment of cell survival and damage fractions discloses the propensity of re-
appearance/healing of tumor cells after treatment.
& 2016 Elsevier Ltd. All rights reserved.

1. Introduction
Breast cancer is one of the common types of cancer found
among the female all over the world (DeSantis et al., 2011). At
present, commonly used breast cancer treatment techniques are
surgery, chemotherapy, radiation therapy, hormonal therapy, tar-
geted cancer therapy, etc. The surgical procedures are complex,
time consuming and relatively expensive with poor cosmetic re-
sults, while radiation therapy and chemotherapy lead to major
health related side effects. Thus, in recent times new treatment
modalities have been developed by scientists which are relatively
fast, having less health related side effects and comparatively in-
expensive with better cosmesis. The newly developed techniques
are either based on rising the temperature, viz., laser (Megan et al.,
2008; Bhowmik et al., 2015a), focused ultrasound (Bhowmik et al.,
2015a), radiofrequency (Haemmerich et al., 2003; Ng and Jamil,
2014; Jamil and Ng, 2015), microwave (Wu et al., 2015) and
magnetic eld (Jordan et al., 2001) assisted thermal ablation
methods, or lowering the temperature, i.e., cryotherapy (Deng and
n
Correspondence to: Department of Mechanical Engineering, Indian Institute of
Technology Ropar, Nangal Road, Rupnagar-140001, Punjab,
Tel.: 91 1881 242159; fax: 91 1881 223395.
E-mail address: ramjee.repaka@gmail.com (R. Repaka).
Liu, 2004) for irreversible cellular damage. Out of all these treat-
ment modalities, RFA has received much attention as a minimally
invasive strategy for the treatment of unresectable primary he-
patic tumors (Tungjitkusolmun et al., 2002), and promising results
have also been obtained in treating tumors in lung, bone, brain,
kidney, prostate gland, pancreas, breast, coronary vascular dis-
eases (Mirza et al., 2001; Fornage et al., 2004; Morady et al., 1993),
etc.
Breast is an ideal case for RF application because of its super-
cial location on the thorax and the absence of intervening organs.
First clinical application of RFA for the treatment of human breast
cancer was performed by Jeffrey et al., (1999), on 5 women (aged
3866 years) with locally advanced (stage III) invasive breast
cancer (tumor size 47 cm). The study reported zero complications
with success in 4 out of 5 patients. Subsequently, few more studies
explored the feasibility of RFA for the treatment of breast cancer
such as Fornage et al., (2004), Burak et al., (2003), and Hayashi
et al., (2003). Especially, the application of RF technique in patients
with early/small breast carcinoma seems to be an appealing
treatment method with improved cosmesis, and hence a compe-
titive method compared to other conventional and new treatment
India.
methods (Grotenhuis et al., 2013). Thus, the present study is fo-
cused on the application of RF technique for the thermal ablation
of small/early and localized breast cancers.

http://dx.doi.org/10.1016/j.jtherbio.2016.04.002
0306-4565/& 2016 Elsevier Ltd. All rights reserved.
 S. Singh et al. / Journal of Thermal Biology 58 (2016) 8090
Nomenclature
A frequency factor (s  1)
c specic heat (J kg  1 K  1)
C concentration of damaged/undamaged cell
d diameter of electrode and trocar (mm)
D diameter of tumor (mm/cm)
E electric eld (V m  1)
Ea activation energy (J mol  1)
f radiofrequency (Hz)
h depth (mm/cm)
J current density (A m  2)
k thermal conductivity (W m  1 K  1)
Q m metabolic heat generation (W m  3)
Q ext external power source (W m  3)
R universal gas constant ( 8.314 J mol  1 K  1)
t time (s)
T temperature
V electric potential (V)/volume (mm3)
x, y, z position/coordinates
Greek symbols
thermal diffusivity
density (kg m  3)
s electrical conductivity (S m  1)
In an RF technique an electrode is inserted into the tissue by a
guiding mechanism and separate grounding pads are used on the
outer surface of the ablation domain to establish an electric circuit.
Once the electrode tips are accurately placed in the ablation zone
of the tumor and the circuit is closed, alternating current (in the
frequency range of 375500 kHz) is supplied to the electrode. The
established electric elds between the electrode tips and the
grounding pads cause the conduction of current from the elec-
trode tip to the tissue. The oscillating electric and magnetic elds
at the frequency of the radio waves cause the ions within the
tissue to agitate due to a change in the direction of current. The
agitation of ions results in the frictional/resistive heating of tissues,
since, biological tissues are poor conductors with high impedance
value. Due to the large value of tissue impedance, the resistive
heating is dominant within few millimeters of tissue surrounding
the electrode. But, the actual thermal ablation zone within the
tumor is mainly due to thermal conduction of heat to the per-
ipheral tissue. The rise in local temperature of tumor tissue above
the threshold value for a certain exposure time causes irreversible
cellular damage (i.e., damage integral, Z1). In general, the
coagulation (i.e., the initial phase of damage) begins when the
temperature of tissue rises to 4345 C but, the desired damage
( Z1) of the entire ablation zone can be achieved at a higher
temperature, i.e., above 50 C. The rapid heating or excessive
heating above 100 C causes boiling, vaporization and carboniza-
tion which reduces the ability of nearby tissue to conduct elec-
tricity due to the rise in impedance and limits the thermal diffu-
sion process. To avoid high ablation temperature within the tumor,
protocols like increasing the electrode surface area and pulsing the
input power have been found to be efcient. Further, the strategy
of permeating saline solution in the ablation zone improves the
effective electrical conductivity. At the same time, it increases the
probability of irregular necrotic zones.
Due to the associated complication during RFA of tumor, many
studies in the recent time used computational simulation to pre-
dict the outcome of the ablation process (Tungjitkusolmun et al.,
81
b blood perfusion (m3 s  1 m  3 tissue)
thermal damage parameter
necrosis time (s)
Subscripts
a arterial blood
b blood
c core
D damaged
e electrode tip
f nal value
s tissue/trocar base surface
t tissue
tb trocar base
tt trocar tip
UD undamaged
0 initial value
Abbreviations
AJCCS America joint committee of cancer staging
PBE Pennes bioheat equation
RFA radiofrequency ablation
2002; Ekstrand et al., 2005; Quaranta et al., 2007). In addition, a
few studies focused on estimating the optimal input conguration
during RFA, viz., voltage and frequency, and predicted the outcome
of induced damage due to combinations of various (a) thermo-
physical properties of tumor and healthy tissues (Jamil and Ng,
2013b) or (b) geometric congurations (Jamil and Ng, 2013c,
2013d). Jamil and Ng (2013b) employed Taguchi design of ex-
periments, involving the Analysis of Variance (ANOVA) on the data
obtained from experiments to quantify the effect of two input
tunable parameters, voltage and frequency and four thermo-phy-
sical properties, the metabolic heating rate in healthy tissue and
tumor, the blood perfusion rate in healthy tissue and tumor. The
study suggested that the voltage and frequency were crucial
parameters that had a direct effect on the treatment outcome.
Thus, RFA of tissue entails to proper treatment plan, including
careful consideration of voltage and frequency. Similarly, Jamil and
Ng (2013d) also employed ANOVA on experimental data to
quantify the effect of voltage and frequency, along with four
geometric parameters, viz., tumor size, tumor location, electrode
size and relative position of the electrodes with respect to the
tumor. The study by Jamil and Ng (2013d) revealed that the reli-
able treatment during RFA largely depends on (a) the optimum
conguration of input and geometric parameters, and (b) the
precision with which thermal damage is imparted to the tumor
and migrating cells within the normal tissue. More recently, Ng
and Jamil (2014) evaluated the efcacy of RFA due to varying
electrical conductivity, thermal conductivity, and blood perfusion
rate corresponding to different types of tissues. The sensitivity of
such variation was tested using Taguchi's design of experimental
method. Particularly, for temperature controlled mode the thermal
conductivity had an additive effect on the ablation volume. While,
the blood perfusion and electrical conductivity had an inverse
effect on the ablation volume. Overall, the computer simulations
facilitated the researchers to make a predictive assessment of as-
sociated queries regarding RFA of tumor based on the hypothesis
of unforeseeable events.
82 S. Singh et al. / Journal of Thermal Biology 58 (2016) 8090
In this work, the performance of the single tip electrode for RFA
of small cancerous tumor within the breast have been predicted,
since earlier in-vivo studies have claried that RFA fails to ablate
large tumor volume completely. In spite of the tunable parameters
discussed in the literature survey, the exposure time also needed
to be optimized to determine the efcacy of thermal damage. The
same was lacking in earlier investigations. Thus, this study de-
termined the optimal exposure time along with voltage for dif-
ferent sizes of tumors during RFA. In addition, the study also
showed the possibility of connecting the damage parameter of the
radial tumor-healthy junction with that of the electrode-tip tem-
perature with the intention to quantify the accuracy of damage
based on the monitoring parameter, e.g., electrode-tip tempera-
ture. The correlations that link the radial damage front with the
electrode-tip temperature is a part of our extended work on the
RFA of breast tumor. The same will be presented in the imminent
publication. Overall, the objectives of the present study are of
three folds, viz., (a) introducing an effective pre-clinical modeling
strategy for the hassle free RFA procedure, (b) dening essential
regulating parameters for preferential thermal damage of tumor
volume sparing healthy tissue, and (c) revealing the role of mon-
itoring the electrode tip temperature to optimize the extent of
damage. Due to lack of any experimental data (or rise in tem-
perature prole) in the literature for RFA of breast tumor, the study
validated the present mathematical model with the results avail-
able on RFA of bovine liver tissue (Sethuraman et al., 2014).
(a)
(b)
Fig. 1. (a) Breast model with tumor stages and (b) trocar model with mono-polar electrode.
2. Problem denition and mathematical modeling
2.1. Problem denition
The problem considers a three dimensional model of breast
(Fig. 1(a)), which has been constructed based on the anatomical
details (Ng and Sudharsan, 2001) available in the literature. Elec-
trical, thermal and thermo-physical properties of different mate-
rials (Ekstrand et al., 2005; Ng and Sudharsan , 2001; Bhowmik
et al., 2014; Bezerra et al., 2013; Jossinet, 1996; Andreuccetti et al.,
1997; e-funda, 2014; Teon, 2014) have been depicted in Table 1.
Electrical properties of tissue presented in Table 1 have been
considered at a radio-frequency of 500 kHz. The dielectric con-
stant, i.e., r (relative permittivity) has been neglected, since the
dielectric losses caused by rotation of molecules in the electric
eld are negligible for frequency below 1 MHz (Ekstrand et al.,
2005). The study considered stainless steel properties for trocar
electrode as well as connecting shaft (e-funda, 2014), and Teon
properties for the insulating material (Teon, 2014). The three
dimensional model for trocar (Ekstrand et al., 2005) with one
mono-polar electrode has been shown in Fig. 1(b). Since the ab-
lation regime/size of tumors are small, thus, the study considered
a trocar with one mono-polar electrode instead of multi-tined
electrodes or bipolar electrodes.
Different sizes of early tumor volume within the breast (Fig. 1
(a)) have been modeled based on the staging guidelines given by
American Joint Committee for Cancer Staging (AJCCS) (Greene
et al., 2006). Breast cancer is mostly originated at the upper outer
 S. Singh et al. / Journal of Thermal Biology 58 (2016) 8090 83

Table 1
Electrical as well as thermal properties of breast tissue, tumor, electrode and trocar insulator and thermo-physical properties of breast tissue and tumor (Ekstrand et al.,
2005; Ng and Sudharsan, 2001; Bhowmik et al., 2014; Bezerra et al., 2013; Jossinet, 1996; Andreuccetti et al., 1997; e-funda, 2014; Teon, 2014).

Tissue/electrode Electrical properties Thermo-physical properties

Properties s (S m  1) c (J kg  1 K  1) k (W m  1 K  1) (kg m  3) b (m3 s  1 m  3) Q m (W m  3)

Breast fata 0.0560b 3000c 0.21d 920c 0.1796  10  3d 400d

Glandular tissuea 0.4274b 3000e 0.48d 1080e 5.3908  10  4gvnd 700d
Musclea 0.4459f 3800g 0.48d 1085g 5.3908  10  4gnd 700d
Tumora 0.4386b 3500d 0.48d 1080e 0.011h 1400h
Electrode (stainless steel) 9.804  105bmi 500i 36.7i 8100i
Trocar tip (stainless steel) 9.804  105i 500i 36.7i 8100i
Trocar base (Teon) 1.0  10  16j 1010j 0.23j 2190j

a
Electrical properties of tissue obtained at 500 kHz.
b
Jossinet (1996).
c
Ekstrand et al. (2005).
d
Ng and Sudharshan (2001).
e
Bezerra et al. (2013).
f
Andreuccetti et al. (1997).
g
Bhowmik et al. (2014).
h
Assumed.
i
e-funda (2014).
j
Teon (2014).

portion of the breast within the glandular tissue (Greene et al.,
2006). Therefore, the present study has modeled different stages
of breast tumor volume within the upper outer portion of the
breast in the glandular tissue (Fig. 1(a)). The tumor diameter (D)
has been varied within the range given for three stages, viz., T1mic
(D r0.1 cm), T1a (0.1 cm oDr 0.5 cm) and T1b
(0.5 cm oDr 1 cm) as suggested by AJCCS (Greene et al., 2006).
According to earlier studies (Ng and Sudharsan, 2001; Vaupel,
1998), the blood ow in breast tumor has been found to be in the
range of 880 ml/100 g per min. Therefore, the study considered
the blood perfusion within the tumor to be 60 ml of blood/100 g of
the tissue per min (Ng and Sudharsan, 2001). In order to obtain
the equivalent blood ow/perfusion in the tumor in m3 of blood
per m3 of tissue per second, the unit 1 ml of blood/100 g of tissue
per min is converted to 1 m3 of blood/m3 of tissue per second by
using the density of tumor tissue given in Table 1. Therefore, 60 ml
blood/100 g tissue per min will be equivalent to blood perfusion of
0.011 m3 of blood/m3 of tissue per second. The metabolic heat
generation per m3 of tumor volume has been assumed twice that
of the healthy cell (i.e., glandular tissue). This is mainly due to the
fact that the considered tumor volume is small in size, i.e., not
more than 1 cm diameter. Metabolic heat generation of breast
tumor size more than 1 cm diameter can be evaluated using the
relation suggested in (Ng and Sudharsan, 2001). The model as-
sesses early/small melanoma volume within the breast, and
therefore the sink effect due to blood vessels is assumed to be
negligibly small. Since, early melanoma volumes are nutrient de-
cient (Frangioni, 2008) as the supply blood vessels fails to extend
to the tumor from chest wall. However, for relatively larger de-
velopment stages, the presence of nearby blood vessels causes a
signicant reduction in the local temperature and formation of
unwanted hot spots/non-uniformity (Tungjitkusolmun et al.,
2002). The consideration of heterogeneous vascular architecture
and its effect on tissue temperature distribution during RFA had
been reported by Tungjitkusolmun et al. (2002); Singh et al.
(2015).
2.2. Mathematical modeling
Tt cb b Q m Q ext
= t 2Tt b Tt Ta + +
t t ct t ct t ct (1)
where t, ct and t are the density, the specic heat and the
thermal diffusivity of the tissue, respectively, t is expressed as kt
/tct,b, b and cb are the perfusion rate, the density and the
specic heat of the blood, respectively, Q m is the volumetric heat
generation due to metabolism, Q ext is the volumetric heating due
to external power source, Ta is the arterial blood temperature
(37 C), Tt is the unknown local tissue temperature, t is the time,
subscripts t and b represent tissue and blood, respectively. It is
worth-noting here that, the Pennes perfusion based model is
capable to estimate the temperature distribution accurately in
vascularized tissue with thermally insignicant vessels. However,
this model overestimates the local temperature for a relatively
large diameter vessels (Bhowmik et al., 2013). Therefore, the effect
of vascular ow should not be ignored while estimating the local
temperature of tumor near large diameter vessels. As discussed
earlier, in the present work large diameter vessels have been
considered to be far from the small volume breast tumor. In such a
case, the only interstitial perfusion effect of blood is dominant,
which can be accurately modeled using Pennes bioheat model.
Pennes bioheat model has been used in a wide range of applica-
tions to accurately estimate the heat transport phenomenon in
perfused biological tissues and organs, viz., RF hyperthermia
(Tungjitkusolmun et al., 2002; Jamil and Ng, 2015), laser-tissue
ablation (Megan et al., 2008; Paul et al., 2014; Bhowmik et al.,
2015a), thermal response due to fat thickening (Bhowmik et al.,
2015b), whole body thermoregulation under non-anesthetized
condition (Vallez et al., 2016), characterization of tumor (Mitra and
Balaji, 2010; Jamil and Ng, 2013a, 2013b, 2013c, 2013d; Ng and
Jamil, 2014; Das and Mishra, 2015), optimization of treatment
parameters (Jamil and Ng, 2013b, 2013c, 2013d), etc.
During RFA, volumetric heating of local tissue is mainly due to
the conduction losses (resistive heating) caused by applied electric
elds to the resistive material, i.e., tissue. Therefore, the absorbed
power density or the external power source due to resistive
heating can be written as (Ekstrand et al., 2005)

The Pennes bioheat equation (PBE) has been considered for
evaluating the change in temperature within the breast during
RFA. The PBE for perfused tissue domain can be written as (Pennes,
1998)
Q ext = JE = E2 (2)
where J is the current density (A m  2), E is the electric eld
(V m  1) and s is the electrical conductivity (Sm  1). The current
density (J) and the electric eld (E) are the vector quantities.
84 S. Singh et al. / Journal of Thermal Biology 58 (2016) 8090
Further, it has been assumed that the real part admittance (i.e.,
electrical conductivity) is contributing to the resistive heating,
since, the imaginary part has negligible effect.
A quasi-static electrical model has been used to demonstrate
the electric eld distribution within the tissue due to the applied
voltage, and can be given by Laplace equation as (Ekstrand et al.,
2005)
( V ) = 0
where V is the electric potential (V). The electric eld (E) and the
current density (J) generated within the tissue can be written as
E = V ; J = E
In the above mentioned equations, the electrical conductivity
(s) has been considered to be independent of temperature, since,
peak temperatures of the tissue in the study have been found to be
well below the charring temperature ( 100 C). Since, earlier
study (Chang, 2003) has revealed that, the difference in computed
temperature with and without the temperature dependent elec-
trical conductivity is approximately 58%, below the charring
temperature.
It is a well-known fact that, the blood perfusion (b) within the
tumor is more than the surrounding healthy tissue and varies with
size. The increase in blood perfusion within the tumor causes a
greater cooling effect as can be interpreted from Eq. (1). Hence, the
ablation temperature and the time required for ablating tumor is
considerably more compared to the healthy tissues. The threshold
temperature at which the onset of cellular damage manifests
within the tumor and within the healthy tissue has been con-
sidered to be at 45 C, but the desired damage ( Z1) of the entire
tumor takes place at relatively higher temperature. Further, during
the ablation blood perfusion changes due to gradual necrosis of
cells with time and is mainly due to leftover undamaged cells. The
perfusion is assumed to be increasing initially due to vasodilation
of capillaries caused by heating of perfused tissue (Abraham and
Sparrow, 2007) and later decreases with time/induced damage
given as
b,0 for (t ) 0
b (t ) = b,0 1 + 25 (t ) 260 (t )2 for 0 < (t ) 0.1
b,0 exp [ (t )] for (t ) > 0.1
where b,0 is the constant blood perfusion of tissue (given in Ta-
ble 1) and (t) is the induced thermal damage. The rst relation in
Eq. (5) suggests that, all cells in the tissue are undamaged and
perfusion remains same. The second relation exhibits a small in-
crease in perfusion during the initial damage due to vasodilation of
capillaries caused by heating of perfused tissue. The third relation
represents ceasing of perfusion within the tissue in an exponential
manner. It is worth-noting here that, perfusion ceases at a com-
paratively large value of , i.e., higher than 1.
The induced thermal damage eld, (t), within the tissue is
equivalent to the rate of protein denaturation for a period of ex-
posed time above the threshold temperature for cellular damage,
which can be expressed using rst-order Arrhenius rate
equation (Henriques, 1947) as
t manual adjustment of voltage has been achieved by trial-and-er-
(t ) = = 0 A exp Ea/RT ( ) d for ( =t tT ) > 0
where is the necrosis time (i.e., ttT), tT is the ablation onset/
threshold temperature, A is the frequency factor, Ea is the activa-
tion energy and R is the universal gas constant
( 8.314 J mol  1 K  1). In the present study, A and Ea for the bulk
tissue domain have been considered as 1.8  1051 s  1 and
327,000 J mol  1, respectively (Jia et al., 2007). It should be noted
here that, induced thermal damage is zero for r0.
The concentration/volume of undamaged (CUD) and damaged
(CD) cells over time can be evaluated from the knowledge of
thermal damage. The concentration of undamaged cells (CUD) over
time can be determined using the relation
CUD (t )
= exp [(t )]
C0 (7)
(3)
where CUD (t) is the concentration of remaining undamaged cells
over time, C0 is the concentration of undamaged cells before the
onset of ablation, i.e., r 0. All cells within the tissue before the
onset of ablation have been assumed to be undamaged, i.e., 100%
(4)
concentration (or C0 1.0). Thus, the concentration of damaged
cells (CD) over time can be easily determined using the relation
CD (t )
= 1 exp [(t )]
C0 (8)
where CD (t) is the concentration of damaged cells over time. From
Eqs. (8), 1.0 means  63% of the tissue molecules are damaged,
which is assumed to produce desired damage as suggested by
(Henriques, 1947). The damage integral, 1.0 or 63% of induced
thermal damage to tissue molecules is sufcient to restrict any
reversible change and to initiate natural death of cells.
2.3. Boundary and initial conditions
The initial voltage of the entire tissue domain before the appli-
cation of RF energy has been considered to be zero. The electrical
boundary condition for the outside surface of breast has been set to
be V0, since the outside surface is grounded using a grounding
pad. An electrical insulation boundary condition (n  j 0) has been
set for the portion of insulated trocar base inside and outside the
tissue domain. The electrode potential Ve and trocar tip potential Vtt
have been set to the constant source potential V ( = P *Zref ). Here,
P refers to the supplied power and Zref is the reference impedance
prole of RFA device. The effect of external radiation on the current
within the electrode is negligible at 500 kHz radio frequency.
Therefore, it has been assumed that the current within the elec-
trode is fully balanced (Ekstrand et al., 2005). The outer surface of
breast and trocar base (outside the tissue domain) have been as-
sumed to be subjected to natural convection, i.e., qconv
(5)
hconv(TsT1), where hconv and T1 are considered as
10 W m  2 K  1 and 22.4 C, respectively. The initial temperature of
breast tissue has been considered to be same as that of the core
body temperature, i.e., Tc (37 C), and the initial temperature of
the electrode, trocar tip as well as trocar base has been set to 25 C.
At present, different modes of controlling unit (source) are used
to control the power delivered to the tissues, such as power-con-
trolled mode, temperature-controlled mode and impedance-con-
trolled mode. The present study modeled the case of manual
temperature control mode instead of programmed adjusted tem-
perature control mode. In the manual control mode, the usual
practice is to apply a constant voltage in the active electrode/trocar
tip by maintaining the tip sensor temperature below the preset
target temperature (Tungjitkusolmun et al., 2002). In this study,
the preset target temperature has been set to 70 C, and the
(6) ror method. For numerical simulations, the specic heat and
density of blood have been considered to be 4200 J kg  1 K  1 and
1060 kg m  3, respectively.
2.4. Numerical procedure
The spatial and temporal temperature distributions during RFA
of cancerous breast tumor have been obtained by solving the
 S. Singh et al. / Journal of Thermal Biology 58 (2016) 8090 85

coupled electric eld distribution equation (Eq. (3)), energy 80

equation for perfused tissue domain (Eq. (1)) and rate equation for Measured (Sethuraman et al., 2014)

Rise in temperature, T ( C)
tissue damage (Eq. (6)) using a commercially available nite ele-

0
ment software, COMSOL Multiphysics 5.0. The present work used Present FEM model
the version 5.0 of the coupled solver that was developed by
COMSOL AB, Stockholm, Sweden.
60
The meshing of physical domain has been done using ne tet-
rahedral mesh elements, i.e., 4,31,486 tetrahedral elements. The
maximum element sizes of muscle, gland, fat and the tumor have
been considered to be 0.3 cm, 0.3 cm, 0.3 cm, 0.05 cm, respectively. 40
Similarly, the maximum element sizes of electrode, trocar tip and
trocar base have been considered to be 0.05 cm, 0.055 cm and Electrode Tip location:
0.1 cm, respectively. The above mesh element sizes have been de- 3 cm from top surface of bovine liver
termined after several runs with varying number of mesh elements 20
and identical properties to attain a grid independent solution. The P=6W
relative tolerance for the electric eld interface and heat transfer tf = 15 min
interface has been set to 0.0001. The numerical convergence has
been found to be taking place below the pre-specied relative tol- 0
erance of the solvers. A direct linear solver, MUMPS (MUltifrontal 0 250 500 750 1000
Massively Parallel sparse direct Solver) with default preordering Time (s)
algorithm has been used for solving the electric eld. Whereas,
Fig. 2. Comparison between the measured electrode tip temperature values (Se-
iterative conjugate gradient method has been used with geometric
thuraman et al., 2014) and predicted temperature values (from the present FEM
multi grid pre-smoothers for solving the temperature eld. The model) at the electrode tip location 3 cm from the top surface of bovine liver tissue
Backward Difference Formula (BDF) has been used for time during RFA of bovine liver at 6 W constant power input (P) and exposure time (tf) of
marching, and care has been taken to store the data at each step by 15 min.
assigning strict steps for the solvers. The Newton-Raphson algo-
rithm has been used for iteration in both direct and iterative solvers. Table 2
Electrical and thermal properties of bovine liver tissue and electrode (Haemmerich
et al., 2003; Sethuraman et al., 2014).

3. Results and discussion Tissue/electrode Electrical Thermal properties

properties
3.1. Model validation Properties s (S m  1) c (J kg  1 K  1) k (W m  1 K  1) (kg m  3)

Bovine liver 0.18a 3600b 0.65a 1060b

In order to verify the model delity, the elevation of electrode tip
temperature above the measured temperature at ambient room
condition (20 C) during the ex vivo study of RFA on bovine liver
tissue (Sethuraman et al., 2014), has been compared with the pre-
dictions made by the present FEM model (considering similar
geometry, electrode conguration and boundary conditions), as can
be seen from Fig. 2. For validation, the electrical and the thermal
properties of bovine liver tissue and electrode have been obtained
from Haemmerich et al. (2003), Sethuraman et al. (2014) (refer
Table 2). The blood perfusion and the metabolic heat generation
effects have been neglected for ex vivo study (Sethuraman et al.,
2014). The electrode has been modeled within the liver up to an
insertion depth of 3 cm from the top surface of liver model. The
bovine liver model has been heated using a constant power P of
6 W for 15 min similar to the operational details opted by Se-
thuraman et al. (2014). Here, the applied constant input electric
potential V of the electrode was  17.30 V (i. e. , V = P *Zref ) cor-
responding to the reference impedance Zref prole around 50 .
The prediction of rise in temperature using the present model
(Fig. 2) has been found to be in satisfactory agreement with the
measured value presented in Sethuraman et al., (2014). Henceforth,
the thermal analysis of RFA of breast tumor has been carried out
using the same model, except few input congurations for tumor
ablation have been modied, since the tissue properties of human
breast tumor differs from that of the bovine tissue.
3.2. Estimation of optimal regulating parameters for RFA of breast
tumor
It is well-known from the earlier study (Ekstrand et al., 2005)
that, the preferential heating of tumor volume depends on mul-
tiple factors, viz., tissue properties, tumor shape, applied voltage,
Active probe tip 108 840 18 6450
(NiTi)
Probe shaft 10  5 1045 0.026 70
(Polyurethane)b
a
Sethuraman et al. (2014).
b
Haemmerich et al. (2003).
heating time, electrode placement within the tumor, etc. Out of all
these factors, exposure time and applied voltage play a critical
role, which needed to be optimized for different stages of breast
cancers, in ensuring risk free clinical application of RFA. The
thermal efcacy of tumor damage has been assessed based on the
thermal damage eld ( Z1.0) and the required exposure time (t)
instead of isothermal eld. Since, isothermal elds overestimate
the amount of tissue injury and produces under heating of tumor
if heating is stopped earlier. To evaluate the optimal voltage and
desired time for thermal damage ( Z1), detail sensitivity study
has been performed by varying the applied voltages for different
tumor diameters, as can be seen from Fig. 3(a) and (b). The study
considered the manual adjustment of voltages by trial-and-error
method, and by keeping the tip temperature below the target
temperature (70 C). The applied voltages and exposure time for
various stages of tumor have been found to be different, since the
thermal capacity (cV) varies with the tumor volume (V), as can be
seen from Fig. 3(a) and (b). Hence, for a given tumor volume a
trade-off has been attained between exposure time and voltage. In
this work, the optimal regulating parameters, viz., voltage and
exposure time (Fig. 3(c)) have been determined based on various
interdependent factors such as (a) maintaining mild temperature
range for tumor ablation, (b) maintaining the tip temperature
below the target temperature, and (c) maintaining the exposure
86 S. Singh et al. / Journal of Thermal Biology 58 (2016) 8090

Fig. 3. Variation of (a) time to reach the threshold temperature (t45 C) and (b) necrosis time () for different tumor diameters (D) with applied voltage (V), and (c) the
temperature front (at T 45 C) and thermal damage front ( 1.0) of different tumor sizes for the associated optimized regulating parameters.

time reasonably longer (preferably in minutes) to attain a better
control during the clinical procedure. Fig. 3(a) and (b) illustrate the
variation of time required to reach the threshold temperature
(t45 C) and necrosis time (), respectively, with applied voltage for
different stages of breast tumor. The necrosis time (
( tf-t45 C)4 t45 C). Where, tf is the total time.
The associated threshold temperature front and damage front
due to optimal regulating parameters, viz., voltage and exposure
time, for different tumor development stages have been shown in
Fig. 3(c). It is worth-noting from Fig. 3(c) that, the entire tumor
reaches the threshold temperature quite earlier (i.e., t450C) than the
necrosis time (). Thus, there is a possibility of under heating if the
temperature is monitored during the thermal ablation. In the
present analysis, the heating of tumor volume has been dis-
continued as soon as the entire tumor volume reaches the desired
damage value (Z1), i.e. equivalent to the time (tf).
The safe/required damage, while minimizing excess heating
and attaining a good control of RFA in clinical practices, can be
achieved by employing mild voltage and by increasing the ex-
posure time reasonably. The considered optimal exposure time (tf)
and voltage (V) for different tumor volumes are less than 10 min
(i.e., necrosis time, o4 min) and not more than 10 V, respec-
tively, as can be seen from Fig. 3(c). These values are chosen in
such a manner that, the clinical practitioners have reasonable time
to damage the tumor, and stop the electric supply. It is worth
noting that, the optimal values of regulating parameters may differ
depending on the type of tissue and electrode.
3.3. Assessment of sensitivity of clinically relevant monitoring
parameter
Fig. 4(a) represents various temperature measurement loca-
tions during RFA, viz., electrode tip, P1 and P4 (tumor-gland
junctions), P2 and P5 (0.5 mm away from the tumor) and P3 and P6
(1.0 mm away from the tumor). For all the tumor diameters, rise in
temperature above the core body temperature ( 37 C) is max-
imum at the tumor-electrode tip junction, as can be seen Fig. 4(b).
For illustration purpose only two stages, viz., T1a (0.5 cm) and T1b
(1.0 cm) are presented in Fig. 4(b). The electrode tip position P4 has
been adjusted for different tumor diameters by trial-and-error
method to ensure damage front spread beyond P4 is minimum. In
case of all tumor sizes (Fig. 4(b)), tumor-gland junctions (P1 and
P4) are above the threshold temperature, and are exposed for time
(tf) to attain the desired thermal damage ( Z1).
During clinical RFA procedure (in case of temperature con-
trolled system), voltage is adjusted either manually or using au-
tomated program by monitoring the in-vivo electrode tip tem-
perature with a temperature sensor placed at the electrode tip
casing (Hayashi et al., 2003). It is worth-noting that, RF electrode is
not directly heated by the RF current due to the low intrinsic
impedance of material, instead it is heated by the surrounding
tissue. Thus, the measurable in-vivo electrode tip temperature
 S. Singh et al. / Journal of Thermal Biology 58 (2016) 8090 87

Fig. 4. (a) Schematic diagram representing different temperature measurement locations during RFA and (b) variation of rise in temperature proles above the core
temperature ( 37 C) for two tumor stages at different locations (viz., electrode tip, tumor-glandular junction and healthy glandular tissue) with exposure time (the
optimized regulating parameter are same as in Fig. 3(c)).

could be the useful data to know the rise in peri-electrode tumor
temperature. In view of this, the present work relates the electrode
tip temperature with that of the induced damage due to rise in
entire tumor temperature (Fig. 4(b)). Further, the present study
proposes that, the pre-clinical modeling would be useful to
(a) quantify the accurate tumor damage/exposure time to mini-
mize the damage of peripheral healthy tissues and (b) facilitate the
estimation of the associated temperature rise of the entire tumor
based on the tip temperature. The knowledge of ablation/exposure
time as well as the temperature difference between the electrode
tip and the peripheral tumor-healthy junctions from the pre-
clinical modeling would be resourceful to achieve a better in-vivo
control of clinical RFA (simply by monitoring the electrode tip
temperature). It can be clearly seen from Fig. 4(b) that, the asso-
ciated exposure time (tf) as well as the temperature difference
between the tip and the peripheral junction of tumor can certainly
estimate the extent of damage (based on the tip temperature at
the end of necrosis time, ). The tip temperature, associated

Voltage = 8 V T (0C) Voltage = 10 V

T (0C) t = 551 s
57.46 t = 385 s 69.69
-5 -4 -3 -2 -1 0 1 2 3 4 5 73.1 -6 -4 -2 0 2 4 6 5684.4
1.0 1.0
56 41.0
T = 45 0 C = 0.7 38.0 0.9 T = 45 0 C = 0.7 0.9
54 65 40.0
37.0 0.8 39.0 0.8
52 36.0 > 1.0 38.0
0.7 60 0.7
50 35.0 37.0
T = 48 0 C 0.6 36.0 0.6
48 34.0 55 T = 50 0 C
> 1.0 35.0
0.5 34.0 0.5
46 33.0
0.4 50 33.0 0.4
44 T = 50 0 C 32.0 32.0
31.0 0.3 31.0 0.3
42 45
30.0 0.2 30.0 0.2
40 T = 55 0 C 29.0
29.0 0.1 T = 60 0 C
38 40 28.0 0.1
0 27.0
37 0
37
Tumor (D = 0.5 cm Tumor (D = 1 cm
(a) (b)
Fig. 5. Thermal and damage eld distributions/maps of two tumor sizes during RFA due to optimal regulating parameters shown in Fig. 3(c); (a) D 0.5 cm, and (b)
D 1.0 cm.
88 S. Singh et al. / Journal of Thermal Biology 58 (2016) 8090
temperature difference and the exposure time at the end of tumor
necrosis for two different breast cancer stages, viz., T1a (0.5 cm)
and T1b (1.0 cm) have been shown in Fig. 4(b), which will provide
a priori information about the controlling constraints (i.e., the
applied voltage, target electrode tip temperature and the total
exposure time) for clinical application of RFA.
The variation in thermal and damage eld distributions/maps
of two different tumor development stages (i.e., T1a and T1b) at
the end of exposure time have been presented in Fig. 5. It is evi-
dent from Fig. 5 that, the isothermal contours (on the left hand
side) overestimate the tumor damage front from that of the actual
damage front (on the right hand side). Although, the onset of tu-
mor damage starts at 45 C but the desired damage of the entire
tumor, i.e., Z1, takes place at a relatively higher temperature
(T 445 C).
3.4. Effect of induced thermal damage on healthy cell volume, tumor
cell volume and blood perfusion
The thermal damage of tumor and healthy tissue volumes have
been calculated by subtracting (a) the portion of the trocar volume
from the original volume of tumor before the trocar is inserted,
and (b) the entire portion of trocar volume and tumor volume
from the volume of damage front, respectively, as shown in Fig. 6
(a). Fig. 6(b) depicts the variation of damaged volume for both
healthy and tumor tissues with the increase in tumor diameter for
the chosen optimal regulating parameters (Fig. 3(c)) during RFA of
breast tumor. For relatively smaller tumor diameters (i.e., upto
5 mm or so), the damaged volume of healthy tissue is more
compared to the tumor tissue. Whereas, opposite trend has been
noticed for the tumor having diameter greater than 7 mm. This
situation arises because trocar geometry has been kept unchanged
for all tumor diameters, though trocar tip from tumor bottom
surface has been adjusted. If the diameter of electrode is com-
parable with the tumor diameter (in the lateral direction), then the
500
Damaged volume (mm )
3
400
300
200
100
0
1 mm 3 mm 5 mm 7 mm 10 mm
Fig. 6. (a) Schematic diagram of damage front ( 1) representing the calculation of damaged tumor and healthy volumes due to RFA, and (b) variation of damaged healthy
and tumor volumes during RFA of different tumor sizes.
heat deposited inside the tumor conducts further towards the
healthy breast tissues. On the other hand, if the tumor diameter (in
the longitudinal direction) is less than the length of electrode then,
a greater portion of the healthy tissue is exposed to the outer
surface of the electrode. It is evident from Fig. 6(b) that, extent of
damage to the healthy volume during RFA of small/large volume
tumor is relatively small in magnitude, i.e., o0.22 cm3, if the
electric supply is discontinued after reaching the exposure time.
However, in practice a signicant amount of surrounding healthy
tissue is subjected to post-treatment damage due to thermal dif-
fusion, and is considered to be desirable in clinical practice to limit
the migrating cancer cells within the healthy tissues.
Fig. 7 illustrates the cell survival fraction and cell damage
fraction due to the rise in tissue temperature and induced thermal
damage during RFA of breast tumor. In Fig. 7, left and right por-
tions of the electrode tip represent the concentration of cell sur-
vival fraction and cell damage fraction, respectively. It is apparent
from Fig. 7 that, the cell damage fraction increases with the ex-
posure time. Whereas, the cell survival fraction decreases with the
increase in exposure time. It can be concluded from the above
results that, the thermal degradation of cell is a gradual process
and complete death of cell takes place at a relatively large value of
induced damage. In practice, the desired damage of tumor cells
can be attained at 1.0, i.e., the threshold cell damage fraction
to be CD/C0 Z63.20% and cell survival fraction to be CUD/C0
r36.80%. The cell survival fraction greater than the threshold
value, and the cell damage fraction less than the threshold value
may cause reappearance/possible healing of tumor cells after the
treatment. It can be further noted that, the tumor cells near the
electrode tip suffers maximum damage fraction (i.e., CD/C0 490%)
and least cell survival fraction (i.e., CUD/C0 o10%).
3.5. Outline of pre-clinical modeling strategy
The pre-clinical computational modeling could play a crucial
(a)
Healthy tissue
Tumor tissue
Tumor diameter, D
(b)
 S. Singh et al. / Journal of Thermal Biology 58 (2016) 8090 89

Fig. 7. Cell survival fraction (CUD/C0) and cell damage fraction (CD/C0) maps/distributions of two different tumor development stages (D 0.5 cm and 1.0 cm) during RFA due
to optimal regulating parameters.

role to understand the complicacies/risks of RFA before its clinical 4. Conclusion

implementation. The gained knowledge could be useful for the
development of strategies or shall provide a route to minimize the The present article demonstrates the role of predictive pre-
complications. The mathematical models may serve as a predictive clinical computational modeling strategies towards an effective
tool and notify the clinical practitioners about the possible out- and risk free thermal ablation of breast tumor using RF technique.
In other words, the article laid down certain clinically relevant
come. Therefore, based on the results discussed in the previous
outlines (a) to minimize the extent of thermal damage to the
sections, the present study outlined a few essential strategies;
healthy volume by optimizing the regulating parameters, (b) to
avoid healing of tumor by ensuring sufcient thermal damage to
(1) Initially, the induced thermal damage to healthy tissue needs
the tumor volume and (c) to enable better control of in-vivo
to be minimized for different development stages of tumor by
thermal damages by monitoring the measurable clinical para-
establishing an optimal input conguration (viz., voltage, ex-
meters. The study adopted systematic modeling steps to predict
posure time, and electrode-tip temperature) of the RFA tech-
the induced damages to the healthy cells during RF ablation of
nique by performing a parametric study.
different development stages of breast tumor. Based on the com-
(2) After establishing the optimal input conguration, the relation
puter simulations, it has been veried that the thermal damage
(i.e., the difference in temperatures at the end of necrosis
eld provides a better estimation of tissue-injury compared to the
time) between the essential monitoring parameters (elec-
isothermal eld. The optimization of regulating parameters reveals
trode-tip temperature) and tumor-healthy junction needs to
that, the undesirable thermal injury can be minimized by achiev-
be extracted from the numerical results to enable a better in-
ing a good control during clinical RFA by employing mild voltages
vivo control of clinical RFA.
and reasonably longer exposure time. The study further proposes
(3) The electrode position within the tumor for all development
that the measurable quantities, viz., electrode tip temperature and
stages needs to be adjusted and electrode geometry needs to
exposure time shall act as potential monitoring parameters in
be congured reasonably to impose sufcient damage to the
clinical practices. Thus, a priori knowledge of change in tempera-
tumor volume sparing healthy tissue. If the electrode position
ture between the electrode tip and peripheral tumor-healthy
is adjusted, earlier steps {steps (a) and (b)} shall be repeated.
junctions as well as the desired exposure time at the end of tumor
(4) Finally, the possibility of reappearance/healing of damaged
necrosis from computational study shall allow better control of in-
tumor cells and healthy cells after the completion of RFA
vivo damages. It has been evaluated that, the thermal degradation
needs to be conrmed numerically by quantifying the cell
of cell is a gradual process and therefore it is possible that tumor
survival and cell damage fractions.
may heal after the treatment due to insufcient damage of tumor
volume. The insufcient damage of tumor volume has been
All the above mentioned essential strategies have been devel-
quantied based on the cell survival and cell damage fractions.
oped using a viable computational model. The obtained results/
input data after following the essential modeling strategies, if used
to assist clinical application of RFA would lead to maximum da- Acknowledgements
mage to the tumor tissue and minimum damage to the healthy
tissue. Authors would like to acknowledge Science and Engineering
90 S. Singh et al. / Journal of Thermal Biology 58 (2016) 8090
Research Board, Department of Science and Technology, Govern-
ment of India, for providing the Grant (Grant number SB/FTP/ETA-
0135/2013) to pursue the present research work. Authors would
also like to acknowledge Indian Institute of Technology Ropar for
providing essential infrastructure to carry out the present
research.
References
Abraham, J.P., Sparrow, E.M., 2007. A thermal-ablation bioheat model including
liquid-to-vapor phase change, pressure-and necrosis-dependent perfusion, and
moisture-dependent properties. Int. J. Heat Mass Transf. 50, 25372544.
Andreuccetti, D., Fossi, R., Petrucci, C., 1997. An Internet resource for the calculation
of the dielectric properties of body tissues in the frequency range 10 Hz to
100 GHz, [Online]. Available: http://niremf.ifac.cnr.it/tissprop/, IFAC-CNR,
Florence (Italy), 1997. Based on data published by C. gabriel et al. in 1996.
(accessed 06.12.14).
Bezerra, L.A., Oliveira, M.M., Rolim, T.L., Conci, A., Santos, F.G.S., Lyra, P.R.M., Lima, R.
C.F., 2013. Estimation of breast tumor thermal properties using infrared images.
Signal Process. 93, 28512863.
Bhowmik, A., Repaka, R., Mishra, S.C., Mitra, K., 2015a. Thermal assessment of ab-
lation limit of subsurface tumor during focused ultrasound and laser heating. J.
Therm. Sci. Eng. Appl. 8 (1), 112. http://dx.doi.org/10.1115/1.4030731.
Bhowmik, A., Repaka, R., Mishra, S.C., 2015b. Thermal analysis of the increasing
subcutaneous fat thickness within the human skina numerical study. Numer.
Heat Transf. A Appl. 67 (3), 313329.
Bhowmik, A., Repaka, R., Mishra, S.C., 2014. Thermographic evaluation of early
melanoma within the vascularized skin using combined non-Newtonian blood
ow and bioheat models. Comput. Biol. Med. 53, 206219.
Bhowmik, A., Singh, R., Repaka, R., Mishra, S.C., 2013. Conventional and newly
developed bioheat transport models in vascularized tissues: a review. J. Therm.
Biol. 38, 107125.
Burak, W.E., Agnese, D.M., Povoski, S.P., 2003. Radiofrequency ablation of invasive
breast carcinoma followed by delayed surgical excision. Cancer 98, 13691376.
Chang, I.A., 2003. Finite element analysis of hepatic radiofrequency ablation probes
using temperature-dependent electrical conductivity. Biomed. Eng. Online 2,
12.
Das, K., Mishra, S.C., 2015. Simultaneous estimation of size, radial and angular lo-
cations of a malignant tumor in a 3-D Human breast-a numerical study. J.
Therm. Biol. 52, 147156.
Deng, Z.S., Liu, J., 2004. Numerical simulation of 3-D freezing and heating problems
for combined cryosurgery and hyperthermia therapy. Numer. Heat Transf. A
Appl. 46, 587611.
DeSantis, C., Siegel, R., Bandi, P., Jemal, A., 2011. Breast cancer statistics 2011. CA
Cancer J. Clin. 61, 408418.
Ekstrand, V., Wiksell, H., Schultz, L., Sandstedt, B., Rotstein, S., Eriksson, A., 2005.
Inuence of electrical and thermal properties on RF ablation of breast cancer: is
the tumour preferentially heated? Biomed. Eng. Online 4, 41.
e-funda, 2014. Material data base e-funda: Properties of steel, [Online]. Available:
http://www.efunda.com/materials/alloys/alloy_home/steels_properties.cfm,
(accessed 06.12.14).
Fornage, B.D., Sneige, N., Ross, M.I., Mirza, A.N., Kuerer, H.M., Edeiken, B.S., Ames, F.
C., Newman, L.A., Babiera, G.V., Singletary, S.E., 2004. Small (r 2-cm) breast
cancer treated with US-guided radiofrequency ablation: feasibility study.
Radiology 231, 215224.
Frangioni, J.V., 2008. New technologies for human cancer imaging. J. Clin. Oncol. 26,
40124021.
Greene, F.L., Compton, C.C., Fritz, A.G., Shah, J.P., Winchester, D.P., 2006. Breast, AJCC
Cancer Staging Atlas, Part VI. Springer, New York, pp. 217234.
Grotenhuis, B.A., Vrijland, W.W., Klem, T.M.A.L., 2013. Radiofrequency ablation for
early-stage breast cancer: Treatment outcomes and practical considerations.
Eur. J. Surg. Oncol. 39, 13171324.
Haemmerich, D., Chachati, L., Wright, A.S., Mahvi, D.M., Lee Jr, F.T., Webster, J.G.,
2003. Hepatic radiofrequency ablation with internally cooled probes: effect of
coolant temperature on lesion size. IEEE Trans. Biomed. Eng. 50, 493500.
Hayashi, A.H., Silver, S.F., van der Westhuizen, N.G., Donald, J.C., Parker, C., Fraser, S.,
Ross, A.C., Olivotto, I.A., 2003. Treatment of invasive breast carcinoma with
ultrasound-guided radiofrequency ablation. Am. J. Surg. 185, 429435.
Henriques, F.C., 1947. Studies of thermal injury V: the predictability and sig-
nicance of thermally induced rate processes leading to irreversible epidermal
injury. Arch. Pathol. 43, 489502.
Jamil, M., Ng, E.Y.K., 2013a. Evaluation of meshless radial basis collocation method
(RBCM) for heterogeneous conduction and simulation of temperature inside
the biological tissues. Int. J. Therm. Sci. 68, 4252.
Jamil, M., Ng, E.Y.K., 2013a. Ranking of parameters in bioheat transfer using Taguchi
analysis. Int. J. Therm. Sci. 63, 1521.
Jamil, M., Ng, E.Y.K., 2013a. Statistical modeling of electrode based thermal therapy
with Taguchi based multiple regression. Int. J. Therm. Sci. 71, 283291.
Jamil, M., Ng, E.Y.K., 2013b. To optimize the efcacy of bioheat transfer in capacitive
hyperthermia: A physical perspective. J. Therm. Biol. 38 (5), 272279.
Jamil, M., Ng, E.Y.K., 2015. Quantication of the effect of electrical and thermal
parameters on radiofrequency ablation for concentric tumour model of differ-
ent sizes. J. Therm. Biol. 51, 2332.
Jeffrey, S.S., Birdwell, R.L., Ikeda, D.M., Daniel, B.L., Nowels, K.W., Dirbas, F.M.,
Griffey, S.M., 1999. Radiofrequency ablation of breast cancer: rst report of an
emerging technology. Arch. Surg. 134, 10641068.
Jia, W.C., Choi, B., Franco, W., Lot, J., Majaron, B., Aguilar, G., Nelson, J.S., 2007.
Treatment of cutaneous, numerical and animal studies. Lasers Surg. Med. 39,
494503.
Jordan, A., Scholz, R., Maier-Hauff, K., Johannsen, M., Wust, P., Nadobny, J., Schirra,
H., Schmidt, H., Deger, S., Loening, S., Lanksch, W., Felix, R., 2001. Presentation
of a new magnetic eld therapy system for the treatment of human solid tu-
mors with magnetic uid hyperthermia. J. Magn. Magn. Mater. 225, 118126.
Jossinet, J., 1996. Variability of impedivity in normal and pathological breast tissue.
Med. Biol. Eng. Comput. 34, 346350.
Megan, J., Raje, S., Kim, K., Mitra, K., Guo, Z., 2008. Bio-heat transfer analysis during
short pulse laser irradiation of tissues. Int. J. Heat Mass Transf. 51, 55115521.
Mirza, A.N., Fornage, B.C., Sneige, N., 2001. Radiofrequency ablation of solid tumors.
Cancer J. 7, 95102.
Mitra, S., Balaji, C., 2010. A neural network based estimation of tumour parameters
from a breast thermogram. Int. J. Heat Mass Transf. 53 (21), 47144727.
Morady, F., Harvey, M., Kalbeisch, S.J., El-Atassi, R., Calkins, H., Langberg, J.J., 1993.
Radiofrequency catheter ablation of ventricular tachycardia in patients with
coronary artery disease. Circulation 87, 363372.
Ng, E.Y.K., Jamil, M., 2014. Parametric sensitivity analysis of radiofrequency ablation
with efcient experimental design. Int. J. Therm. Sci. 80, 4147.
Ng, E.Y.K., Sudharsan, N.M., 2001. An improved three-dimensional direct numerical
modelling and thermal analysis of a female breast with tumour. Proc. Inst.
Mech. Eng. H 215, 2537.
Paul, A., Narasimhan, A., Kahlen, F.J., Das, S.K., 2014. Temperature evolution in tis-
sues embedded with large blood vessels during photo-thermal heating. J.
Therm. Biol. 41, 7787.
Pennes, H.H., 1998. Analysis of tissue and arterial blood temperatures in the resting
human forearm. J. Appl. Physiol. 85, 534.
Quaranta, V., Manenti, G., Bolacchi, F., Cossu, E., Pistolese, C.A., Buonomo, O.C.,
Carotenuto, L., Piconi, C., Simonetti, G., 2007. FEM analysis of RF breast ablation:
multiprobe versus cool-tip electrode. Anticancer Res. 27, 775784.
Sethuraman, S., Anand, A., Li, J., 2014. Integrated ultrasound thermometry and
multiphysics modeling for liver RF ablation monitoring: Ex vivo studies. In:
Proceedings of the IEEE Ultrasonics Symposium, Chicago, USA. pp. 16501653.
Singh, S., Bhowmik, A., Repaka, R., 2015. Radiofrequency ablation of malignant
breast tumor: A numerical study. In: Proceedings of the 23rd National Heat and
Mass Transfer Conference and 1st International ISHMT-ASTFE Heat and Mass
Transfer Conference, Thiruvananthapuram, India, 20 December 2015.
Teon, 2014. Teon (Polytetrauoroethylene), [Online]. Available: http://www.phy.
davidson.edu/fachome/dmb/PY430/Friction/teon.html, (accessed 0.12.14).
Tungjitkusolmun, S., Staelin, S.T., Haemmerich, D., Tsai, J.Z., Cao, H., Webster, J.G.,
Lee Jr, F.T., Mahvi, D.M., Vorperian, V.R., 2002. Three-dimensional nite-ele-
ment analyses for radio-frequency hepatic tumor ablation. IEEE Trans. Biomed.
Eng. 49, 39.
Vaupel, P., 1998. Tumor blood ow. In: Molls, M., Vaupel, P. (Eds.), Blood Perfusion
and Microenvironment of Human Tumors: Implications for Clinical Radio-
oncology. Springer-Verlag, Berlin, pp. 4146.
Vallez, L.J., Plourde, B.D., Abraham, J.P., 2016. A new computational thermal model
of the whole human body: Applications to patient warming blankets. Numer.
Heat Transf. A Appl., 115. http://dx.doi.org/10.1080/10407782.2015.1080573.
Wu, T., Li, P., Shao, Q., Hong, J., Yang, L., Wu, S., 2015. A simulation-experiment
method to characterize the heat transfer in ex-vivo porcine hepatic tissue with
a realistic microwave ablation system. Numer. Heat Transf. A Appl. 64,
729743.