A MINIMALLY INVASIVE MICROWAVE HYPERTHERMIC

APPLICATOR WITH AN INTEGRATED TEMPERATURE

SENSOR

Guido Biffi Gentili and Mariano Linari

Department of Electronic and Telecommunications, University of Florence, Via S. Marta 3, 50139 Firenze, Italy

Keywords: Interstitial microwave applicator, endocavitary, hyperthermia, temperature sensor.

Abstract: In the field of microwave hyperthermia and thermo-ablation, the use of minimally invasive applicators is

recognized as a very promising means for the treatment of small, early stage, cancer lesions because a very

thin applicator can be easily introduced inside the body and precisely directed towards a deep seated tumour

using the most advanced 3D imaging techniques and surgical stereo-navigation. Minimally invasive

applicators have been successfully employed for the treatment of bladder carcinoma and brain tumours but

the accurate temperature monitoring of the heated tissue volume still remains an open problem. In this

paper we propose a new minimally invasive applicator, integrating a low-cost metallic wired temperature

sensor. The miniaturised endocavitary applicator consists of a asymmetric isolated dipole operating at 2.45

GHz. The very slim shape of the applicator allows to easily insert it into the lesion through a soft plastic

tube (catheter) while a temperature sensor, properly embedded in the applicator body, measures the tissue

temperature at the interface with the catheter surface. An electromagnetic analysis based on the Finite

Integration Technique (FIT) and experimental verifications over a tissue sample proved that a coaxial

choke, enclosing the temperature sensor wires, allows localize the heating pattern in a restrict volume while

drastically reducing measuring artefacts due to the perturbing effects induced by the probe leads.

1 INTRODUCTION

Microwave endocavitary and interstitial

hyperthermia has been widely investigated in the

past decades as localised thermal therapy for cancer

treatment. Many thin applicators have been

developed and used in therapeutic applications with

valuable results. Most of them are essentially

constituted of an insulated monopole, dipole or helix

feed through a thin coaxial cables. However several

technical solutions (Turner, 1986; Tumeh and

Iskander, 1989; Camart et al., 1996; Lin and Wang,

1987; Cerri et al.; 1993; Saito et al., 2000) have

been proposed in order to localize the heating in a

restrict area of tissue around the tumour and to avoid

accidental and unwanted hot spots in the healthy

tissue.

The effectiveness of a microwave thermal

therapy depends not only on the radiative properties

of the applicator used but also on the reliable and

accurate temperature control during the therapeutic

treatment. Impedance tomography, microwave

radiometry, magnetic resonance imaging (MRI) and

also methods based on ultrasounds are very

attractive non-invasive temperature-monitoring

techniques. Despite these promising prospects, up

till now a very accurate measure of depth tumours

temperature can be obtained only by invasive

techniques, because only thermocouples, thermistors

or optical fibres sensors seem to give the required

measuring accuracy, spatial resolution and real-time

response. Temperature monitoring by invasive

sensors is common practice in the hyperthermic

treatments. Usually, very thin probes are separately

inserted in the tissue near the antenna and moved

ahead and back in order to better estimate the

effective heating pattern. In that case only optical

fibre sensors could be employed because metallic

wired probes located near the antenna strongly

interact with the radiating element, producing

uncontrollable electromagnetic fields distortion and

false temperature readings.

When the tumour is at its early stage a single

applicator can be employed to heat the small volume

of the lesion by inserting it as sketched in Figure 1.

113

Bifﬁ Gentili G. and Linari M. (2008).

A MINIMALLY INVASIVE MICROWAVE HYPERTHERMIC APPLICATOR WITH AN INTEGRATED TEMPERATURE SENSOR.

In Proceedings of the First International Conference on Biomedical Electronics and Devices, pages 113-118

DOI: 10.5220/0001052801130118

 SciTePress

In this case it is advantageous to integrate the

temperature sensor directly in the applicator body in

order to avoid additional traumas in the patient and

to simplify the hyperthermic treatments itself.

Figure 1: Single applicator heating a small tumour.

When lesions are more extended in volume an array

of applicators could be required to uniformly heat

the tumor, as sketched in Figure 2.

Figure 2: Array of applicators heating a medium-sized

tumour.

Also in this case applicators with integrated

temperature sensors can be usefully employed

because the temperature distribution inside the

tumor volume can be numerically estimated starting

from the point measurements taken in

correspondence of each radiating element, by

combining tomography algorithms and bio-heat

equation.

From the mechanical and structural point of view

it is not difficult to integrate a very thin fiber-optic

probe inside the core of the applicator without

altering its original diameter. It is worth to note,

however, that fiber-optic sensors are very expensive

and delicate devices and thus not properly suited for

the production of a rugged mono-use applicators, as

required for the routined clinical practice.

Another sensor that is practically unaffected by

the strong EM fields existing near the antenna is the

Bowman thermistors (Bowman, 1976) because its

high resistance leads can not carry significant RF

currents. Unfortunately also this sensor type is

delicate and expensive to be fabricated.

Ordinary low-cost and sturdy thermometric

sensors, as thermocouples and thermistors, use high

conductance metallic leads for the connection to the

measuring unit. Unfortunately, when they are placed

too close to the applicator body, metallic wires cause

unwonted electromagnetic (EM) interferences. In

these cases dangerous and uncontrolled hot spots in

the tissue can occur, as well erroneous (or very

noisy) temperature readings.

This work suggests a novel technique for

integrating a low-cost wired temperature sensor

inside the body of a minimally invasive Microwave

Hyperthermic Applicator (MHA) without perturbing

the radiated fields.

2 METHODS

2.1 Applicator Design

The proposed interstitial/endocavitary MHA,

depicted in Figure 3, essentially consists of a coaxial

asymmetric dipole type antenna, radiating in the

biological tissue through an insulating tube

(catheter).

Figure 3: Microwave Hyperthermic Applicator (MHA)

(length: 65 mm; thickness: 2 mm) working at 2.45 CHz.

In order to avoid unwanted heating of the healthy

tissue, a coaxial balun has been introduced to block

the back currents flowing on the surface of the

coaxial feeding line (Longo et al., 2003). The

diameter of the radiating upper arm of the dipole has

been properly increased to improve the matching of

BIODEVICES 2008 - International Conference on Biomedical Electronics and Devices

114

the MHA input impedance to the tissue (Jones et al.,

1988; Biffi Gentili et al., 1995). Thank to its very

thin shape (65 mm in length and 2 mm in thickness)

the coaxial applicator can be easily introduced inside

a small catheter (3-4 mm in diameter) and

subsequently inserted into the lesion. The position of

the applicator can be fixed in loco using an anchor

balloon.

Starting from this basic applicator configuration,

a wired temperature probe (thermocouple or

thermistor) has been successively introduced into the

catheter inside the balun as depicted in Figure 4.

Figure 4: MHA with a temperature probe embedded inside

the insulator of the coaxial choke.

The temperature sensitive tip of the probe is

embedded in the catheter body in order to measure

the temperature of the tissue at the interface with the

catheter itself, where higher temperatures are

expected. Miniature (SMT) chip inductors can be

also inserted close to the sensing tip of the probe to

isolate it from the radiating conductors of the coaxial

applicator.

2.2 Numerical Analysis

Electromagnetic full-wave analysis has been

employed to investigate the Specific Absorption

Rate (SAR) distribution produced by the applicator

and the perturbations due to the closeness of the

metallic wires of the temperature probe. EM fields

have been calculated with a time domain Finite

Integration Technique (FIT) (CST Studio Suite,

2006) in a three-dimensional spatial domain

constituted by a muscular tissue volume including

the MHA model.

Perfect Matched Layers (PML) boundaries condition

(Berenger, P. 1994) have been used in order to limit the

computational domain to 140 × 60 × 60 mm

volume.

However the thin profile of the applicator and the high

permittivity of the human muscle tissue (ε

= 52 − j13 @

2.45 GHz) required a very little mesh size (< λ/50 at 3

GHz) which made heavy the computation load. EM

simulations, performed in the 2 to 3 GHz frequency range,

run for 2 hour with an Intel Pentium III @1GHz with 1.5

GB RAM memory

2.3 Experimental Set-up

Different types of soft material phantoms with the

same dielectric properties of human tissue have been

experimented upon in the recent years to investigate

SAR and temperature distribution produced by

microwave applicators. An accurate phantom should

closely represent the electromagnetic properties of

the human body in the frequency range of interest

and it should be easy to prepare and to handle.

When small volume of tissue are involved in the

heating process, true biological tissue sample as a

suine liver or a chicken breast appear more suitable

to investigate power deposition near the applicator.

In our case the realized MHA prototype has been

tested by using chicken breast. The prototype,

ending with a SMA male connector, is connected to

a microwave source working at 2.45 GHz, capable

of a maximum power of 300 W continuous (CW) or

pulsed. The input and reflected power of the MHA

is monitored by a power meter connected to a

bidirectional coupler while the temperature,

measured by a thermocouple or a thermistor, is

acquired by an A/D converter, recorded in a data file

and simultaneously displayed on a PC monitor for

the real-time direct control of the heating process.

3 RESULTS AND DISCUSSION

In order to accurately define the heating pattern

volume and avoid unpredictable field distortion or

undesirable tissue overheating, back currents

flowing on the surface of the coaxial feeding line

and currents induced on the metallic wires of the

thermocouple was blocked by a common coaxial

balun. As a result, at the 2.45 GHz operative

frequency, the power deposition in the tissue is

confined within a small well defined ellipsoidal

volume wrapping the radiating section of the

applicator body as shown in Figure 5.

As expected, numerical simulations evidence a

focusing point of the EM fields near the sensing tip

of the sensor that could be responsible of a localised

hot spot and temperature overestimation. In order to

reduce this unwonted spot, miniaturised chip

inductors have been properly inserted in series to the

leads of the temperature sensor near the tip to block

any RF current flowing in the sensing element. In

A MINIMALLY INVASIVE MICROWAVE HYPERTHERMIC APPLICATOR WITH AN INTEGRATED

TEMPERATURE SENSOR

115

our EM model we used lumped inductive elements

with an inductance of 1 nH and we compared

numerically this solution (Figure 6c) with the

reference cases where the tip was kept floating

(Figure 6a) or short circuited to the external coaxial

conductor of the applicator (Figure 6b) that

schematizes the applicator shown in Figure 5.

Figure 5: Normalized SAR distribution of a MHA with a

temperature probe embedded inside the insulator of the

coaxial choke.

If a thermistor is used as temperature sensing

element instead of a thermocouple, the sensor tip is

constituted by a semiconductor with conductivity

ranging between 10

-2

and 10

-4

S/m. In both cases the

simulation result of Figure 6c shows that the

insertion of the micro-chokes practically eliminates

the arising of hot spots near the sensing termination

and therefore drastically reduces temperature

measuring errors.

Figure 6: Normalized SAR distribution of a MHA with the

integrated temperature probe: (a) floating tip (ideal case),

(b) tip short circuited to the coaxial conductor (worst

case), (c) tip RF isolated with inductors (actual case).

At a radial distance of 5 mm from the applicator, in

the direction of maximum radiation, SAR is less

than 50 % of the maximum value calculated at the

surface of the catheter and reduces to 90 % at a

distance of 10 mm. This confirms that the applicator

can be used both for hyperthermic treatments of

small tumours ad also for thermo-ablation surgery,

depending on the maximum applied power and time.

It is worth to note that the device can tolerate 30 W

CW or average power and up 150 W pulsed power

without any damage or excessive self-heating.

The MHA input matching is also numerically

calculated in the frequency range from 2 to 3 GHz,

in absence of the thermocouple and in presence of a

temperature sensor with the probe close to the

applicator and the metallic leads embedded into the

coaxial balun. In both cases a reflection coefficient

less than -20 dB is assured at the 2.45 GHz working

frequency as shown in Figure 7.

Figure 7: Input reflection parameter vs. frequency of the

MHA with (continuous line) and without the integrated

temperature probe (dotted line).

Power distribution in the tissue was experimentally

evaluated by applying 20 W continuous microwave

power to the MHA over a period of about 15

minutes (Figure 8) and a pulse power of 60 W for 10

seconds. The results are in good agreement with the

numerical predictions and proved that the induced

heating pattern in the biological medium assumes

the typical ellipsoidal shape around the applicator

radiating end. The overall dimension of the

ellipsoidal heating pattern clearly depends on the so

called thermal-dose, i.e. on the quantity of the EM

energy delivered to the medium. It is also evidenced

that the metallic sensor tip (thermocouple or

thermistor) do not produces a local hot spot,

authorizing us to state that an accurate temperature

monitoring can be obtained. This was afterward

confirmed through a measure made with an auxiliary

fiber-optic temperature sensor. Using the integrated

BIODEVICES 2008 - International Conference on Biomedical Electronics and Devices

116

sensor, the temperature at the interface between the

catheter and the tissue can be carefully monitored

during the microwave heating process because not

significant self-heating of the sensor has been

observed. Figure 9a shows the evolution of the

temperature vs. time when 20 W CW power is

applied for 15 minutes at the input of the applicator,

while Figure 9b depicts the thermal response of the

medium to a 10 second pulse of 60 W peak power.

It is worth note that the increase of the

temperature is very fast in both cases because the

chicken breast used as phantom for our heating

experiments is not perfused by the blood and

therefore the thermal response of the tissue is

determined only by its thermal conductivity.

(a)

(b)

Figure 8: Thermal pattern in a splitted chicken tissue

obtained by applying 20 W CW power to the MHA for 15

minutes (a) and 20 minutes (b) respectively.

(a)

(b)

Figure 9: Temperature evolution at the tissue/applicator

interface obtained by applying 20W CW input power to

the MHA for 15 minutes (a) and a 10 second pulse of 60W

peak power (b).

4 CONCLUSIONS

The proposed minimally invasive MHA integrates a

very cheap temperature sensor and therefore it is

very suited for the mass-production of mono-use

devices. The integration of the radiating element and

the temperature sensor inside the same applicator

case allows to heat a small tissue volume (target)

and to measure the temperature accurately at the

same time. Due to the use of a simple coaxial balun

the microwave energy is confined around the

applicator body reducing the risk of accidental

overheating of healthy tissue close to the tumour.

The thermistor (or the thermocouple junction) peeps

out from the catheter surface in the point where the

EM field, and hence the temperature, reaches the

maximum. Temperature in deeper zone of the tissue

surrounding the applicator can be extrapolated by

mathematical models based on the bio-heat equation

(Pennes, 1948) if the EM and thermal parameters of

the tissue are known.

A MINIMALLY INVASIVE MICROWAVE HYPERTHERMIC APPLICATOR WITH AN INTEGRATED

TEMPERATURE SENSOR

117

Multiple applicators (arrays) with integrated

temperature sensors could be used in order to treat

larger tissue volumes and more accurately estimate

the temperature distribution through the combined

application of bio-heat equation and tomography

algorithms.

The pliability of the miniaturized coaxial cable

and of the associated silicon catheter make easier the

insertion of the applicator using the natural way of

the body or minimally invasive surgical procedures.

Many coaxial applicators for hyperthermic

treatments show a heating pattern characterized by a

typical tear drop shape. The implementation of a

coaxial choke in the MHA investigated in this paper

reduces appreciably the drop tail and allows to

precisely localize the tissue volume involved during

the microwave treatment. Moreover the high degree

of miniaturization due to the availability of

miniaturized coaxial cable to use in medium-high

power applications (dimensions about 1-2 mm in

diameter are easily available), permits to extend the

clinical applications of this minimally invasive

applicator. Hyperthermic treatments of bile-ducts in

cancer therapy, impracticable in the past for the

restrict dimensions of the ducts, could be possible

nowadays as well other delicate surgical

interventions that require high precision and reduced

invasivity.

The originality of this MHA, from an

engineering point of view, lies in the peculiar

integration of the metallic wires of a low cost

temperature sensor inside of the choke body without

perturbing the SAR distribution.

The invasivity of the clinical hyperthermic or

thermo-ablation treatment is highly reduced using

this type of applicator because in fact no separated

insertions for temperature probes, no additional

external electrodes as for RF treatments or any other

kind of devices are required. Therefore many deep-

seated tumors (e.g. certain brain, liver,

gastrointestinal or gynaecological tumours), could

be effectively and easily treated with the proposed

MHA.

The integrated temperature sensing element

permits to accurately monitor the maximum

temperature reached into the tissue and it can be

used to close the control loop of a specific

microwave hyperthermic or ablative process by

defining the appropriate thermal-dose to be

administered to the lesion.

By monitoring the temperature in time domain,

very useful data on blood perfusion rate, thermal

conductivity and specific heat could be directly

extrapolated by the bio-heat transfer equation and

used to construct more complex and realistic

biological tissue models. As well thermal properties

difference between healthy and pathological tissue

could be relieved in order to extrapolate diagnostic

information.

REFERENCES

Turner, F., 1986. Interstitial equal-phased arrays for EM

hyperthermia. IEEE Trans. Microwave Theory Tech.,

vol. 34, no. 5, pp. 572-578.

Tumeh, A.M., Iskander, M.F., 1989. Performance

comparison of available interstitial antennas for

microwave hyperthermia. IEEE Trans. Microwave

Theory Tech., vol. 37, no. 7, pp. 1126-1133.

Camart, J.C., Despretz, D., Chive, M., Pribetich, J., 1996.

Modeling of various kinds of applicators used for

microwave hyperthermia based on the FDTD method.

IEEE Trans. Microwave Theory Tech., vol. 44, no. 10,

pp. 1811-1818.

Lin, J.C., Wang, Y. 1987. Intertitial microwave antennas

for thermal therapy. Int. J. Hyperthermia. vol. 3, no. 1,

pp. 37-47.

Cerri, G., De Leo, R., Primiani, V.M. 1993. Thermic

endfire’ interstitial applicator for microwave

hyperthermia. IEEE Trans. Microwave Theory Tech.,

vol. 41, no. 6, pp. 1135-1142.

Saito, K., Hayashi, Y., Yoshimura, H., Ito, K., 2000.

Heating characteristics of array applicator composed

of two coaxial-slot antennas for microwave

coagulation therapy. IEEE Trans. on Microwave

Theory and Techniques, vol. 48, no. 11, pp. 1800-

1806.

Bowman, R.R. 1976. A probe for measuring temperature

in radio-frequency-heated material. IEEE Trans. on

Microwave Theory and Techniques, pp. 43-45.

Longo, I., Biffi Gentili, G., Cerretelli, M., Tosoratti, N.,

2003. A Coaxial Antenna with Miniaturized Choke for

minimally Invasive Interstitial Heating. IEEE Trans.

on Biomedical Engineering, vol. 50, no. 1, pp. 82-88.

Jones, K., Mechling, J.A., Trembley, B.S. 1988. SAR

distributions for 915 MHz interstitial microwave

antennas used in hyperthermia for cancer therapy.

IEEE Trans. on Biomedical Engineering, vol. 35, no.

10, pp. 851-857.

Biffi Gentili, G., Leoncini, M., Trembly, B.S., Schweizer

S.E. 1995. FDTD Electromagnetic and Thermal

Analysis of Intertitial Hyperthermic Applicators. IEEE

Trans. on Biomedical Engineering, vol. 42, no. 10, pp.

973-980.

CST Studio Suite 2006, Computer Simulation Technology

GmbH, D-64289 Darmstadt, Germany.

Berenger, P. 1994. A perfectly matched layer for the

absorption of electromagnetic waves. J. Computat.

Phys. vol. 114, no. 2, pp. 185-200.

Pennes, H.H. 1948. Analysis of tissue and arterial blood

temperatures in the resting human forearm. J. Appl.

Physiol. vol. 85, no. 1, pp. 93-122.

BIODEVICES 2008 - International Conference on Biomedical Electronics and Devices

118