Advances In Experimental Medicine and Biology 1990; 267:327-44 TRIPAS : A TRIAPPLICATOR SYSTEM WITH RELOCATABLE HOT SPOT

AT TISSUE DEPTH

Haim I. Bicher, Samuel A. Afuwape, Ralph S. Wolfstein, Duane F. Bruley, and Kenneth Reesman Valley Cancer Institute,!Panorama City, CA 91402

ABSTRACT
oblem of heat focusing and standardization Solving the pr application of hyperthermia requires a of the clinica! mathematical predi ction model. The model should include the ve parameter, and be able to predict medium constituti the microwave applicators to optimize positioning of g and provide for reproducible treatment treatment p!annin set-up. We pre sent a configuration of 3 applicators subtended by an equilateral triangle in order to target and ot for improved treatment of deep tumors. relocate a hot sp A simple geometri c analysis is illustrated. The microwave ofile~ from the three power sources~ was beam absorption pr obtained from pha ntom studies depicting the radiative heat pattern for the triapplicator system (TRIPAS). A complex mathematical model was developed to demonstrate interaction of the beams in the medium. It was observed empirically that under coherent propagation in the near field electromagnetic (EM) waves tend to add at the center, while varying the propagation axial focal length caused a relocation of the summing focal points. Mathematical prediction correlated very well with the phantom studies. SAR values above 100 W/kg were achieved at 12.5 cm phantom depth, creating a relocatable hot spot at the concentric loci of the 3 air cooled horn microwave applicators operating at 300 MHz.

This Project is supported in part by a computational from !BM LASC, Los Angeles Ca, USA!.

grant

Consensus on Hyperthermia for the 1990s Edited by H. I. Bicher et al. Plenum Press, New York, 1990

327

INTRODUCTION The oroblem of microwave deep heating has been extensively analyzed by several authors (Andersen 1984, Johnson et al. 1985, and Turner and Kumar 1982). Although some devices are available for regional deep hyperthermia applied to different areas of the body, the probem of insufficient energy penetration persists. Heretofore this has been only partially solved by multiple phased array applicator configurations (Bach Anderson 1984, Gee et a! !98~, Samulski et al. 1987 and Sato et al. !986). The rationale behind these technioues is the in-ohase, coherent addition of electromagnetic (EMI energy from several strateaically located sources. The selection of a source freauency lower than 300 MHz theoretically enhances deeper penetration while oaradoxically limiting localization of the energy to a defined tumor volume. In prior publications ( Bicher et a!.].979,1980, 1982, 1984, 1985, 1986) we have described the use of single or parallel opposed (POPAS) 300 MHz air cooled dielectric loaded microwave applicators. The purDose of this oape~ is to define the characteristics, mathematical rationale and the oenetration of EM energy of a triaDolicatnr system (TRIPAS) focused on a deep target in muscle phantom experiments. A mathematical model is ~resented based on synthesizing EM waves oroDagated in the biomedium, using its constitutive parameters, the incident energy and effective wavelength to predict and graBhica!Iv demonstrate the convergent hot spot and to relocate the target within the concentric configuration of the three .apD!icators.

MATERIALS AND METHODS Device Construction The TRIPAS system 2* is based on the use of three air cooled dielectric loaded applicators, 20x22 cm in aperture, which have been previous!y described (Bicher et al 1985). The system operates at 300 MHz, and is driven by fixed frequency generator capable of output of up to 1000 watts. The three applicators are mounted on a graduated circular stand (see Figure I) by the use of movable brackets. This allows the applicators to be placed in clos.e proximity to the skin overlying tne target area in a convergent fashion. The a~plicators are attached to brackets by universal ball joints, which allow angling to conform to the body surface. In vitro Experiments !n vitr o experiments were conducted by placing applicators on the three sides of an equilateral triangular phantom box bu~t of plexiglass, measuring 41.5 c{ oer side. The aPplica tors were placed flush against each side, with aperture fo cal points converging at the center of the triangle.
2~

Equipment SUDO!ied by HBC! Medical Group, Inc., 14427 Chase Street #203 manorama City, Ca 91402 USA

328

Figure I. TRIPAS Applicator Stand Construction of TRIPAS system. Note graded stand supporting 3 concentric relocatable air cooled horn applicators. For explanation see text.

329

The distance from the center of the triangle to the midpoint of each aperture was 12.5 cm. The ohantom was split in a plane peroendicular to one applicator, to allow placement of microthermocouDles and liquid crystal thermochromic paper (~CP) between the aDp!:cators (see Figure 2). Changes in the color of the thermochromic sheet indicated the thermal field and ~oint of wave convergence. Colors ranged from black (22 C~ throuQh red to b!ue (30 C), making heat field pattern clearly discernible. The comoosition of the muscle eouivalent ohanzom material was that tabulated by Stuchly and Stuch!y (1980), having a dielectric constant of 50-58 F!m and conductivity of 0.909-0.952 S/m at 300 MHz. SAR Determination
De~erminat ion of the power depositior pattern within the phantom involved monitoring 27 precisely oositioned microthermoco uDles in rectangular matrix on the LCP. The lit phantom was used to place sensoring devices triangular sp in the centr al Diane, normal to one of the three radiating apertures. T he LCP gave the additional advantage of an ure of the heat ~attern induced. The specific overall nict absorption r ate (SAR) from each sensor was determined using the formula: SAR = a!84"c* (A~/Xe ) W/kg [17,18], where c ~s the s~ecific heat capacity of musc:e tissue phantom (0.86

cal/gm/deg. CI, T is the rise in temoerature above normalized temperature (deg. C), and t is the period of time the volume was exposed to the microwave energy. The microwave field was applied to the phantom for 60 seconds prior to each reading. Peak temperatur e changes induced were 2.6 deg. C at 4cm depth, 0.7 deg C at 8cm depth, and 1.8 deg at the summation point 12.5cm deep. A longer duration of exposure to the microwave field was re muired to Drovide the thermocontours simulated in FiQure 2.
Model Derivation

The complex propagation constan~ of EM energy in a medium is readily derivable from hyperbolic wave (Helmholtz) equations using Maxwells fundamental wave equation ~

~ x ~ =( (/- + iw~ ) "~ ............

2 propagation

or a stratified heterogeneous medium (single constant)

................. 3 )
from Maxwells continuity law .-

................. .................

5) 6)

330

APPLICATOR #2

APPLICATOR #3

Significantly Heated Areas

Figure 2. TRIPAS - Muscle Phantom Heating Pattern Positioning of the applicators for a typical measurement of temperature increment using thermochromic paper. Note the high intensity thermal field in front of each applicator, at the central convergent area and within a circle of positive interference surrounding it.

331

Eo.(5) specifies the wave equation while Eq.(6) specifies the complex ~ropagation constant. The general solution of E~.(5) for a homogeneous case is given : "~y : Ae^(-(~z) + Re (~/z) ....... ; ....... 7)

for both forward and backward propagation on z-axis, where A and R are constants described by the medium and excitation characteristics. The complex propagation constant, Ea.(6), can be broken down into its real and imaginary parts :

whereo. (alpha) describes the attenuation factor, expressed in terms of the mediums constitutive parameters :

= w~Z~/ 2) 1/2 [((I + (~-/g w)A 2)A 1/2

1)^1/2 ......

9)

and where~ (beta) describes .the phase factor, expressed in terms of the mediums constitutive parameters :

: w.~A&~12)"~ 112[((I + (,{I~ w)~2)~ 112 + 1)Al12] ........ 10)

A harmonically excited, linearly polarized plane wave traveling in a medium or media with known constitutive parameters can be described. ~y : Er.e^{ -o~ z) e^ i)wt-,~ z) ...................... I].)

Thus precomputing Eq. (9) and Eo. (10) respectively,Eo. (11) can be easily computed and graphically simulated. Plane Wave Interactive Simulation ~he prediction of the spatial heating effect on deep seated neoplastic tissue is mathematically tractable, and can be deduced from E~.(11). !t is well established that a radiating source is most efficient when the physical parmeters of its aperture are comparable to the wavelength of the source being used (Guy et al. !974). Thus, the structure of the applicator and the source wavelength dictate in part, the distribution of an EM field profile. Studying the behavior of a plane wave in an excited medium with its characteristic complex propagation constant, does produce observable interaction of electromagnetic waves within KD the medium (Guy 1971), Kantor and Cetas 1977). The selection of these variables as shown by this simulation, depicts its potential value in Understanding electromagnetic interaction with biological media. This is further enhanced by the use of 3D graphical analysis. For an approximate linearly polarized plane wave, an axial E-field could be written

Er:

(1/( ~r~, +

-I/(~ r)^2, + i/(~ r)^3 + !/(~" r)^n)e^( - ~r) ...........

12)

332

where r is written in rectangular coordinate system : 2 2 2 1/2 r = ((x-x) + (y-y) + (z-z)) ...............
and where n is an integer

13)

2.6. TRIPAS Three applicators ar e configured around an encircled eouilateral triangle to target and relocate the calculated hot spot in order to achieve significant heating in the clinically desired area Simulation of the analytical solution, Eo. (!I), fo r the plane electromagnetic wave propagated in the medium is then implemented. The rationale behind the design of TRIPAS is not different from the theory of wire antennae. Applicator apertures are considered as planar co llections of small electric ( or magnetic) dipoles, each o f which radiates complex energy waveforms as the source to impinge on the tissue individually. A monotone excitation is modelled in the biomedium ~lane by convolut ing ( superimposing) olane wave solutions on a surface expre ssions : Ey = Er. e~(-c~ Ix^2 + ~?-))~ 1/2.e^((x^2 + #A2) 1/2) .cos(6.238(#^2 + y^2)^1/2/ ) ................. ].a) (see notation for definition parameter). It is presupposed that small electric (or magnetic) dipoles radiate complex energy waveforms, the energy contributions of which are s patially superimposed and summed over individual contributio ns at the point where they meet in the biomedium when in ph ase, while negating each other when out of phase. TRIPAS Analysis Assumptions 1) A convergent linearly polarized beam axially propagates on the positive z-axis from a strategically located source. 2) the Each beam simulates the characteristic excitation and mediums complex propagation constant.

3) Heterogeneous layers are stratified to homogeneous layers to enhance oaractical single oarametric values for the medium dielectric constant and conductivity. 4) Each applicator is designed for 300 MHz, EM energy absorption is enhanced by electrical dipoles characteristics in Rayleigh region ( Kritikos et al. 1976 (~R < 2D~F1)). 5) A muscle phantom with dimensions consistent with the effective wavelength of 300 MHz is selected to enhance constructive interference. (This is derived from Nilsson, P. et al 1985.

333

Figure 3 shows the simplified analytical geometry which illustrates the convergence of the beams from three dipoles on the vertices of an equilateral triangle. The point of maximum summation, or the point of minimum substration, is located at the center of an encircled equilateral triangle resulting from the three tangential dipoles (Figure 3, case 1). Assuming the point of convergence as the aperture focal point (R) in the medium, R is selected to be proportional to the effective wavelength of the medium to meet Rayleigh region such as: Target Relocation The relocation of the target hot spot is postulated by angular tilting of two adjacent apertures so as to display a constructive focal point within the quadrants of the circle. Figure 3, cases 2 to 4, display theoretical results of tilting a pair of adjacent applicators. The tilting angle of the pair of adjacent applicators can be derived from simple geometrical analysis: ~ ( theta ) = arcsin ( L/R - 1) ................. 16) where L is the distance from the third aperture to the perpendicular line of adjacent tilted applicators. In other words knowing the precise depth of a tumor in a strategical oreconfiguration of TRIPAS, a constructive interference is postulated to give differential heating effect resulting in the formation of a hot spot. In order to demonstrate the feasibility of this postulation, 3D graphic simulations and thermographic split muscle phantom experiments were implemented as previously described (see 2.1 -2.3). Graphic Analyses For computer simulation of dipoles plane wave, Eq (14), three element sources were located 120 degrees apart around a rectangular co-ordinated circle with its center as the origin ( Figure 3, case 1) Two commercial 3D graphics packages (3*, 4*) were invoked to implement the computer simulation. A fortran 77 routine was written on main-frame to a priori compute the and valves, with the mediums constitutive parameters: relative permittivity, conductivity and the excitation frequency. These were passed into the 3D Curve package to generate a data file. The flexible Golden Surfer package was invoked to produce the 3D output response of the simulations. Selection of the co ordinate system can be obtained from : (O,r), (r,-rsin30), (-r, -rsin30) ................. 17)

3* Curve Three D West Coast Consultants, a202 Genesee Ave, Suite 308~.San Diego CA 92117 4* Surfer Golden Software Inc., 80i I~ Street, P.O.Box 281 Golden CO. 80402

334

Basic Dipoles Configuration

Case 1

Case

Case 3

Case 4

Figure 3. TRIPAS Analysis - Postulated Target Relocation Arrows indicate dipoles strategically located around encircled equilateral triangle. Dipoles sum their plane - wave contributions at the Case l: center. Case 2: Pair dipoles rotated at their tangential mid points by theta to enhance relocation of summation point (target) at the RHS of circular segment. Case 3. Pair dipoles rotated at their tangential mid - point by theta to enhance relocation of summation point (target) at the Bottom circular segment. Case 4. Pair dipoles rotated at their tangential mid points by theta to enhance relocation of summation point (target) at the LHS circular segment.

335

Three ~i~oles circularly located at the above rectanQular co- ordinates with center {0,0). The valve (r) is matched to (R), the axial focal length, though (r) can be arbitrariTy selected. Each co ordinate goes into Eq. 4, respectively :

~y : Er.e (-~c (x+12.5)~2 + (y~5.3)^2)~1/2) ^ .e (-/~ ((x+12.5)^2 + (v-_6.3)^2) 1/2) .cos(6.283 (x+12;5)^2 + (y-6.3)^1/2~Y~) ...... ~y : Er.e (-~ ((x-12.5)^2 -+ (~v~-6.3)~I/2) .cos(6.283(ie (~.5((x-12 5)- 2) 2 ~+ (y-6.3)^2)^1/2) x-1 )~2 + iy-6.3) 1/2~e(~) 18c)

for r = R = 12.5 Creating the plane wave superimposed on the traveling surface wave ( see Figure 3a and Figure 5 ) Results SAR measurement The results of the SAR measurements indicate that the temperature increases obtained were greatest close to each radiating aperture, and then lessened substantially, until a zone of summation is reached, where they create a hot spot The SAR measurements at the 12.5 cm hot spot closely approached those measured at the near field of the aperture (see Figure ~). Phantom Temperature Determination
Thermal responses visualized on the liquid crysta! sheets showed three different zones of preferential heating :A hot spot was located close to #ach applicator. A second clearly defined heat zone appeared at the center of the phantom, at the point of beam interaction. Both of these zones are about 6 cm in diameter. A third heat zone is circumferential about midway between the peripheral and central hot spot ( see Figure 2), about 3.~ cm wide ( one Quarter of the effective wavelength of 13.4 cm at 300 MHz in the muscle Phantom).

Simulation The results of the simulation of EQs. 18a - c are disolayed 3-dimensionally in Figure 5. The inputs for computing the complex propagation constant and its components (~,~) simulated in the muscle tissue excited at

336

SUMMARY RESULTS 11/19/86 PM 12.5 cm (TRIPAS) PILANTOM 60 Second Time 600 WATTS Power

ANTENNA #I

TENNA #2

85

73

38 " ~ 9cm2~8 = _ ~

109

84

ANTENNA #3 SAR TRIPAS 600 WATTS (I MINUTE)

Figure 4. TRIPAS - SAR Profiles SAR (specific absorption ratio) measured in a triangular muscle phantom, using 300 MHz at 600 Watts equally split among three equivalent applicators (dimension,design) for 60 seconds. SARs in the area of summation approximate those near the front of the eo~licator aDeratures.

337

300 MHz, were : permittivity value of 54 F/m and conductivity value of 1.07 S/m, with free space permeability value of !2.57x10 H/m and the permittivity of free space as 8.854xi0 F/m. These parameters were apDropiately scaled as input variables into the graphic package for the data file. The selection of effective wavelength (~eff) can be calculated theoretically, but in this case was extrapolated from the muscle ohantom heating pattern studies. A typical value of ~eff= 13.a cm was used ( see SAR measurements ). The dispersive effect of the plane wave was accomplished by the addition of electromagnetic plane wave solution with surface expression ( see Eas.(14) and (15)). Figure 5 displays the 3D output surf ace plot of the simulation. The letter marker A,B and C i ndicate the origin of the diooles , strategically located at 120 degree separation in circular configuration. The arrow marker points to the central summation hot spot. The main features highliahted by these 3D plots are : 1) high intensity singularity in front of each simulated aperture, 2) rapid decay in amplitude intensity due to generation of destructive interference from subtractive surface waveforms, 3) at even greater depth a second high intensity area due to additive interference causina a peaking standing wave, 4) followed by subtractive interference creating an intensity valley and 5) finally producing the theoretically predicted summation at the centrally located hot spot The high intensity peak at the aPerture (source) is a common observation in experimental and theoretical modellings (Bach Andersen 1984, Gee et al .1984, Kantor and Cetas 1977), that accounts for near- field effects. The surface dispersive effect is predominant at the viewing angle in this simulation. The amplitude steering (modulation) of the target hot spot is implemented by the inclusive term E Eqs. 18a-c. It should be noted that the above surface plot is in very good agreement with experimental observations as described earlier, (comparing Figures 2 and 5). Note, this specifically applied to case i of Figure 3 relating to the theoretical postulation.
Discussion

The present results make use of a mathematical model of radiating dipoles to predict the coherent and noncoherent interference of three converging electromagnetic wave sources interacting within biological media to produce a relocatable hot s~ot The muscle-~hantom experimental results in part verified this postulation. Previous studies (Bach Andersen 1984,Turner and Kumar 1982, Wait 1986),solved domain-integral equations for electric field vector ~otential as a result of induced magnetic current elements derived from fundamental Maxwells classic microscopic equations, provided versatility of implementing the medium heterogeneity. Turner and Kumar (1982,84) used Huygens principle to simulate a horn-tyoe apolicator as an array of point-source dipole radiators, in order to predict the heating patterns and ~erformance of the

338

Figure 5. TRIPAS - Simulatien 3D surface plot displaying pattern of split 300MHz microwave beams into muscle phantom. Note:l) the area of central converaence indicated by arrow(as postulated in Fig. 3 case I), ~) high intensity thermal field in front of each applicator (A,B, and C) and 3) circular area of positive interference at approximate I/8 wavelength depth between the applicator and the convergent center.

339

aperture. Such mathematical model was limited in its ability to soatial!y predict the interaction of harmonically excited plane waves in biological medium. Johnson et al. (1984,85) and Sato (1986) analyzed circular compact arrays simulating resonance conducting p~ates as magnetic current sources. A mathematical integration over the aperture surface yields field distribution. This computational method is limited by the convergence or divergence of the numerical technioue used for the integration. Bach Andersen (1984] modelled an array of dipoles to explicitly define radiative antenna gain for oower density at a given ~oint relative to the total input Dower. This concept was used to analyze radiative apertures and their penetration capability in biological tissue. This author (Bach Andersen 1984) further showed the theoretical synthesis of one- dimensional convergent beams meeting at a to produce constructive interference. He focal ooint p~ctorially demonstrated that two waves traveling in oDoosite directions produce either an additive coherence or subtractive noncoherence effect. The present investigation followed the same theoreticaI postulation. For cylindrical geometry the larQe number of waves add to produce power intensity convergence at the axis (Lin 1982). This central cylindrical ~ocusing is oenerally applied in regional heating of cylindrical tissue, (Samu!ski et al., 1987 and Turner 1984), but lacks the flexibility to relocate a localized hot spot Nilsson et al. (1985) tilted two applicators at 45 degrees. They reported a distinct heatin~ pattern when both the radiative source had no phase relationship, while constructive and destructive interference were observed when both app!icators were excited in Phase coherently. Gee et al. (1984), Bach Andersen (1984) and Ling et al (1984) Presented theoretical and experimental models of near ~eld ~hased array focusing analyzed in 3D plane wave profiles. The alternative use of an interactive Diane waves model might ease the limitations inherent in the above simulation techniques. Earlier, we ~e~orted (Bicher et al 1985) an experimental parallel opposed s~stem that achieved therapeutic temperatures in the 42 to 43 degree Celsius range at 8 cm depth in both muscle e~uivalent phantom and in pigs 20-30 cm thicK. !n other papers (Bicher et al 1982, 84, 86) we defined the clinically usable penetration using one air coo~ed applicator operated at 300 MHz to be 5 cm. Usinm POPAS (two o~oosed applicators operated in phase), we were able to heat theroeutically to a depth of 8 cm. By olacinm one aDolicator anteriorly and the other posteriorly, treatment of tumors at moderate depth, in thorax, abdomen and pelvis, have proved successful. Our results of in vitro experiments and mathematicaI simulation ( Figure 3, case i) clearly show effective heatinQ at a depth of 12.5 cm using 3 convergin~ microwave

340

beams from standard 300 MHz external applicators. The flexibility of the system (TRIPAS) and the ability to change the incident angle of the beams to relocate the hot spot within the target volume, make the system extremely practical for the treatment of deep seated tumors, without the need to heat an entire body region as is required with current phased array systems which operate at lower freauencies. P~sitioning of the air gap applicators around the patient is quick and easily reproducible. These aDolicators have poor coupling however, so there is some energy loss. ~n the present system this is compensated by the use of a high power generator operating at fixed freouency (300 MHz), capable of 300 watts per applicator. Thus a 900 watts generator for TRIPAS wi!l be required for clinical use. Such a device has already been built and is operational in our laboratory. The good correlation between our mathematical predictions and phantom studies validates the theoretical approach used. The ability of the TRIPAS system to predict the location of the hot spot within the treatment field makes computer ~lannina o~ the applicators ~ositionino in the clinical situation feasible. Nomenclature D = aD~licator aperture dimension (cm) Er = axial spatial electric field magnitude Ey : electric field complex vector (v/cm) f = frequency (MHz) = magnetic field omplex vector (A/cm) = imaginary operator (Sq.root (-1)) = numerical approximate integer for axial E field = distance from the aperture to target - tumor (cm) = axial aperture focal length (cm) : graphical radius x,y,z = rectangular coordinate system = 2~f ; angu!art frenquency Irad) = attenuation factor (Np/m) = phase factor (rad/m) =conductivity of the medium (S/m) = relative dielectric constant ./~ = free space ~ermeabiliti (H/m)= complex ~ropagation constant (H.Np/m) eff = effective wavelength (cm/ cycle) (theta) = tilting axial aperture angle (deg) = vector differential operator ~= laplace o~erator = variable incrememt REFERENCES 1) Bach Andersen, J~ 1984, Eletromagnetic heating - a review. Proceedings of the 4th International symposium on Hyperthermia Oncology, Aarhus, Vol. 2, ed. J. Overgaard (London, Philadelphia: Taylor & Francis, 1984), pe. 113 128.

(v/cm)

341

2) Bach Andersen, J: 1984, Theoretical limitation on radiation into muscle tissue. International Journal of Hyperthermia, 1, a5 55. 3) Arcangeli, G., Lombardini, P.P., Lovisolo G. A. Marsiglia G. and Piattelli, M., 1984, Focusing of 915 ~Hz electromagnetic power on deed human tissue: a mathematical model study. IEEE Transactions Biomedical Engineering, BME 31, 47 -5a. ~) Bicher, H. I., Sandh~, T. S., and Hetzel, F. W., 1979, Hyperthermia as a radiation adjuvant : An effective fractionati.on regime. International Journal of Radiation, Oncology, Biology. Physics. 5,!71. 5) Bicher, H.I., Sandhu, T.S., and Hetzel, F.W., 1980, Hyperthermia and radiation ~n combination. International~ Journal of Radiation, Oncology, Biology, Physics, 6, 867 870. 6) Bicher, H. I., Hetzel, F. W., and Sandhu, T. D.S., 1982. Results of a ohase I/II clinical tria! of fractionated hyperthermia in combination with low dose ionizing radiation. Advance Exoerimental Medical Biology, !57, 87 97. 7) Bicher, H. I., 1984, The Set up of the state of art HyDerthermia clinic - therapeutic, "ethical, and mractical considerations. Frontiers in Radiation Theraoy Oncology, 18, 153 161. 8) Bicher, H. I., and Wo!fstein, R. S., 1984, Microwave hyDerthermia as an adjuvant to radiation therapy. Summary experience of 208 mu~tifraction treatment cases. H.v~erthermia Oncology, vol. 1, ed. by J. Overmaard (London, Phi!adelDhia : Taylo~ & Francis, 1984),~. 363--366. 9) Bicher, H. I., Moore, D. W., and Wolfstein, R. S., 1985, Air cooled 300 MHz ADplicators used in a marallel opposed phased (POPAS). International Journal of Radiation, Oncology, Biology, Physics, 11,$I, 2!7. !01 Bicher, H. I., and Wolfstein, R. S., 1986, Microwave hyperthermia as an adjunct to radiation therapy: Summary experience of 256 multifraction treatment cases. Interationa~ Journal of Radiation, Oncology, Biology, Phisics 12,1667 1671. 11) Bicher, H. I., and Wo~stein, R. S., 1986 Loca! hymerthermia for deep tumors experience with technimues. Journal of Microwave Power, 21, 103. three

12) Gee, W., ~ee, S., Bong, N. K., Cair, C. A., Mittra, R., and Magin, R. L., 1984~Focused array hyperthermia apnlicator: therory and ex.periment. IEEE Transactions. Biomedica!; Engineering, BME 31, 38-46.

342

13) Guy, A W., 19781, Electromagnetic fields and relative heating petterns due to a rectangular aperture source in direct contact with bilayered biological tissue. IEEE Transactions. Microwave Theory and Technique, MTT -19, 214 223. 14) Guy, A. W., Lehmann, J.T., and stonebridge, J.B., 1974, Therapeutic application of electromagnetic Dower. Proceedings of the IEEE, 22,55 -75. 15) Guy, A.W., Lehmann, J.F., stonebridge, J.B., and Sorensen, C.C., 1978, Development of a 915 MHz direct contact apDlicator for therpeutic heating of tissues, leee Transactions. Microwave Therory and TechniQues, MTT 26, 550 -556. !6) Johnson, C.C., Durey C.H., and P!enk, H. P., 1977, Exnerimenwith 915 MHz hyperthermia treatment of malignant tumors. Journal of Bioengineering. 1, 511 - 530. !7) Johnson,R.H., James, J. R., Hand, J.W , 1984, Hopewell, J. W, Dun!op, P.R.C, and Dickson, R.J., New low- profile apnlicators for local heating of tissues. IEEE Transactions. Biomedical Engineering, BME - 31, 28 - 37. 18) Johnson, R. H., Andrasic, g., Smith, O. L., and James, J. R., 1984, Field penetration of arrays of compact a~plicators in localized hyperthermia. International Journal of Hyperthermia I, 29 -43. 19) Kantor, G., 1981, Evaluation and survey of microwave and radiofreQuency applicators. Journal of Microwave Power 16, 135 - !50. 20) Kantor, G., and Cetas, T. C., 1977, A comparative heating pattern study of direct - contact applicators in microwave diathermy. Radio Science, 12, 6S, 111 - 120. 21) Kritikos, H. N., Foster, K. R. and Schwan, H. P., 1981, Temperature profiles in spheres due to electromagnetic heating. Journal of Microwave power, 16 (3&4), 327 - 344. 22) Lin,J.C.,Kantor G., and Ghods, A., 1982, A class of new microwave therapeutic applicators. Radio science~ 17 5S, ]195 - 1235. 23) Ling, H., Lee, S.W., and Gee, W., 1984, FreQuency optimization of focused microwave hyperthermia applicators. Proceedings of the IEEE, 72, 224 -225. 24) Nilsson, P., Larsson,T., and Persson, B., 1985, Absorbed power distribution from two tilted waveguide applicators. ~nternational Journal of hyperthermia, i, 29 -43. 25) Samulski, T. \I., Lohrbach, A., !987, Kapp~ D. S.~ Fessenden, P.~ and Heating deed seated eccentrically

343

located tumors with an annular phased array system: A comoarative clinical study usino two annular array operating confiauration. International Journal of Radiation, Oncology, Biology, Physics, 13, 83 94. 26) Sato, G., Shibata, C., Sekimukai, S., Wakabayashi, H., and Mitsuka, K., 1986, Phase - controlled circular array heating eauioment for deep - seated tumors: Preliminary experiments. IEEE Transaction. Microwave Theory and Techniques ~TT 34, 520 524. 27) Turner, P. F., and Kumar, L., 1982, Comouter solution for applicator heating patterns. National Cancer Institute Monograph, 61, 52! 523 28) Turner, P. annular phased Engineering, BME F., 1984, Regiona! hyperthermia with an array. IEEE Tranaction. Biomedical 31, !06 - 1!9.

29) Wait, J. R., !986, Analysis of the radiation leakage for a Four - aperture phased - array applicator in hyperthermia therapy. IEEE Transactions, Microwave Theory and Techniaues Mtt 34 539 541.

344