Improved Localization of Energy Deposition in Adaptive Phased ...

• FENN, SATHIASEELAN, KING, AND STAUFFERImproved Localization of Energy Deposition in Adaptive Phased-Array Hyperthermia Treatment of Cancer

VOLUME 9, NUMBER 2, 1996 THE LINCOLN LABORATORY JOURNAL 187

Improved Localization ofEnergy Deposition in AdaptivePhased-Array HyperthermiaTreatment of CancerAlan J. Fenn, Vythialingam Sathiaseelan, Gerald A. King, and Paul R. Stauffer

■ Elevated cell tissue temperature (hyperthermia) due to radio-frequency (RF)energy absorption is known to produce an improved response for malignanttumors in humans when applied in combination with other anticancermodalities. However, clinical studies in thermotherapy have shown the difficultyof localizing RF energy deposition in malignant tissue deep within the bodywithout damaging surrounding healthy tissue. The study presented in thisarticle involves a preclinical investigation of adaptive feedback and computercontrol of amplitude and phase from coherent RF antenna arrays to provideimproved distribution of electromagnetic energy deposition in the body.Measurements in a treatment-planning phantom irradiated with an adaptiveantenna-array applicator show that noninvasive adaptive nulling can reduce theRF energy absorption in nearby healthy tissue while focusing energy into adeep-seated tumor site.

O , many clinicalstudies have established that elevated celltissue temperature (hyperthermia), induced

by electromagnetic energy absorption in the radio-frequency (RF) band, significantly enhances the effectof chemotherapy and radiation therapy in the treat-ment of malignant tumors in the human body [1–8].Ideally, hyperthermia treatments with RF radiatingdevices are administered in several treatment sessions,in which the malignant tumor is heated to a tempera-ture above approximately 42°C for thirty to sixtyminutes. Figure 1 illustrates how this hyperthermiatreatment (or thermotherapy, as it is also called) isperformed. During treatments with noninvasive RFapplicators, clinicians have had difficulty adequatelyheating deep tumors while preventing surroundinghealthy tissue from incurring pain and damage due toundesired hot spots greater than 44 to 45°C [9]. Two

previous articles in this journal have discussed thetopics of adaptive nulling and adaptive focusing [10,11]. Since then we have significantly improved thespeed of the adaptive-nulling process and expandedthe scale of our experiments. This article describes apreclinical investigation of adaptively controlledphased-array transmitting antennas with multiple RFfeedback sensors to provide improved distribution ofelectromagnetic energy deposition in malignant andhealthy tissues within the body.

The electromagnetic-energy absorption rate in tis-sue, sometimes referred to in the literature as the SAR(specific absorption rate, or absorbed power per unitmass), has units of Joules/kg-sec (or W/kg) and maybe expressed as

SAR = 12

2σρ

E , (1)


188 THE LINCOLN LABORATORY JOURNAL VOLUME 9, NUMBER 2, 1996

where σ is the tissue electrical conductivity (S/m), ρ isthe tissue density (kg/m3), and E is the magnitudeof the local electric field (V/m) [4]. In Equation 1, thequantity

12

2σ E

is the time-average RF power density converted toheat energy, and is called the dissipated power. If we ig-nore thermal-conduction and thermal-convection ef-fects, which are not important until after a significanttemperature rise occurs, the initial temperature rise∆T (°C) in tissue is related to the specific absorptionrate by

∆ ∆Tc

t= 1SAR , (2)

where c is the specific heat of the tissue (in units of

Joules/kg-°C), and ∆t is the time period of exposure(sec) [4]. Substituting Equation 1 in Equation 2yields a relation between the induced temperature risein tissue and the applied electric field as

∆ ∆Tc

E t= 12

2σρ . (3)

Thus by modifying the local electric-field ampli-tude, we directly affect the local energy absorptionand induced temperature rise in tissue. For example,in malignant tissue we would like to deposit an elec-tric field of sufficient magnitude to heat the tumorvolume to a therapeutic temperature typically in therange of 42 to 45°C. At the same time, we would liketo limit the SAR magnitude in nearby healthy tissueto be less than that within the tumor in order to keepthe healthy tissue temperature below approximately44°C. Multi-element incoherent or phase-coherent

FIGURE 1. Hyperthermia treatment with radio-frequency (RF) radiating devices. A noninvasive adaptive phased-array applicator produces RF electromagnetic energy to heat deep-seated tumors in the human body. Theadaptively controlled phased-array transmitting antennas, along with RF receiver feedback probes located onthe skin and inside the tumor, provide improved distribution of electromagnetic-energy deposition in both ma-lignant and healthy tissues within the body. The computer screen shows a cross section of the patient’s torsowith the tumor in red. Current clinical hyperthermia systems, which do not utilize adaptive phased-array tech-niques, produce undesired RF hot spots that can result in pain and damage to healthy tissue.

Water bolus

Patient

Adaptivephased-array

controller

Adaptive arraytransmitting antennas

Adaptive arraytransmitting antennas

RFtransmitter/receiver

Adaptive nulls

Tumor



array antenna systems can provide significant flexibil-ity for shaping the SAR distribution [4, 12–16].

During clinical treatments, electromagnetic radiat-ing-array applicators have often been used in an inco-herent mode, in which the power delivered to eachradiating applicator is automatically adjusted on thebasis of temperature-sensor feedback measurements.In many cases, incoherent 915-MHz microwave arrayradiation has provided effective heating of superficialtumors, but does not provide adequate heating of tu-mors deeper than two centimeters. In propagatingthrough human tissue, the electric field produced byradiowave antennas attenuates rapidly, with increas-ing attenuation at the higher microwave frequencies.Lowering the frequency into the RF region, to belowapproximately 150 MHz, helps the electromagneticwave penetrate more deeply, but still does not pro-duce a higher value of SAR at depth compared to thevalue of SAR at the surface.

To increase the value of SAR at depth relative tothe surface SAR, we must geometrically focus energydeposition from multiple electric fields. Because ofconstructive interference of electric fields at the in-tended focus and destructive interference of electricfields away from the focus, multichannel coherentphased-array applicators can theoretically providedeeper tissue penetration and improved localizationof the absorbed energy in deep-seated tumor regions

compared to incoherent array applicators [4, 14]. Un-fortunately, because of complex scattering within thehuman body and instrumentation variations of hy-perthermia phased-array system hardware [17], clini-cians cannot always accurately predetermine ormanually adjust the optimum settings for outputpower and phase of each antenna to focus heat reli-ably into the deep-seated tumors.

Initial investigations of nonadaptive pretreatmentplanning at Northwestern Memorial Hospital in Chi-cago used a commercial deep-heating RF phased di-pole ring-array hyperthermia system [18]. The com-mercial ring array used for this experiment has a60-cm diameter and normally surrounds a patient’storso. A temperature-controlled water bolus fills theregion between the patient’s torso and the ring array.A light-emitting diode (LED) matrix phantom,which simulates a cross section of the human torso,was constructed to display the effects of manually ad-justing the amplitude and phase of the array antennas[19, 20]. Figure 2 shows the LED matrix phantom[21], which consists of 137 LED sensors positionedin an elliptically shaped Plexiglas plate with a square-grid diode spacing of 2 cm. The LED matrix is posi-tioned within an elliptical cylinder of homogeneoussaline solution contained within a 2-mm-thick hardplastic shell. The electrical conductivity of the salineis chosen to be similar to body tissues, which results

FIGURE 2. A light-emitting diode (LED) matrix phantom, which simulates a cross section of a human torso, is used inpretreatment planning for clinical hyperthermia treatments. The photographs show the LED dipole-sensor array re-moved from the inside of the elliptical phantom shell. The LEDs glow with an intensity proportional to the local electro-magnetic field generated by a hyperthermia phased-array applicator. For treatment planning, the LED sensor array isplaced inside the phantom shell and the phantom shell is filled with clear saline solution.

5 cm

SalineLED

dipolesensor array



in a 100-MHz RF power-density attenuation of ap-proximately a factor of two for each 3.0-cm distancethat the RF waves propagate into the phantom. Theelliptical cylinder has a cross section of 24 cm by 36cm, similar to the cross section of a human torso, andhas transparent ends for viewing the LEDs. Thelength of the LED metallic leads forming the dipolereceive sensor is nominally 5 cm from tip to tip. Thelight output from the LEDs is directly proportionalto the local electric-field strength generated by the RFring array.

This LED matrix phantom has been used for pre-treatment planning of nonadaptive clinical hyper-thermia trials in the following manner [19]. With thematrix phantom load centered in the RF ring-arrayaperture, the operator begins by manually adjustingthe RF power amplifiers and phase shifters until theLED phantom visually demonstrates maximum elec-tric-field strength in the planned tumor target with asfew hot spots as possible in healthy tissue, as shown inFigure 3. These manual adjustments can take overtwo hours to complete. Then the patient is substi-tuted for the phantom and the clinical treatment isconducted—with the assumption that the irradiationpattern does not change substantially after the patientis substituted for the phantom. Because pretreatmentplanning with this manual trial-and-error adjustmentprocedure often produces unacceptable RF hot spotsin healthy tissue for the required deep-tumor heating

[19], we are now experimentally investigating pre-treatment planning using much faster computer-con-trolled adaptive nulling [22, 23] for eliminating theunwanted hot spots in the LED phantom.

For us to localize energy deposition for an appro-priate temperature rise in a deep-seated tumor, Equa-tion 3 indicates that we must first monitor the elec-tric-field magnitude E received at one or morefeedback probes [24] located within the tumor andadjacent healthy tissue, and then adjust the amplitudeand phase of each transmitting antenna of the arrayfor maximum RF power deposition within the tumorand minimum power deposition in nearby healthytissue. Figure 4 illustrates the equipment setup thatperforms this process. A gradient-search computer al-gorithm that modifies antenna-array input param-eters (drive signals) on the basis of the rate of changeof system output parameters (power-deposition pat-tern) can be used to adaptively determine the indi-vidual antenna power and phase input signals tomaximize (focus) or minimize (null) the electromag-netic radiation measured at one or more feedbackprobe positions [22, 23].

The adaptive-nulling approach used in this articleis based on algorithms and testing techniques devel-oped for adaptive phased-array radar and communi-cations signal processing systems [25]. The resolutionwidth of an adaptive null is approximately equal tothe half-power radiation beamwidth of the adaptive-

FIGURE 3. Sample nonadaptive pretreatment planning session with the LED matrix phantom. (a) The computed-tomog-raphy-scan data show the presence of a large rectal tumor. (b) After approximately two hours of manually adjusting thephase and power settings of the phased-array transmitter, the LED sensor array displays an electric-field pattern withlocal heating of the tumor area but with other hot spots in healthy tissue.

(a) (b)

Tumor region

Hot spots

Tumor



array antenna. This property allows an adaptive nullformed on the surface of the body to reduce the elec-tric field in regions, such as healthy tissue, that extendto some depth below the null. For hyperthermia ap-plications, adaptive-array nulls that reduce the RFpower deposition by approximately 50% or more areusually sufficient to eliminate undesired hot spots inhealthy tissue.

Experimental Setup

For the current experiments, a frequency of 100 MHzwas used for the ring-array system, and the sum of theinput power to all four channels was held constant at860 W. An invasive catheter with a dipole electric-

field sensor one millimeter in diameter was posi-tioned at a depth of approximately eight centimetersin the lower half of the phantom and used to measurethe local electric field at the simulated deep-seated tu-mor site. Three independent noninvasive RF feed-back probes, spaced circumferentially at ten-centime-ter intervals, were attached to the surface of thephantom, as shown in Figure 5. These probes wereused to measure feedback signals for reducing localpower deposition on the upper surface [26].

The goal of the experiment was to irradiate onlythe lower portion of the phantom, which containedthe simulated tumor, while minimizing irradiation ofthe upper portion, which contained simulated

FIGURE 4. A minimally invasive system of four coherent electromagnetic radiating antennas used to heat a deep-seatedmalignant tumor. The RF input power and phase of each radiating antenna are computer controlled with power-deposi-tion feedback measurements from RF feedback probes attached to the body’s surface and inside the tumor. The powerand phase delivered to the radiating antennas can be adjusted so that the electromagnetic radiation is simultaneouslyincreased at the tumor and decreased at the surface sensors. At RF frequencies around 100 MHz, the nulls formed onthe surface of the body are sufficiently broad (as indicated by the gray shaded regions) that nearby regions of healthytissue within the body are protected from RF irradiation.

RFsource

Tumor target

Power amplifiers,phase shifters

Power,phase

commands

Adaptive nulling/focusing

algorithm

Receiver

Computer

RF radiatingantenna

Adaptive phased-array

controller

Signals fromRF feedback probes

Bodycross section

P4, 4

RF sensor

Null zones

Focus

RF sensor

RF sensor

φ

P3, 3φ

P2, 2φ

P1, 1φ



healthy tissue. At 100 MHz, the RF wavelength inthe saline phantom is approximately thirty centime-ters. The half-power beamwidth (or null width) of anadaptive ring array is approximately equal to one-halfthe wavelength, or fifteen centimeters. Thus an in-tense null formed on the surface of the phantomshould reduce the electric field by about 50% at adepth of fifteen centimeters. Less intense surface nullswould have less effect on reducing the electric-fieldintensity at depth.

The input power and phase delivered to each of thefour RF radiating dipole pairs of the ring array weremanually set to equal initial values of 215 W and 90°,respectively. The computer, as illustrated in Figure 4,started the adaptive-array algorithm by automaticallyadjusting, via digital-to-analog converters, the poweramplifiers and phase shifters in each of the four chan-nels of the phased array. The computer software per-formed calculations of the rate of change of the mea-sured RF power at the surface sensors (simulatedhealthy tissue regions) after each adjustment of RFpower and phase to the array transmit channels. Forthis experiment we used a modified method-of-steep-est-descent algorithm to determine the input powerand phase commands that minimize the summationof the local power deposition measured by each sur-face RF feedback sensor. All adjustments were com-

pleted and the adaptive nulls were formed in approxi-mately two minutes, which is an appropriate speedfor real-time use in optimizing clinical treatments.

Before adaptive nulling, both the RF feedback sen-sors and the LED phantom indicated multiple hotspots. The light-toned bars in Figure 6 indicate themeasured RF feedback data before adaptive nulling,and the light-intensity pattern in the photograph inFigure 7(a) indicates the associated deposition of RFenergy, including hot spots. Then the adaptive-null-ing algorithm was executed for three iterations to re-duce the RF feedback signal at each surface-nullingsensor by at least a factor of two. The RF feedbacksensors and the LED phantom then displayed theelectric-field distributions after adaptive nulling, asshown by the dark-toned bars in Figure 6 and thelight-intensity pattern in Figure 7(b). The simulatedtumor position in the lower half of the phantom isfully irradiated while the upper half of the phantom,containing the region of simulated healthy tissue, hasa substantially reduced electric-field intensity.

We then ran a second experiment to attempt tonull the electric field noninvasively over the right halfof the phantom, rather than the upper half. Figure 8shows the measured RF feedback data before and af-ter nulling. As in the first experiment, adaptive null-ing significantly increases the RF power deposited in

FIGURE 5. Test configuration for adaptive-nulling hyperthermia phased-array experiments. (a) LED matrix phantom withRF feedback probes mounted on the surface of the phantom. (b) The LED phantom is placed within the RF ring arrayand the water bolus is filled. The adaptive-nulling phased array consists of four active transmit channels, each con-nected to a pair of RF radiating dipoles.

(b)(a)

RF feedbackprobes

Waterbolus

LEDdipole

sensors

RF ring arrayRF

radiatingdipoles



the tumor site while significantly decreasing thepower delivered to the healthy tissue sites. Figure 9shows photographs of the phantom before nullingand after the nulling algorithm has converged. Thesedata clearly show that it is possible to restrict the heat-ing to the left side of the target body.

In order to protect large volumes of healthy tissuewhile focusing RF energy deep within the body, wemust have the essential ability to produce multipleelectric-field minima. Fortunately, these minima canbe easily generated by using noninvasive RF feedback

0.0

0.2

0.4

0.6

0.8

1.0

1.2

TumorHealthy tissues

No

rmal

ized

po

wer

dep

osi

tio

n

Before adaptive nulling

After adaptive nulling

FIGURE 6. Measured electric-field probe data in a pre-treatment-planning phantom. These data demonstratethe effect of adaptive nulling at three independent sur-face sites. The adaptive nulling protects regions ofhealthy tissue while an 8-cm-deep tumor site is irradi-ated with a coherent four-channel ring phased-array sys-tem operating at a radio frequency of 100 MHz. The light-toned bars indicate the normalized RF power depositionat each electric-field sensor before adaptive nulling. TheRF power deposition in healthy tissue, prior to adaptivenulling, is greater than the RF power deposition in the tu-mor. The RF power deposition after adaptive nulling, in-dicated by the dark-toned bars, measured at the simu-lated deep-seated tumor site increases by 19%, while theRF power deposition measured by the three electric-fieldfeedback probes on the surface of the phantom is re-duced by 91%, 57%, and 87%, respectively. The measure-ments demonstrate that the adaptive-nulling process re-sults in a stronger irradiation of the tumor compared tothe irradiation of the superficial healthy tissue.

FIGURE 7. Pretreatment-planning LED matrix phantomirradiated by an adaptively controlled coherent RF ringarray operating at 100 MHz. (a) Before adaptive nullingthe light-intensity pattern of the two-dimensional LEDdisplay reveals hot spots along the top, bottom, and leftand right surfaces of the phantom as well as in the cen-tral region. (b) After adaptive nulling at three positionson the upper-half surface of the phantom, the RF irradia-tion is concentrated over the lower portion of the phan-tom (the position of the simulated tumor), while irradia-tion of the upper portion of the phantom (the position ofhealthy tissue) is significantly reduced. The experimentshows that the effect of the noninvasive adaptive-nullingprocess is to shift the RF irradiation away from thehealthy tissue areas and toward the tumor.

(b)

(a)LED

phantom

probes located on the tissue surface near each desiredelectric-field minimum. Our measurements in asimple homogeneous LED phantom indicate that we

Treatment region

NullNull

Null



can adaptively control the transmit power and phasedistributions in an electromagnetic hyperthermiaphased-array system, and thus optimize the RFpower-deposition distribution prior to treatment byusing a treatment-planning phantom. In addition tothe control of power-deposition distributions byadaptive-array phase and power adjustments with RFfeedback as reported here, subsequent minor adjust-ments of temperature in a perfused tumor can bemade during the clinical treatments by using localtemperature feedback measurements to adaptivelycontrol the total power delivered to the hyperthermiaapplicator as a function of time [27, 28].

Conclusion

An adaptive power-deposition feedback and controltechnique has been investigated experimentally intwo dimensions with an LED matrix phantom and anelectromagnetic four-channel dipole-array hyperther-mia system with adjustable amplitude and phase. Themeasured data demonstrate that the distribution ofelectromagnetic-energy absorption generated by a

FIGURE 9. Pretreatment-planning LED matrix phantomirradiated by an adaptively controlled coherent RF ringarray operating at 100 MHz. (a) Before adaptive nulling,the light-intensity pattern of the two-dimensional diodedisplay reveals hot spots along the top, bottom, and leftand right surfaces of the phantom as well as in the cen-tral region. (b) After adaptive nulling at three positionson the right-half surface of the phantom, the RF irradia-tion is concentrated over the left portion of the phantom(site of the simulated tumor position), while irradiationof the right portion (simulating healthy tissue) is signifi-cantly reduced. As in Figure 7, the experiment showsthat the effect of the noninvasive adaptive-nulling pro-cess is to shift the RF irradiation away from the areas ofhealthy tissue and toward the tumor.

0.0

0.2

0.4

0.6

0.8

1.0

1.2Before adaptive nulling

After adaptive nulling

TumorHealthy tissues

No

rmal

ized

po

wer

dep

osi

tio

n

hyperthermia phased-array applicator can be im-proved by an adaptive-nulling feedback and controlalgorithm with multiple independent RF feedbackprobes. Further investigation of this adaptive phased-array control procedure should be initiated for im-proving the localization of heating in deep-seated tu-mors by providing real-time compensation forvariable blood flow.

FIGURE 8. Measured RF feedback data before and afternulling (second experiment). The normalized RF powerdeposition after adaptive nulling, indicated by the darkbars, measured at the simulated deep-seated tumor siteincreases by 106%, while the RF power deposition mea-sured by the three feedback probes on the surface of thephantom is reduced by 77%, 45%, and 78%, respectively.

LEDphantom

(b)

(a)

Treatmentregion

Null

Null

Null



Acknowledgments

We would like to thank Alden Hayashi for a criticalreview of the manuscript. The support of the LincolnLaboratory Advanced Concepts Committee is sin-cerely appreciated. The authors are grateful to D.H.Temme of Lincoln Laboratory for technical discus-sions. This work was sponsored by the Department ofthe Air Force.

R E F E R E N C E S1. P.K. Sneed and T.L. Phillips, “Combining Hyperthermia and

Radiation: How Beneficial?” Oncology 5, Mar. 1991, pp. 99–108.

2. E.J. Hall, Radiobiology for the Radiologist (Lippincott, Philadel-phia, 1994), chap. 16.

3. J. Overgaard, D. Gonzales Gonzalez, M.C.C.H. Hulshof, G.Arcangeli, O. Dahl, O. Mella, and S.M. Bentzen, “Hyperther-mia as an Adjuvant to Radiation Therapy of Recurrent orMetastatic Malignant Melanoma. A Multicentre RandomizedTrial by the European Society for Hyperthermic Oncology,’’Int. J. Hyperthermia 12 (1), 1996, pp. 3–20.

4. S.B. Field and J.W. Hand, eds., An Introduction to the PracticalAspects of Clinical Hyperthermia (Taylor & Francis, London,1990).

5. P. Fessenden and J.W. Hand, in Medical Radiology: RadiationTherapy Physics, A.R. Smith, ed. (Springer-Verlag, Berlin,1995), pp. 315–363.

6. M.H. Seegenschmiedt, P. Fessenden, and C.C. Vernon, eds.,Thermoradiotherapy and Thermochemotherapy 1: Biology, Physi-ology, Physics (Springer-Verlag, Berlin, 1995).

7. C.A. Perez, B. Emami, R.J. Myerson, and J.L. Roti Roti, inPrinciples and Practice of Radiation Oncology, C.A. Perez andL.W. Brady, eds. (Lippincott, Philadelphia, 1992), chap. 16.

8. K. Sugimachi, H. Kuwano, H. Ide, T. Toge, M. Saku, and Y.Oshiumi, “Chemotherapy Combined with or without Hyper-thermia for Patients with Oesophageal Carcinoma: A Prospec-tive Randomized Trial,’’ Int. J. Hyperthermia 10 (4), 1994, pp.485–493.

9. J.W. Strohbehn, “Hyperthermia Equipment Evaluation,” Int.J. Hyperthermia 10 (3), 1994, pp. 429–432.

10. A.J. Fenn and G.A. King, “Adaptive Nulling in the Hyperther-mia Treatment of Cancer,” Linc. Lab. J. 5 (2), 1992, pp. 223–240.

11. A.J. Fenn, C.J. Diederich, and P.R. Stauffer, “An AdaptiveFocusing Algorithm for a Microwave Planar Phased-ArrayHyperthermia System,” Linc. Lab. J. 6 (2), 1993, pp. 269–288.

12. W. Gee, S.-W. Lee, N.K. Bong, C.A. Cain, R. Mittra, andR.L. Magin, “Focused Array Hyperthermia Applicator:Theory and Experiment,’’ IEEE Trans. Biomed. Eng. 31 (1),1984, pp. 38–46.

13. C.J. Diederich and P.R. Stauffer, “Pre-Clinical Evaluation of aMicrowave Planar Array Applicator for Superficial Hyperther-mia,’’ Int. J. Hyperthermia 9 (2), 1993, pp. 227–246.

14. P.F. Turner, “Mini-Annular Phased Array for Limb Hyperther-mia,” IEEE Trans. Microw. Theory Tech. 34 (5), 1986, pp. 508–513.

15. E.R. Lee, T.R. Wilsey, P. Tarczy-Hornoch, D.S. Kapp, P.Fessenden, A. Lohrbach, and S.D. Prionas, “Body Conform-able 915 MHz Microstrip Array Applicators for Large SurfaceArea Hyperthermia,’’ IEEE Trans. Biomed. Eng. 39 (5), 1992,pp. 470–483.

16. M.K. Gopal, J.W. Hand, M.L.D. Lumori, S. Alkhairi, K.D.Paulsen, and T.C. Cetas, “Current Sheet Applicator Arrays forSuperficial Hyperthermia of Chestwall Lesions,’’ Int. J. Hyper-thermia 8 (2), 1992, pp. 227–240.

17. P. Wust, H. Fähling, R. Felix, S. Rahman, R.D. Issels, H.Feldmann, G. van Rhoon, and J. van der Zee, “QualityControl of the SIGMA Applicator Using a Lamp Phantom:A Four-Centre Comparison,’’ Int. J. Hyperthermia 11 (6),1995, pp. 755–768.

18. P.F. Turner, T. Schaefermeyer, and T. Saxton, “Future Trends inHeating Technology of Deep-Seated Tumors,” Recent Results inCancer Research 107, 1988, pp. 249–262. Our group modifiedthe software that controls the transmitter RF power and phasein each channel of the array, allowing the adaptive nulling ex-periments to be conducted.

19. B.B. Mittal, V. Sathiaseelan, R.M. Shetty, K.D. Kiel, M.C.Pierce, W. Adelman, and M.H. Marymont, “Regional Hyper-thermia in Patients with Advanced Malignant Tumors: Expe-rience with the BSD 2000 Annular Phased-Array System andSigma 60 Applicator,’’ Endocurietherapy/Hyperthermia Oncol-ogy 10, 1994, pp. 223–236.

20. C. Schneider and J.D.P. van Dijk, “Visualization by a Matrixof Light-Emitting Diodes of Interference Effects from a Radia-tive Four-Applicator Hyperthermia System,’’ Int. J. Hyperther-mia 7 (2), 1991, pp. 355–366.

21. C.J. Schneider, J.D.P. van Dijk, A.A.C. De Leeuw, P. Wust,and W. Baumhoer, “Quality Assurance in Various RadiativeHyperthermia Systems Applying a Phantom with LED Ma-trix,’’ Int. J. Hyperthermia 10 (5), 1994, pp. 733–747.

22. A.J. Fenn and G.A. King, “Experimental Investigation of anAdaptive Feedback Algorithm for Hot Spot Reduction in Ra-dio-Frequency Phased-Array Hyperthermia,’’ IEEE Trans.Biomed. Eng. 43 (3), 1996, pp. 273–280.

23. A.J. Fenn and G.A. King, “Adaptive Radio-Frequency Hyper-thermia Phased-Array System for Improved Cancer Therapy:Phantom Target Measurements,’’ Int. J. Hyperthermia 10 (2),1994, pp. 189–208.

24. Miniature electric-field measurement probes with a diameterequal to 1 mm are commercially available for clinical use. Theseprobes can fit inside standard implantable catheters. See M.Astrahan et al., “Heating Characteristics of a Helical Micro-wave Applicator for Transurethral Hyperthermia of BenignProstatic Hyperplasia,’’ Int. J. Hyperthermia, 7 (1), 1991, pp.141–155.

25. R.A. Monzingo and T.W. Miller, Introduction to Adaptive Ar-rays (Wiley, New York, 1980), chap. 4.

26. With a four-channel phased array system, it is possible to formas many as three independent nulls. To form more than threeindependent nulls it is necessary to utilize an increased numberof transmitter channels.

27. A. Hartov, T.A. Colacchio, J.W. Strohbehn, T.P. Ryan, and P.J.Hoopes, “Performance of an Adaptive MIMO Controller fora Multiple-Element Ultrasound Hyperthermia System,” Int. J.Hyperthermia 9 (4), 1993, pp. 563–579.

28. W.-L. Lin, R.B. Roemer, and K. Hynynen, “Theoretical andExperimental Evaluation of a Temperature Controller forScanned Focused Ultrasound Hyperthermia,’’ Med. Phys. 17(4), 1990, pp. 615–625.



. is an assistant leader in the RFTechnology group. He re-ceived a B.S. degree from theUniversity of Illinois in Chi-cago, and M.S. and Ph.D.degrees from The Ohio StateUniversity, all in electricalengineering. He joined thestaff of Lincoln Laboratory in1981 and was a member of theSpace Radar Technology groupfrom 1982 to 1991, where hisresearch was in phased-arrayantenna design, adaptive-arraynear-field testing, and antennaand radar cross-section mea-surements. Since 1990 he hasconducted research in theapplication of adaptive-nullingtechniques to radio-frequencyhyperthermia treatment. He iscurrently investigating opti-cally controlled phased-arrayantennas for mobile satellitecommunications applications.

In 1990 Alan was a co-recipient of the IEEE Anten-nas & Propagation Society’sH.A. Wheeler ApplicationsPrize Paper Award for “PhasedArray Antenna Calibrationand Pattern Prediction UsingMutual Coupling Measure-ments,” a paper that he co-authored for the IEEE Transac-tions on Antennas and Propa-gation. He is a senior memberof the IEEE, a member of theNorth American Hyperther-mia Society, and an associatemember of the AmericanSociety for Therapeutic Radi-ology and Oncology. He hasalso been appointed to a five-year term of membership inthe Institute for Systems andComponents of the Electro-magnetics Academy.

received a B.Sc. degree inelectronics engineering fromthe University of Sri Lanka in1976 with First Class honors,and the Ph.D. degree in mi-crowave engineering from theUniversity of Bradford in1982. In 1975 he received theUNESCO Team Gold Medalas the best engineering stu-dent, and the Dr. O.P. Kul-shresha Silver Medal as thebest engineering studentamong the electrical powerand electronics groups. From1982 to 1986 he was a re-search scientist with the Medi-cal Research Council’s ClinicalOncology and Radiotherapeu-tics Unit in Cambridge, En-gland. He was an assistantprofessor in the WashingtonUniversity School of Medicinefrom 1986 to 1988. He iscurrently employed by North-western Memorial Hospital asa hyperthermia engineer in theRadiation Oncology Center.He also holds an appointmentas assistant professor in clinicalradiology at NorthwesternUniversity Medical School. Heis a member of several profes-sional societies, including theInstitution of Electrical andElectronics Engineers, theAmerican Association ofPhysicists in Medicine, theNorth American Hyperther-mia Society, and the AmericanSociety for Therapeutic Radi-ology and Oncology. He is theauthor or coauthor of thirty-five articles in his field, and hehas presented twenty-fivepapers at national and interna-tional society meetings.

. studied electrical engineeringand biology at the RensselaerPolytechnic Institute, and wasa research fellow at the StateUniversity of New York(SUNY) Upstate MedicalCenter in Syracuse, New York,where he received an M.D.degree in 1965. He completedhis internship at St. Joseph’sHospital in Syracuse and hisresidency at the StanfordUniversity Medical Center inPalo Alto, California. From1969 to 1971, he was the chiefof radiation therapy serviceand director of residencytraining for radiation therapyat the Fitzsimmons GeneralHospital in Denver. In 1971,he returned to Syracuse, wherehe has been an attendingradiation oncologist at theUniversity Hospital, SUNYHealth Science Center; theVeterans AdministrationHospital; and the CrouseIrving Memorial Hospital. Heis now the Associate Directorof the Division of RadiationOncology, University Hospi-tal, SUNY Health ScienceCenter in Syracuse, and theDirector of the hospital’sHyperthermia Cancer Treat-ment and Research Center. Heis also a professor in the SUNYHealth Science Center’s De-partment of Radiology in theCollege of Medicine. Anassociate editor of MedicalPhysics, he is the author orcoauthor of more than thirtyarticles in his field and haspresented nearly fifty papers atnational and internationalsociety meetings.

. received a B.A. degree inphysics from the College ofWooster in 1975 and an M.S.degree in electrical engineeringfrom the University of Arizonain 1979. Additional specializa-tion in clinical engineering ledto a position as research associ-ate in the Division of Radia-tion Oncology at the Univer-sity of Arizona from 1979 to1983. He received boardcertification in Clinical Engi-neering in 1983, and boardcerftification in Medical Phys-ics in 1991. He is currently anassociate professor in theDepartment of RadiationOncology at the University ofCalifornia, San Francisco,where his focus of researchcontinues to be the engineer-ing development and testing ofimproved RF, microwave, andultrasound technologies forhyperthermia and thermaltherapy applications. He is amember of the North Ameri-can Hyperthermia Society,IEEE, Association of theAdvancement of MedicalInstrumentation, and theAmerican Association ofPhysicists in Medicine. He haspublished over seventy papers,reports, and book chapters inthe field of hyperthermia, andis associate editor for theInternational Journal of Hyper-thermia and the InternationalJournal of Radiation Oncology,Biology, and Physics.

Improved Localization of Energy Deposition in Adaptive Phased ...

Documents