close

Вход

Забыли?

вход по аккаунту

?

Coaxial microwave antennas for interstitial and intracavitary hyperthermia

код для вставкиСкачать
INFORMATION TO USERS
The most advanced technology has been used to photograph and
reproduce this manuscript from the microfihn master. UMI films the
text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any
type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand corner and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in
reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6" x 9" black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly
to order.
University Microfilms International
A Beil & Howell Information Company
300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA
313/761-4700 800,521-0600
Order Number 9080864
Coaxial microwave antennas for interstitial and intracavitary
hyperthermia
Wong, Terence Zekon, Ph.D.
Dartmouth College, 1990
UMI
SOON.ZeebRd.
Ann Arbor, MI 48106
COAXIAL MICROWAVE ANTENNAS FOR INTERSTITIAL
AND INTRACAVITARY HYPERTHERMIA
A Thesis
Submitted to the Faculty
in partial fulfillment of the requirements for the
degree of
Doctor of Philosophy
by
TERENCE ZEKON WONG
Thayer School of Engineering
Dartmouth College
Hanover, New Hampshire
MAY, 1990
Examining Committee:
Chairman
Member
Member
DeSn of Graduate Studies
© 1990 Trustees of Dartmouth College
TPAIMI.^
Author
Thayer School of Engineering
Dartmouth College
COAXIAL MICROWAVE ANTENNAS FOR INTERSTITIAL
AND INTRACAVITARY HYPERTHERMIA
Terence Z. Wong
Doctor of Philosophy
May, 1990
ABSTRACT
Hyperthermia, the elevation of tissue temperature from 42°C to 50°C, is currently being
investigated as an adjunct to radiation therapy in the treatment of cancer. The interstitial
microwave antenna array hyperthermia (IMAAH) system is designed to be used in
combination with interstitial brachytherapy. Hyperthermia is induced by li \,n.iature coaxial
antennas which are inserted into the 2.2 mm O.D. nylon brachytherapy catheters. This
thesis explores three different aspects of interstitial and intracavitary microwave
hyperthermia. The first investigation is the development of a two-dimensional timedependent numerical model of a non-homogeneously perfused tumor heated with a typical
4-antenna interstitial array. The objective of this study was to determine the conditions
under which temperature changes measured immediately after turning the microwave power
on or off could be used to estimate SAR (W/kg) or blood flow in the tissue. The second
study investigates a choke as a possible improvement to the present interstitial antenna
design and sets the ground work for an intracavitary antenna design. The choke would
address one of the problems inherent in the coaxial microwave antenna, which is that the
performance of the antenna depends on its insertion depth in tissue. Finally, an
intracavitary applicator is described which is based on the present interstitial dipole antenna
design and incorporates a choke. This applicator is designed specifically to deliver
hyperthermia transurethrally to the prostate for treating benign prostatic hyperplasia (BPH),
but similar designs would be applicable to other intracavitary sites.
ii
Ul
ACKNOWIFnaFMRNTS
It has been a privilege to work with the members of my Thesis Committee. Their job
was made much more difficult by the fact that this research was frequently interrupted by
my medical school obligations. I am particularly indebted to the Committee Chairman,
Professor Stuart Trembly, who allowed me to work independently but provided the
requisite guidance and encouragement when it was necessary. He was also instrumental in
the development of the theoretical description of the antennas in this work. Professor John
Strohbehn was a major guiding force during the earlier part of this thesis work, and
continued to provide insightful suggestions throughout the research. I would like to thank
Dr. Evan Douple for his support throughout this effort; his enthusiasm has always been a
driving force behind the hyperthermia project. Finally, Dr. Chris Coughlin has continually
provided useful suggestions and encouragement from both the medical education and
clinical engineering standpoints.
The experimental work in the dog prostate could not have been accomplished without
the participation of Drs. Eirik Jonsson and Jack Hoopes. I am grateful to them for their
contributions and their camaraderie. Thomas Ryan helped with the automated thermal
mapping system and was always available for elaborating new ideas. Drs. John Heaney
and James Taylor provided valuable input for the development of the clinical protocol for
the treatment of BPH.
Finally, I wish to thank my wife, Ellen, for her encouragement, patience, and
understanding throughout this endeavor, and for her technical help in completing the final
manuscript. My parents also deserve credit for their financial assistance and moral support
during this long educational process.
This work was funded in part from the following sources:
Predoctoral Fellowship, Friends of the Norris Cotton Cancer Center, Hanover, NH.
NIH/NCI grants CA23594, CA37245, and CA19379.
ACS grant IN157D.
iv
TABLE OF CONTENTS
1.0
1.1
1.2
1.3
1.4
INTRODUCTION
HYPERTHERMIA
INTERSTITIAL MICROWAVE HYPERTHERMIA AT DARTMOUTH
INTRACAVITARY HYPERTHERMIA
INTRACAVITARY HYPERTHERMIA OF THE PROSTATE FOR BPH...
Benign prostatic hyperplasia
Intracavitary applicators for hyperthermia of the prostate.
Hyperthermia for BPH
1.5 GOALS
2.0 TRANSIENT FINITE ELEMENT ANALYSIS OF THERMAL METHODS
USED TO ESTIMATE SAR AND BLOOD FLOW IN HOMOGENEOUSLY
AND NON-HOMOGENEOUSLY PERFUSED TUMOR MODELS
2.1 ABSTRACT
2.2 INTRODUCTION
2.3 MATERIALS AND METHODS
Bioheat transfer equation
Tumor model
Finite element model
Estimates from transient calculations
2.4 RESULTS
Homogeneous perfusion
Inhomogeneous perfusion
2.5 DISCUSSION
2.6 CONCLUSIONS
1
2
3
4
8
8
9
12
15
17
18
19
21
21
.23
.25
.28
.31
.31
.35
.42
.50
3.0 COAXIAL MICROWAVE ANTENNAS WITH CHOKE FOR
HYPERTHERMIA
.52
3.1 ABSTRACT
3.2 INTRODUCTION
3.3 MATERIALS AND METHODS
Experimental antennas
Impedance measurements
.53
.54
.57
.57
.59
Theoretical calculations
3.4 RESULTS
.60
.64
V
Impedance measurements
Theoretical results
3.5 DISCUSSION
3.6 CONCLUSIONS
64
70
82
87
4.0 A COAXIAL MICROWAVE APPLICATOR FOR TRANSURETHRAL
HYPERTHERMIA OF THE PROSTATE
4.1 ABSTRACT
4.2 INTRODUCTION
4.3 MATERIALS AND METHODS
Theoretical Considerations
89
90
91
92
92
Transurethral microwave antenna
Impedance measurements
SAR measurements
In-vivo experiments
4.4 RESULTS
92
95
95
96
Impedance measurements
SAR measurements
In vivo experiments
4.5 DISCUSSION
4.6 CONCLUSIONS
5.0 CONCLUSIONS
Transient temperature measurements
Investigation of a Microwave Choke
Design of a Transurethral Applicator
Future Directions
,99
,99
,101
.106
.111
.117
.119
.120
.121
.121
.122
6.0 REFERENCES
.124
7.0 APPENDICES
.131
.132
.133
Antenna RG58-2
Antenna RG174-2
Antenna RG174-3
Antenna RG174-4
Antenna RG178-12
.134
.135
.136
Antenna F12-2
.137
»•
vi
HP software for HP vector voltmeter
Description of Software for calculation of Z
Steps for calculating the junction impedance (Zjct) of a choked antenna
Sample session
Fortran code - theoretical impedance software
SAR Measurements on the transurethral microwave antenna
138
149
150
151
154
177
Summaiy of dog prostate hyperthermia (10/9/89)
Summary of dog prostate hyperthermia (3/9/90)
Summary of dog prostate hyperthermia (3/26/90)
Hyperthermia protocol for the treatment of benign prostatic hyperplasia
192
200
204
210
Patient Consent Form
230
vii
LIST OF TABLES
Table 1.1 Microwave intracavitary applicator designs
Table 1.2 Microwave intracavitary applicator designs for hyperthermia of the
prostate
Table 1.3 Hyperthermia for benign prostatic hyperplasia - Summary of clinical
data
5
10
15
Table 2.1 Blood flow values used in homogeneous and non-homogeneous tumor
models
25
Table 2.2 Finite difference time step intervals used for calculating transient tem­
peratures
Table 3.1 Antenna parameters
Table 3.2 Parameters used for theoretical calculation of impedances
28
58
63
viii
LIST OF FIGURES
Figure 2.1
Three models of tumor blood flow, as described by Rubin and
Figure 2.2
Figure 2.3
Casarett (1966)
Two-dimensional finite element grid used in the computer simulations
Typical heat-up temperatures based on transient finite element
24
27
30
Figure 2.5
calculations
Percent error in SAR estimation based on heat-up temperatures for
homogeneously perfused tissue
Blood flow estimates based on cool-down temperatures for
34
Figure 2.6
homogeneously perfused tissue
Percent error in SAR estimation based on heat-up temperatures for
nonhomogeneously perfused tissue
Blood flow estimates based on cool-down temperatures for
nonhomogeneously perfused tissue
Transient temperature data for an extremely nonhomogeneous test
case
36
Figure 2.4
Figure 2.7
Figure 2.8
Cool-down temperatures at two locations in the extreme nonhomoge­
neous test case
Figure 2.10 The effect of using an early time window for estimating blood flow
Figure 2.11 Comparison between the theoretical SAR and estimated SAR in the
32
37
39
Figure 2.9
region of the antenna
Figure 2.12 Variance of the linear regression used to estimate blood flow is plotted
Figure 3.1
as a function of the per cent error
Conventional coaxial microwave antenna
40
41
44
48
55
Figure 3.2 Microwave antenna with coaxial choke
Figure 3.3 Vector voltmeter system used to measure complex impedance
Figure 3.4(a) Measured impedance, antenna RG58-2 (no choke)
Figure 3.4(b) Measured impedance, antenna RG174-2 (thin choke)
56
60
65
66
Figure 3.4(c) Measured impedance, antenna RG178-12 (thick choke)
Figure 3.5(a),(b) Effect of choke thickness, theory and experiment
Figure 3.5(c) Effect of choke length, experiment
Figure 3.6(a) Comparison of theoretical and experimental junction impedance,
antenna RG-58-2 (no choke)
Figure 3.6(b) Comparison of theoretical and experimental junction impedance,
antenna RG174-2 (thin choke)
67
69
70
73
74
ix
Figure 3.6(c) Comparison of theoretical and experimental junction impedance,
antenna RG178-12 (thick choke)
Figure 3.7 Choke impedance Zc for thin and thick chokes
75
76
Figure 3.8(a) Feedline impedance Zf for thin (RG174-2) choked antenna
Figure 3.8(b) Feedline impedance Zf for thick (RG178-12) choked antenna
77
78
Figure 3.9(a) Theoretical feedline impedance, 433 MHz
Figure 3.9(b) Theoretical feedline impedance, 915 MHz
Figure 3.9(c) Theoretical feedline impedance, 2450 MHz
Figure 3.10 Theoretical performance of antenna with ideal choke dielectric
Figure 4.1 Diagram of transurethral hyperthermia applicator
Figure 4.2 Antenna junction impedance as a function of insertion depth
79
80
81
.85
94
100
Figure 4.3
Figure 4.4
Figure 4.5
Figure 4.6
Liquid crystal temperature distributions
Longitudinal SAR distributions
Radial SAR distributions
Schematic diagram for temperature catheters in dog prostate
103
104
105
108
Figure 4.7
Figure 4.8
In-vivo transurethral temperature distributions in dog
In-vivo temperature distributions in prostate
109
110
1.0 INTRODUCTION
1
2
1.1 HYPERTHERMIA
Hyperthermia, the elevation of tissue temperatures from 42°C to 50°C, is currently
being investigated as an adjunct to radiation therapy for the treatment of cancer. The
biological rationale for using hyperthermia as an adjunct to radiation therapy in the
treatment of cancer is well established. In vitro studies using cultured cells and in vivo
studies in animals demonstrate not only that these elevated temperatures alone can kill cells,
but also that radiation and hyperthermia have a strong synergistic cytotoxic effect (Urano
and Douple, 1988, 1989). The lethal effects of hyperthermia are greatest on cell pop­
ulations which are situated in nutrient-poor, low pH environments which often exist in
tumors. These hyperthermia sensitive cells tend to be more radioresistant due to poor
oxygenation. Hyperthermia is preferentially delivered to these poorly perfused tissues, and
is also particularly effective against late S-phase cells, which also tend to be radioresistant.
Although the radiobiological rationale for hyperthermia is very compelling, the eventual
success or failure of hyperthermia as a viable clinical modality may well depend on the
ability of engineers to design a system to preferentially deliver therapeutic temperatures
(43°C to 50°C) to a well-defined treatment volume while maintaining safe temperatures
(<42°C) in all of the remaining tissues.
This thesis is composed of chapters representing three manuscripts which address
several different aspects of microwave interstitial and intracavitary hyperthermia. In both of
these techniques, microwave antennas are embedded in the tissue to produce hyperthermia.
For interstitial hyperthermia, the antennas are inserted into small plastic catheters implanted
directly into the tumor, for intracavitary hyperthermia, the microwave antenna is inserted
directly into a natural body cavity (i.e. mouth, rectum). The first paper (Chapter 2)
investigates a two-dimensional time-dependent numerical model of a non-homogeneously
perfused tumor heated with a typical 4-antenna interstitial array. The objective of this study
was to determine the conditions under which temperature changes measured immediately
after turning the microwave power on or off could be used to estimate power deposition
3
(specific absorption rate or SAR, W/kg) or blood flow in the tissue (L/kg«s). The second
paper (Chapter 3) investigates a choke (or dielectric sleeve) as a possible improvement to
the present interstitial antenna design and sets the groundwork for an intracavitary antenna
design. The choke would address one of the problems inherent in the coaxial microwave
antenna, which is that the performance of the antenna depends on its insertion depth in
tissue. In the third paper (Chapter 4) an intracavitary applicator is described which is based
on our present interstitial dipole antenna design and incorporates a choke. This antenna was
designed specifically to deliver hyperthermia for treating benign prostatic hyperplasia
(BPH), but similar designs would be applicable to many intracavitary sites. The remainder
of this introduction will provide a background for the development of this intracavitary
applicator.
1.2 INTERSTITIAL MICROWAVE HYPERTHERMIA AT DARTMOUTH
Interstitial brachytherapy is a technique in radiation oncology in which the tumor is
implanted with an array of catheters through which radioactive sources
ribbons) are
inserted to deliver radiation to the tumor. One approach that has been studied extensively at
Dartmouth is the interstitial microwave antenna array hyperthermia (IMAAH) system,
which is designed to be used in combination with interstitial brachytherapy. Using this
system, the tumor is implanted with an array of flexible 2.2 mm O.D. nylon brachytherapy
catheters, which accommodate the
ribbons for the brachytherapy treatment as well as
microwave antennas and/or thermometry for hyperthermia treatments. The reasoning
behind this technique includes: 1) higher radiation doses and temperatures can be
selectively delivered to the tumor because it is directly implanted, 2) the invasiveness of the
system is not a major issue since the catheters are required for thermometry and
brachytherapy, 3) the use of an array of antennas enables several parameters to be adjusted
to control the temperature distribution, and 4) intraoperative placement of these catheters
extends the ability of this technique to selectively deliver hyperthermia to deep-seated
4
tumors. To date, the IMAAH system has been applied to treat superficial (Coughlin et al.,
1983) and deep-seated (Coughlin et al., 1985) tumors, as well as brain tumors (Roberts et
al., 1986) and biliary tract tumors (Coughlin and Wong, 1986).
The initial insulated antenna design work at Dartmouth was primarily empirical
(deSieyes et al., 1981). Subsequently, an insulated antenna theory was developed to
describe the behavior of these antennas (King et al., 1983; Trembly, 1982). Experimental
SAR measurements have shown good agreement with theory (Ryan, 1990; Wong et al.,
1986), and the theoretical formulation has recently been applied in the development of a
computer model for treatment planning (Mechling, 1989).
The current IMAAH system is controlled by an IBM AT computer, and is designed to
drive up to 12 microwave antennas coherently at either 915 MHz or 2450 MHz. Microwave
power is provided by a single-output high-power microwave generator and is divided
among the antennas; in addition, a stand-alone computer-controllable 4-channel 433 MHz
generator is available. Power to each antenna can be adjusted either manually or through
automatic feedback control to maintain the desired temperature distribution. More recently,
efforts are being directed toward improving the temperature distribution of interstitial
microwave antenna arrays by air cooling (Eppert et al., 1990; Trembly et al., 1990) and
manipulation of driving phase (Trembly et al., 1986; Trembly et al., 1990).
1.3 INTRACAVITARY HYPERTHERMIA
One modality which is beginning to receive attention at Dartmouth is intracavitary
hyperthermia, where the applicator delivers treatment through a naturally occurring body
cavity. Treatment sites amenable to this approach include gynecological malignancies of the
vagina, cervix, or uterus, rectal cancers, esophageal cancers, and prostatic disease.
Depending on the site to be treated, the intracavitary applicators are designed to be either
rigid or flexible. Several types of microwave intracavitary applicators have been designed
by other groups and are summarized on the next page in Table 1.1. Some of the more
Table 1.1: Microwave intracavitary applicator designs.
Frequency
£MHzl
O.D.
l2mni
300 - 915
10
14
Flexible
Insulated sleeve dipole
(general design)
433
7.9
17.8
Flexible
Roos (1988)
Biconical elements
(uterus)
680
2450
8
8
7
3.6
Rigid
Rigid
Leybovich et al. (1987)
Coaxial sleeve w/ airway lumen
(trachea)
Not stated
Not stated
5
Helical coil (9-tum)
(general design)
433,915
Investigator
Applitfltpr Pg?'8" / SUg
Li et al. (1984)
Lossy helical outer conductor
(rectum and vagina)
Broschat et al. (1988)
Luk et al. (1984)
Length
1ml
Flex/Rjgid
Flexible
8
21
9.4
Rigid
6
general applicator designs will be discussed here, while the intracavitary applicators
designed specifically to treat the prostate will be discussed later in this chapter.
Li et al. (1984) describe a 1 cm diameter antenna design consisting of a 14 cm length
of RG214AJ coaxial cable modified to have a helically wound wire forming the outer
conductor to produce a lossy transmission line. This applicator was tested at frequencies
ranging from 300 MHz to 915 MHz. The resulting SAR patterns were a function of
frequency and insertion depth. In general, the heating pattern was displaced away from the
antenna tip and at certain frequencies extended up the feedline cable. These investigators
also noted that the radial temperature distribution was improved by covering the antenna
with an acrylic "spacer".
Broschat et al. (1988) investigated an "insulated sleeve dipole" applicator designed to
operate at 433 MHz. This antenna was an asymmetrical design based on an RG178 (1.8
mm O.D.) feedline. One antenna section was formed by the inner conductor of the feedline
(0.25 mm O.D., 9.9 cm length) while the other section was formed by a quarter
wavelength (7.9 cm) of braid folded over the feedline jacket to form a section having a
diameter of 2.2 mm. The entire antenna was thickly insulated with heat shrink layers, an
outer layer of latex tubing, and petroleum jelly in the space between the antenna structure
and the latex tubing. Measured SAR distributions were compared with calculations, based
on insulated antenna theory and showed reasonable agreement, with the largest discrepancy
noted at the antenna junction.
Roos (1988) designed a rigid 8 mm O.D. biconical antenna design for treatment of
uterine cancer. The objective of the biconical design was to extend the heating pattern
longitudinally. Two applicators were tested, one with a total active length of 7.0 cm
designed for operation at 680 MHz, and the second having an active length of 3.6 cm for
operation at 2450 MHz. The heating lengths of the 680 MHz and 2450 MHz antennas were
5.5 cm and 3.3 cm, respectively, for > 60% of maximum SAR. For both frequencies, the
60% SAR distribution was 2 cm in diameter.
7
Leybovich et al. (1987) constructed an interesting applicator which included an airway
lumen for treatment of tracheal tumors. They compared the SAR distribution of a single
large coaxial dipole antenna with that obtained by surrounding an airway tube with four
small interstitial antennas spaced at 90° intervals. In this case, the coaxial antenna consisted
of two thin tubular copper elements which were fed by an eccentric RG178 feedline.
Although the overall diameters of these two designs were presumably similar (dimensions
are not stated), the single large antenna was found to be superior to the multiple antenna
approach in terms of heating pattern uniformity and longitudinal SAR distribution.
Luk et al. (1984) discuss a general design for a rigid helical intracavitary applicator.
By adjusting the pitch angle and number of turns on the helix, the applicator could be
optimized for any desired frequency. Axial SAR patterns showed a periodic distribution
with 2 peaks at 433 MHz and 3 peaks at 915 MHz with much higher SARs measured at
915 MHz.
From a practical standpoint, the intracavitary antenna is easier to construct and
implement than the interstitial antenna, because of its larger size. The larger overall diameter
also allows more complicated antenna designs to be realized; features such as a choke, air
or liquid cooling, thermometry catheters, or extra lumens are more easily incorporated.
From a theoretical standpoint, these larger antennas are not as easily characterized as the
smaller interstitial antennas, because they are not electrically thin in the surrounding tissue.
Some of the investigators listed in Table 1.1 include a dielectric sleeve or choke in their
antenna. Such a design, if properly implemented, could improve the performance of the
antenna by making its radiation characteristics independent of insertion depth. The design
considerations and performance of a practical choke are discussed in Chapter 3. Attention
will now be directed toward the specific application of intracavitary hyperthermia which is
addressed in this thesis, that is the treatment of benign prostatic hyperplasia.
8
1.4 INTRACAVITARY HYPERTHERMIA OF THE PROSTATE FOR BPH
1.4.1 Benign prostatic hyperplasia
Benign prostatic hyperplasia (BPH) is a common disease of older men; more than 90%
of all men develop BPH by the eighth decade of life (3). Pathologically, BPH is a nodular
hyperplasia with islands of growth in the prostatic tissue particularly prominent in the
periurethral region (Robbins et al., 1984). As yet, there is no convincing evidence that
BPH is associated with prostatic cancer. The etiology of BPH is not fully understood, but
contributing factors are thought to include hormonal changes and other effects of aging,
environmental factors, and genetic factors (Geller, 1989). With time, the growth often
becomes severe enough to impair urinary flow. BPH is the most common cause of urinary
obstruction in men, and the majority of men over 60 have clinical symptoms of this disease
(Geller, 1989). In total, 10%-20% of all men will require prostatic surgery at some time in
their lives to relieve obstructive symptoms (Lepor, 1989), and approximately 300,000 of
these operations are performed annually in the U.S.
At the present time, the treatment of choice for symptomatic BPH is surgery, for
which there are several major indications:
1) unacceptable decrement in urinary flow,
2) persistent residual urine,
3) acute urinary retention secondary to obstruction,
4) recurrent urinary tract infections secondary to obstruction,
5) hydronephrosis, and
6) acquired bladder diverticula and/or vesicular calculi.
The fact that many patients suffering from BPH are elderly and may not be candidates for
surgery suggests that nonsurgical alternatives for therapy warrant consideration. Non­
surgical treatments for BPH have included medications (a-adrenergic antagonists, 5-a
9
reductase blockers, hormones) and mechanical dilatation (Castaneda et al., 1987; Lepor,
1989). Recently, there has been evidence that hyperthermia may be a useful modality for
the management of symptomatic BPH (Lindner et al., 1989; Sapozink et al., 1989;
Servadio etal., 1986, 1987; Yerushalmi etal., 1985; Yerushalmi, 1986). Hyperthermia
has also been used experimentally to treat the prostate for other diseases, including cancer
(Servadio and Leib, 1984; Szmigielski et al., 1988; Yerushalmi, 1986; Yerushalmi etal.,
1982, 1986) and chronic nonbacterial prostatitis (Servadio et al., 1986, 1987). For the
purposes of this thesis, however, we will be concerned only with hyperthermia as a
possible treatment for BPH.
1.4.2 Intracavitary applicators for hyperthermia of the prostate
Table 1.2 is a compilation of intracavitary microwave applicators that have been
designed specifically for treating the prostate. The prostate is accessible through two
intracavitary approaches: hyperthermia applicators may be inserted either through the
rectum or through the urethra. The transrectal approach is preferable for treating prostatic
cancers for several reasons: 1) the rectum can accommodate a larger applicator which in
turn can heat a larger volume of tissue, 2) most prostatic cancers lie in the posterior portion
of the prostate and are accessible transrectally, and 3) the urethra offers an easily accessible
central location in the prostate for monitoring temperature. All transrectal applicator designs
feature water cooling to prevent damage to the rectal mucosa.
Although the transrectal hyperthermia applicators are advantageous for treating large
prostatic lesions such as cancer, the transurethral approach has significant advantages for
treating BPH. First, the water-cooled transrectal applicator delivers a maximum temperature
several millimeters beneath the rectal mucosa and in the posterior prostate, whereas the
transurethral microwave applicator delivers maximum temperature periurethrally and
concentrates the hyperthermia around the symptomatic lesion. Secondly, the transurethral
applicator can be easily localized within the prostate using a balloon catheter and/or imaging
Table 1.2: Microwave intracavitary applicator designs for hyperthermia of the prostate.
InygstisatQrs
Annlicator Design
Astrahan etal. (1989)
3 interstitial antennas on catheter
(transurethral)
Petrowicz etal. (1982),
Scheiblich and Petrowicz
(1982)
Cylindrical slot, liquid cooled
(transrectal)
Mendecki et al. (1980)
Eccentric coaxial sleeve
(transrectal)
Clini-Therm Corp.
(1990)
BSD Corp. (1990)
Frequency
IMHzi
O.D.
Imml
Length
(cm)
Not stated
630, 915
Flexible
Rigid
434
"
12
20
915,2450
Not stated
Not stated
3 helical interstitial antennas
in catheter (transurethral)
915
6.1
3.5
Single large helical antenna
in foley catheter (transurethral)
915
Rigid
Flexible
Flexible
11
techniques; the transrectal applicator must be properly "aimed" at the prostatic lesion.
Finally, the transurethral approach is less likely to cause complications resulting from
damage to the rectal mucosa.
Petrovicz and Scheiblich (Scheiblich and Petrovicz, 1982; Petrovicz et al., 1982) have
designed a cylindrical slot antenna which operates at 434 MHz, A balloon is attached to the
rigid antenna which couples the antenna to the rectal mucosa through circulating distilled
water, which also serves as a coolant. This antenna was tested on 35 dogs and found to be
capable of generating temperatures > 42°C in the central region of the prostate.
Mendecki et al. (1980) describe a transrectal applicator designed using a coaxial dipole
antenna inserted eccentrically into a teflon bulb. The antenna is a coaxial sleeve design
similar to that described by Broschat (1988). The eccentric placement of the antenna in the
dielectric bulb was shown to result in a directional antenna. Two antennas were designed
for frequencies of 915 MHz and 2450 MHz. Experiments in the canine prostate showed
that temperatures of 41°C to 42°C could be obtained in the center of the prostate (urethra)
with a surface temperature of 43.1°C to 43.5°C at the rectal wall. This was an interesting
result, in view of the fact that no cooling was used in the rectal applicator.
Due to the growing interest in using hyperthermia for treatment of BPH, several
transurethral applicators have been developed. Astrahan et al. (1989a, 1989b) have
developed a transurethral microwave applicator consisting of a modified Foley catheter to
which three microwave antenna catheters are attached at 120° intervals. This apparatus
permits hyperthermia to be delivered to the prostate using standard interstitial microwave
antennas in an arrangement similar to the 4-antenna system described by Leybovich et al.
(1987). One disadvantage of the small interstitial antennas is that they tend to heat proximal
to the junction along their feedlines; this means that parts of the urethra distal to the
prostatic urethra may be heated.
At the most recent North American Hyperthermia Group Annual Meeting in April,
1990, two commercial vendors were in the process of developing dedicated systems for
12
transurethral hyperthermia. Clini-Therm has developed a catheter similar to the Astrahan 3antenna design, while BSD has embedded a helical antenna into the wall of a Foley
catheter.
In Chapter 4, the design of a Foley-based coaxial dipole is described. This design
should offer several advantages over the other transurethral antennas. First, the results of
Leybovich et al. (1987) suggest that the single large coaxial dipole delivers a superior SAR
distribution compared to smaller antennas located along the catheter. Second, based on
experience with interstitial dipole and helical antennas (Ryan, 1990), the SAR patterns of
the single coaxial dipole and helical antennas may be similar, but the dipole antenna would
be expected to perform much better in arrays. This would enable hyperthermia to be
delivered to the posterior region of the prostate (where most cancers occur) by adding a
transrectal antenna.
1.4.3 Hyperthermia for BPH: clinical results
Lindner et al. (1987) treated 6 patients with symptomatic BPH requiring an indwelling
catheter. Previous attempts to wean these patients from their catheters had failed (following
a 5-day course of phenoxybenzamine). The patients were given 5-10 hyperthermia
treatments (1-2 treatments per week) using a microwave (915 MHz) water-cooled
transrectal applicator, and were followed for 6 months. Five of the six patients showed
subjective and objective improvement of their symptoms, and were able to be relieved of
their indwelling catheters. The only treatment failure was a patient with a large, tender
prostate which could not be treated effectively. No complications were observed in this
study.
The largest series of patients undergoing prostate hyperthermia is currently under
study by two investigators in Israel. To date, Servadio and Yerushalmi have used a watercooled transrectal microwave applicator to deliver 500 hyperthermia treatments to 74
patients with benign and malignant prostatic diseases (Servadio et al., 1987). Yerushalmi et
13
al. (1985) summarize their results for the treatment of BPH. In this study, 29 patients with
severe symptoms and for whom surgery was contraindicated were given an average of 14
hyperthermia treatments on a twice weekly schedule. Eleven of the patients had chronic
indwelling catheters. The patients were followed for up to 26 months and evaluated using a
scoring scale for symptoms of frequency, nocturia, urgency, and hesitancy. The authors
noted that symptoms markedly improved after 6-8 treatments, but they felt that a total of
12-15 treatments was optimum. All of the patients who did not have an indwelling catheter
showed symptomatic improvement, and eight of eleven (73%) patients who had indwelling
catheters resumed normal voiding, with a post-void residual of <60 ml. At 18 months
follow-up, none of the previously catheterized patients had developed recurrence of severe
obstructive symptoms or urinary retention. No side effects were noted in this study, and
the transrectal hyperthermia was found to cause no damage to the rectal mucosa.
Most recently, Astrahan and Sapozink et al. have used a transurethral microwave
hyperthermia applicator (Astrahan et al. 1989a,b; Sapozink et al., 1989) for the treatment of
BPH. They have treated 21 men with a total of 177 hyperthermia treatments and a mean
follow-up of 10 months. In terms of objective response (residual urine volume and urine
flow rate), 17/21 (81%) of the patients showed improvement. In terms of subjective
parameters (frequency and stream force), 15/21 (71%) of the patients showed
improvement.
Table 1.3 summarizes the results for hyperthermia as a treatment for BPH. Overall, the
clinical studies to date support the following conclusions:
1) Hyperthermia alone is effective in managing the symptoms of BPH in terms of
both subjecuve and objective parameters. Overall response rate of current studies
is 84%.
2) Hyperthermia may be useful in patients requiring an indwelling catheter, as 76%
of these patients were able to have their catheters removed after treatment.
14
3) Twice weekly treatments with a total of 10-14 hyperthermia sessions has
resulted in the above response rates.
4) Complications and toxicities have been minimal.
Table 1.3.
Hyperthermia for Benign Prostatic Hyperplasia - Summary of clinical data
Lindner et al.
Yerushalmi / Servadio et al.
Sapozink et al.
# Patients
# Treatments
6
47
29
340
21
177
Applicator
Frequency (MHz)
Rectal, water-cooled
915
Rectal, water-cooled
2450
Transurethral
635,915
Treatment length
# treatments / patient
60 min.
5-10
45 - 50 min.
7-18
(?)
8.4 (avg.)
Temperature
Thermal dose equivalent
39.4° - 45.2°C
131 - 510 (total)
42°-43°C
14 - 50 (per treatment)
44°-46°C
(?)
Follow up period
6 mo.
2 - 2 6 mo., (9.7 avg.)
10 mo. (avg.)
Response (subjective)
Response (objective)
Catheter removed
5/6 (83%)
5/6 (83%)
5/6 (83%)
26/29 (90%)
15/21 (71%)
17/21 (81%)
Overall response rate:
Pts. with catheter:
(47/56)
(13/17)
8/11 (73%)
84%
76%
LA
16
1.5 GOALS
The goal of this thesis is to investigate several specific aspects of the coaxial microwave
antenna as they relate to hyperthermia. Specific objectives are listed below:
1) One of the most important factors determining the temperature distribution in the
tumor and normal tissue is blood flow; yet this parameter is unknown and may be
different for each treatment. Our group, as well as others, have observed the
thermal decay after microwave power is turned off and estimated the local blood
flow from the resulting time constant. The goal of this investigation was to
determine the conditions under which this technique could be used to accurately
estimate blood flow, and is the subject of Chapter 2.
2) A fundamental problem with the present interstitial microwave system is that the
antennas radiate differently at different insertion depths. The objective of this study
was to investigate a practical microwave choke design, which could enable the
antenna to perform ideally at all insertion depths. This investigation is discussed in
Chapter 3, and is applicable to both interstitial and intracavitary antenna designs.
3) The third goal of this research was to design a transurethral hyperthermia applicator
for treatment for symptomatic BPH. This antenna design is described in Chapter 4
and can form the basis of other intracavitary applicators.
2.0 TRANSIENT FINITE ELEMENT
ANALYSIS OF THERMAL METHODS USED TO
ESTIMATE SAR AND BLOOD FLOW IN
HOMOGENEOUSLY AND NONHOMOGENEOUSLY PERFUSED TUMOR
MODELS
International Journal of Hyperthermia 4(6): 571-592 (1988)
17
18
2.1
ABSTRACT
A two-dimensional time-dependent finite element model was developed to evaluate
thermal techniques for estimating blood flow and specific absorption rate (SAR). In these
computer simulations, homogeneously and nonhomogeneously perfused tumor models
were heated by a 915 MHz interstitial microwave antenna array. Representative blood flow
values were assigned within the tumor, and the applied SAR distribution was based on a
previously developed antenna theory. SAR values were estimated from the power-on
transient temperatures, and blood flow values were estimated from thermal clearance data
after power was discontinued. These estimated parameters were then compared to the
known "true" blood flow and SAR values throughout the treatment region. SAR values
could be predicted with reasonable accuracy throughout most of the heated region
independent of local blood flow. For a homogeneous model, thermal clearance was found
to yield reasonably accurate blood flow estimates at high perfusion rates and less accurate
estimates at lower perfusion rates. However, for the inhomogeneous model, the blood
perfusion estimates were generally poor, and an average blood flow value for the tumor
was obtained with little ability to resolve the differences in perfusion between regions.
Using temperatures observed early in the cool-down curve resulted in improved spatial
resolution, but increased the contribution of thermal conduction to the blood flow
estimates. A single time-constant exponential thermal decay curve was found to be a
necessary but not sufficient condition for reliable blood flow estimates using this technique.
19
2.2
INTRODUCTION
One of the major problems in clinical hyperthermia is the production of a uniform
temperature distribution in an inherently inhomogeneous tissue volume. Presently available
thermometric techniques permit clinical temperature measurements to be made at only a
limited number of points. Computer simulations are being used at several institutions in an
effort to predict more complete temperature distributions and to evaluate various
hyperthermia treatment modalities (Brezovich et al., 1983,1984; Mechling and Strohbehn,
1986; Paulsen et al., 1984; Sathiaseelan et al., 1986; Strohbehn et al., 1982; Strohbehn and
Mechling, 1986; Strohbehn and Roemer, 1984), Other workers have suggested the
possibility of an interactive model in which the clinically measured temperatures are directly
incorporated into the computer model to produce estimated complete temperature
distributions (Clegg et al., 1985; Divrik et al., 1984), Clegg et al. (1985) report on using
transient temperature data (temperatures measured over time during power-on and poweroff sequences) to obtain additional information about the complete thermal distribution from
the measured data, Many clinics are using thermal washout data to estimate blood flow at
various points in the heated tissue. This technique has been described in detail and
evaluated by several authors (Lyons et al., 1989; Milligan et al., 1983; Milligan and
Panjehpour, 1985; Roemer era/,, 1985; Samulski ef a/., 1989; Sandhu, 1986; Waterman gf
al., 1987). In addition to thermal clearance data acquired at the power-off time period,
temperature changes observed immediately after power is turned on has been used to
estimate power deposition or specific absorption rate (SAR, W/kg) (Roemer et al., 1985).
In this study, a two-dimensional computer model was developed to simulate the
transient thermal behavior of homogeneous and inhomogeneous tumors heated by an
interstitial microwave antenna array hyperthermia (IMAAH) system. The temperature rises
calculated immediately after power was applied were used to estimate SAR at each point in
the treatment region. After steady state temperatures were reached, power was discontinued
and thermal decay data were used to estimate blood flow. This procedure enabled both
20
SAR and blood flow estimates based on these transient temperatures to be compared to
their respective known "true" values throughout the treatment region. The objective of this
work was to model the transient behavior of the IMAAH system and to determine the
circumstances, if any, that would be amenable to SAR and blood flow estimation.
21
2.3 MATERIALS AND METHODS
Bioheat tranter equation
The blood flow and SAR estimates mentioned above and used in this study are based
on the bioheat transfer equation (Bowman, 1982):
Pt Ct (3T/at) = k V^T - pt Pb Cy m (T-Ty) + QH ,
where,
(1)
Pt
= tissue density (kg/m^)
C[
= tissue specific heat (W*s/kg*°C)
k
= thermal conductivity of tissue (W/m*°C)
Pb = blood density (kg/m^)
Cb
= blood specific heat (W*s/kg*°C)
m(x,y) = volumetric blood perfusion rate (m^/kg's)
Tb(t,x,y)
= temperature of blood entering region (°C)
QH (T,x,y)
= power absorbed per unit volume of tissue (W/m^),
= olEP /2 in this study.
T(t,x,y) is taken to be the relative temperature (°C) above core body temperature.
Therefore, we assume that Tb(t,x,y) is uniformly zero throughout the region. The
contribution of metabolic heat production is assumed to be insignificant. Before power is
applied, the treatment region is assumed to be thermally homogeneous and at baseline
(normal core) temperature; thus V^T=0 at t = 0. Since the blood is also at baseline
temperature, the blood flow term also does not make any contribution at this time.
Therefore, it seems reasonable to assume that the resulting simplification of the bioheat
equation applies in the time period preceding and immediately following power turn-on
(time t = ton):
Pt Ct (9T/8t) lt=t%n - QH •
(^)
22
The above assumption is often used to estimate SAR from temperatures measured
immediately after power
is applied at time t = t^n.
SAR = Qh/Pi
(3)
Note that the underlying assumptions are true at time t < t^^, but become less valid as the
temperature rises above baseline, because V^T and the perfusion term both increase with
time.
In a similar manner, thermal washout data have been used in an effort to estimate blood
flow. In this case, the applied power and hence Qjj, goes to zero at time t = t^ff. Three
assumptions made at this point are that (1) the perfusion term overwhelms t he conduction
term (p[pym(T-Ty) » I kV^T I), (2) the blood entering the measurement region is at body
core temperature (Ty = 0), and (3) blood perfusion is not a function of temperature. With
these assumptions, the temperature falls exponentially from its initial value (Tg) after power
is turned off:
or
T(t) = To exp( -pyCymt / c^),
(4)
m = -A, (cj pyCy),
(5)
where t is the time after power-off and X represents the slope of the curve In(T(t)/To) over
time. The time constant of this thermal decay is often used clinically to estimate blood flow.
One of the challenges of using thermal clearance to estimate blood flow is to either 1)
determine a period of time during cool-down when the conduction term is insignificant, or
2) devise a method to "subtract out" the conduction component. However, the conduction
term may increase or decrease over time. If the initial thermal gradients are large, the
conduction term will tend to decrease over time after power is discontinued as the thermal
distribution becomes more uniform. On the other hand, if the steady state thermal
distribution is uniform but the blood flow is nonhomogeneous, the conduction term may
begin near zero (no thermal gradient) and grow with time after power is stopped as the
inhomogeneous blood flow carries the heat away unequally producing thermal gradients.
23
Tumor model
Although it can be expected that every tumor has its own unique blood flow distribution,
Rubin and Casarett (1966) have described three major tumor perfusion patterns: 1) peripheral
vascularization with penetrating vessels, 2) peripheral vascularization without penetrating
vessels, and 3) an arborial pattern with the main blood supply entering the tumor core and
branching toward the periphery. These perfusion models are illustrated schematically in
Figure 2.1.
In this study, a 4 cm diameter tumor having three separate regions was considered
(Figure 2.2). The tumor consisted of a 1 cm diameter core (region 4) surrounded by a 1.2 cm
thick annular intermediate region (region 3) and a 0.3 cm thick outer shell (region 2). The
outer diameter of the tumor was surrounded by normal tissue (region 1). For the purposes of
this study it was assumed that the four tissue regions had identical physical and dielectric
properties (density, permittivity, permeability, electrical conductivity, and thermal
conductivity) equal to that of skeletal muscle. In all cases, it was assumed throughout the
tissue that Pt = Pb = 1-0 x 10^ kg/m^, Ct = Cy = 3.5 x 10^ W*s/kg*°C, and k = 0.63 W/m*°C
(Guy et al., 1974). By appropriately varying the blood flow parameters in each of the four
regions, each of the three major perfusion patterns described above could be simulated. Table
2.1 summarizes the blood flow values used to simulate homogeneous and inhomogeneous
tumor models.
24
M
(h)
(c)
Figure 2.1. Three models of tumor blood flow, as described by Rubin and Casarett
(1966): (a) peripheral vascularization with penetrating vessels, (b) peripheral
vascularization wUhout penetrating vessels, and (c) central vascularization.
25
Table 2.1. Blood flow values used in homogeneous and non-homogeneous tumor models.
Region 4 is the 1 cm diameter central core of the tumor, and is surrounded by concentric shell
regions 2 and 3 (Figure 2.2). Region 1 is the normal tissue surrounding the tumor.
Case name and description
HL
HM
HH
homogeneous, low blood flow
homogeneous, medium blood flow
homogeneous, high blood flow
N1
N2
N3
N4
Blood flow (ml /lOO g/min) in region
_1_
_2_ _4_
5
40
100
5
40
100
5
40
100
5
40
100
inhomogeneous pattern. Figure 2.1(a)
inhomogeneous pattern. Figure 2.1(b)
inhomogeneous pattern, Figure 2.1(c)
40
40
40
150
150
60
40
0
80
0
0
150
inhomogeneous pattern, extreme case
0
0
0
500
Finite element model
The four-region model described above was discretized to form a two-dimensional finite
element grid. All data entry and subsequent analyses were performed on a Digital Equipment
VAX 11/785 computer. Software previously developed at the Thayer School of Engineering
was used to facilitate entry of the finite element grid. The central portion of the grid is shown
in Figure 2.2; the complete grid extends to 5.5 cm along the x and y axes. The grid consisted
of 885 nodes and 1650 triangular elements and represented a cross section through the center
of the tumor. In this simulation, the tumor was implanted with four microwave antennas in a
2 cm square array.
The four antennas were assumed to lie perpendiculai- to the plane of the grid, and the grid
plane was assumed to pass through the junctions of the antennas. In this plane, the origin
was at the tumor center and antennas were located at (x,y) = (-1,-1), (-1,1), (1,-1), and
(1,1), where x and y are given in cm. The microwave antennas have been described
previously (Coughlin et al, 1983, 1985; King et al., 1983; Roberts et al., 1986; Trembly,
1982,1985) and were assumed to be 1.6 mm in diameter and placed concentrically (with no
air gap) in nylon catheters (2.2 mm O.D.). The four antennas were assumed to be driven at
915 MHz with equal power and in phase. Based on this symmetry, calculations were done in
26
only one quadrant (x,y > 0). We were primarily interested in examining the temperatures
along the x and y axes (which by symmetry give the same data) and along the x=y diagonal;
therefore, nodes were placed directly on these lines to avoid the need for interpolation. The
antennas were assumed to be operating as dipoles, with radiating element lengths h^ = hg =
3.0 cm. The electric field (and hence SAR) distribution from this antenna array was
calculated using the theory developed by Trembly and King for an insulated antenna in an
electrically conductive medium (King et al., 1983; Trembly, 1982,1985).
The bioheat transfer equation was solved using a Galerkin finite element method
(Paulsen et al., 1984; Zienkiewicz, 1971) to determine temperatures at each of the grid
nodes. A Crank-Nicholson finite difference time stepping scheme was used to calculate the
grid temperatures over time. The Crank-Nicholson weighting parameter 0 = 0.7 was used to
guarantee unconditional stability. Time step intervals were variable for computational effi­
ciency, being smaller during periods of rapid temperature changes and larger near equili­
bration temperatures (Table 2.2). Type 1 boundary conditions (T=0) were enforced at the
outer grid boundaries x=5.5 cm and y=5.5 cm (not shown in Figure 2.2). No-flux ((p=0)
type 2 boundary conditions were applied along the positive x and y axes and around the
surface of the antenna catheter. In the two-dimensional formulation, it is implicitly assumed
that no-flux conditions also exist perpendicular to the calculation plane.
Two sets of transient simulations were performed. To evaluate power-on transients,
power was applied beginning at time zero (t=0), and temperatures were calculated for 60 s.
To evaluate power-off transients, power was applied at time zero and input power to the
antennas was continuously adjusted over time to maintain a prescribed temperature at a
specified control node (located on the antenna catheter). This control temperature was set to
8°C in all cases, and was maintained for 30 min so that a steady state temperature distribution
was obtained. At this time, power was discontinued and cool-down temperatures were
calculated over a period of 400 s.
27
In both of these simulation types, temperature data used to estimate SAR or blood flow
were recorded (sampled) at intervals which could be realistically obtained in the clinic (i.e. at
4 to 10 s intervals). However, the time steps used in the actual calculations were generally
smaller to improve temporal resolution. For example, in the power-on simulation actual
calculations were made using a time step of 0.1 s immediately after power-on, but the first
data point used in the SAR estimation occured at 4 s. The time intervals used in the finite
difference time-stepping scheme and the intervals used in the actual data analysis are
summarized in Table 2.2.
Figure 2.2. Two-dimensional finite element grid used in the computer simulations.
Symmetry was assumed, so that calculations were made in one quadrant. Only the central
portion of the grid is shown; the complete grid actually extends to x,y = 5.5 cm. The antenna
(perpendicular to the calculation plane) in this quadrant is represented by the open hole along
the diagonal at x = y = 1 cm. The four regions of the treatment region are delineated by the
concentric dense element groups (0.5, 1.7, and 2.0 cm radii). Blood flow values could be
independently assigned to each of these regions. Data were evaluated along the x and y axes
(same data, but not identical data points) and along the diagonal.
28
Table 2.2. Finite difference time step intervals used for calculating transient tem­
peratures. A limited number of these time points (simulating the clinical situation) were
sampled during the heat-up or cool-down periods for SAR and blood flow estimates (see
text).
Experimental series
from t =(s)
tot = (s)
Cool-down experiments:
(power on at t = 0,
power off at t = 1800 s)
0
100
300
600
1790
1900
2100
100
300
600
1790
1900
2100
2200
1
2
5
10
1
2
5
0.0
1.0
1.0
60.0
0.1
1.0
Heat-up experiments:
(power on at t = 0 s)
time step intervals)
Estimates from transient calculations
Selected temperatures calculated immediately after power-on were used to generate an
estimate of SAR which could then be compared with the known "true" value at each node.
Although in theory it would be desirable to obtain temperatures as rapidly as possible after
applying power, present thermometric techniques (i.e. fiberoptic probes in catheters) have
thermal time constants which allow reliable temperature data to be obtained only every few
seconds. Therefore, temperatures were sampled at 4 s intervals for 40 s after power was
turned on and an 11-point least-squares quadratic fit was used to estimate 9T(t,x,y)/9t at t=0.
This was in turn applied to equations (2) and (3) to estimate SAR(x,y). This parabolic
approximation allows the clinically obtainable temperature data to be fit more accurately, and
has been used for determining SAR in tissue-equivalent phantom materials (Wong et al.,
1985,1986). Figure 2.3 illustrates typical heatup temperatures (data points) along with a line
corresponding to the theoretical initial temperature rise (9T/9t)l[_Q+ and the fitted parabola
used to estimate the SAR at this point. These estimated SAR values were then compared to
the true known values at each point and evaluated in terms of percent error:
( estimated SAR(x,y) - true SAR(x,y) )*100 / (true SAR(x,y)).
29
In a similar fashion, temperatures calculated after power-off were used to estimate blood
flow. In this case, temperatures were expected to decay in an exponential fashion if
conduction was negligble. Analysis was thus based on the natural log of the observed
temperature changes. A linear regression was applied to ln(T(t,x,y)/To(x,y)), where t^ < t <
tg and To(x,y) was the initial temperature before cool-down. Equation (5) was then used to
estimate blood flow. If the thermal decay is not truly exponential, this estimate depends on
the choice of t^ and
A was found that blood flow estimates generally improved as t^ and t2
were increased (due to the fact that the errors attributable to initial thermal gradients decrease
over time). Also, when the IMAAH system is used clinically for thermal cool-down
measurements, the first 60 s or so of data must be ignored, particularly when the
termperature probes are adjacent to an antenna. This is necessary due to the conduction
occurring along the metallic antenna, which results in a large initial temperature change.
However, temperatures measured beyond 5 min are small, and in practice both noise and the
resolution of the thermometry equipment set an upper limit to the time window which can be
used to estimate blood flow. For the purposes of this study, a compromise was made by
using ti=120 s and t2=240 s. This time period was felt to be early enough to be a clinically
practical time window, yet late enough to avoid much of the conduction effects. Cool-down
data were taken at 10 s intervals within this time period, and a linear regression was applied
to the data points to determine k(x,y); from this, the blood flow was estimated using equation
(5). Representative cool down curves (and corresponding blood flow estimates) from points
in regions having high and low thermal gradients are discussed later and illustrated in Figure
2.9. Estimated blood flow values were plotted as a function of distance from the origin along
the x-axis, y-axis, and diagonal to allow comparison with the true values.
30
o
o
Case HH
(x,y) = (0,0)
Q)
(0
£ 21
Initial slope based on
true SAR
Parabola used for
SAR estimate
2
(0
E
Error in SAR estimate = -5.96%
0)
10
20
30
40
50
60
Time after power-on (sec)
Figure 2.3. Typical heat-up temperatures (A) based on transient finite element calculations.
These data were calculated at the center of the tumor in case HH, where the blood flow was
100 ml/100 g/min throughout. The solid line represents the theoretical initial linear
temperature rise based on the true SAR at this point. Also shown is the best-fit parabola
based on 11 of the temperature data points taken at 4 s intervals for 40 s. This parabola,
based on clinically obtainable temperature measurements, was used to make the SAR
estimation. In this example, the estimated SAR based on the observed temperature rise is low
by about 6 per cent.
31
2.4
RESULTS
Homogeneous perfusion
For homogeneous models, blood flows in each of the three tumor regions were set equal
to the surrounding tissue blood flow values. Cases were run to represent low (case HL, 5
ml/100 g/min), intermediate (case HM, 40 ml/100 g/min), and high (case HH, 100 ml/100
g/min) blood flow conditions. These values are consistent with typical blood flows measured
in tumors and normal tissues at hyperthermic temperatures (Pence and Song, 1986).
The results from the SAR estimates are shown in Figure 2.4. In these figures, the SAR
values were estimated from the heat-up data at points traversing the array along the x and y
axes and along the diagonal. The estimated values were compared with the known SARs at
each point and the percent error plotted as a function of distance (0 to 4 cm) from the origin.
The errors range from -2 per cent nominally for the low blood flow case to -5 per cent
nominal error in the high blood flow situation. It is evident that the maximum errors (as large
as 25 per cent) are encountered near the antenna, but are low throughout most of the heated
region. Blood flow and conduction tend to remove heat from points within the antenna array
as the temperature rises, so that the SAR estimates tend to underestimate the true values,
especially at the tumor center. As expected, this underestimation becomes more apparent at
high blood perfusion rates (Figure 2.4(c)). Figure 2.4 suggests that SAR can be determined
to within 10 per cent for homogeneously perfused tissue throughout the region of interest,
except within the area immediately surrounding the antenna catheter.
32
+150
Case HL: Homogeneous Blood Flow (5 mt/IOOg min)
(a)
+100-
I
1
S
c
3
+50Along diagonal
£
I
Along axes
-SO'
DIsUnce from Center (cm)
Case HM; Homogeneous Blood Flow (40 ml/IOOg mIn)
(b)
E
r
<
^ Along diagonal
I
^ Along axes
DIsUncB from Center (cm)
Case HH; Homogeneous Blood Row (100 ml/IOOg mIn)
«
I
C
+100-
+50-
8
Along diagonal
.s
* Along axes
-SODistance (ram Center (cm)
Figure 2.4. Percent error in SAR estimation based on heat-up temperatures as a function of
distance from the tumor center for homogeneously perfused tissue. Data are shown along the
X and y axes (combined) and along the diagonal. Three cases are shown: (a) case HL, low
blood flow = 5 ml/100 g/min, (b) case HM, moderate blood flow = 40 ml/100 g/min, and (c)
case HH, high blood flow = 100 ml/100 g/min.
33
Estimates of blood flow based on cool-down data 120 to 240 s after power was turned
off are also plotted as a function of distance from the origin in Figure 2.5 for homogeneously
perfused tissue. In the low perfusion case (HL), the estimated blood flow ranged from
approximately 2 ml/100 g/min to 12 ml/100 g/min at the tumor center (Figure 2.5(a)).
Predictive accuracy greatly improved in the moderate blood flow case (Figure 2.5(b)), where
blood flow estimates were generally within 20 per cent of the true value. In the highly
perfused tissue model (Figure 2.5(c)), blood flow was generally predictable to within 10 per
cent. These results indicate that this technique can yield reasonable blood flow estimates for
homogeneously perfused tissue, and the accuracy improves as the blood flow increases.
34
Case HL
i
§
i
u.
m
, Along axes
True BF- 5 ml/IOOg min
1
a
Along diagonal
DliUnca from Center (cm)
True BF
40 ml/IOOg min
Along axes
1 30Along diagonal
#
20Case HM
Distance Irom Center (cm)
120
True BF- 100 ml/IOOg min
I 100
^Al^g_a»s^^
80-
Along diagonal
60-
0
1
2
3
4
Oiatance from Center (cm)
Figure 2.5. Blood flow estimates based on cool-down temperatures as a function of distance
from the tumor center for homogeneously perfused tissue. Data are shown along the x and y
axes (combined) and along the diagonal. Three cases are shown; (a) case HL, low blood
flow = 5 ml/100 g/min, (b) case HM, moderate blood flow = 40 ml/100 g/min, and (c) case
HH, high blood flow = 100 ml/100 g/min.
35
Inhomogeneous perfusion
The results for SAR estimates from inhomogeneously perfused tumor models are
illustrated in Figure 2.6, where graphs (a), (b), and (c) correspond to the respective
perfusion patterns shown in Figure 2.1 and blood flow cases Nl, N2, and N3 from Table
2.1. These figures suggest that SAR estimates can be made from heat-up temperatures in
nonhomogeneously perfused tissue with accuracy similar to that obtained in the
homogeneously perfused model.
Blood flow estimates made in the inhomogeneously perfused tissue are plotted in Figure
2.7 for these three cases. In general, it is apparent that none of the blood flow
inhomogeneities are accurately predicted by the cool down estimates. Instead, the estimated
blood flow values over the region tend to smear toward an intermediate value. There are
regions in which the blood flow value is predicted with reasonable accuracy, but these
regions tend to have assigned blood flow values similar to the intermediate value mentioned
above. This suggests that blood flow estimates from thermal cool down data cannot be ex­
pected to yield accurate predictions for local blood flow in nonhomogeneous tissues, even if
blood flow rates are very high.
36
CasaNI; Blood Row (ml/IOOg min)
Along diagonal
DMuios Irom Center (cm)
+150CasaN2: Blood Row (ml/IOOg inin)
0
Hisol-"
4
+100-
*.
S
I8
+50-
<
c
§
a
•50DlBtanoe from Center (cm)
+150"
Case N3: Blood Row (ml/IOOg min)
— 80
&
I
I®
g
s
I
H 60
4
+50Along diagonal
0":
Along
-50 •
Distance from Center (cm)
Figure 2.6. Percent error in SAR estimation based on heat-up temperatures as a function of
distance from the tumor center for nonhomogeneously perfused tissue. Data are shown along
the X and y axes (combined) and along the diagonal. The three cases shown here: (a) case
Nl, (b) case N2, and (c) case N3 correspond to the three tumor models ((a), (b), and (c),
respectively) illustrated in Figure 2.1.
37
1S0
CassNI:
True Blood Row.
E
100-
Esl. BF along axes.
50-
Est. BF along diagonal
DIstancs (rem Center (cm)
150-1
Case N2:
True Blood Flow.S
E
8
2
E
I
50-
EbI. BF along axes.
(•
Est BF along diagonal
DIstanoa from Center (cm)
Est. BF along axes.
Est. BF along diagonal
Distance Irom Center (cm)
Figure 2.7. Blood flow estimates based on cool-down temperatures as a function of distance
from the tumor center for nonhomogeneously perfused tissue. Data are shown along the x
and y axes (combined) and along the diagonal. The three cases shown here: (a) case Nl, (b)
case N2, and (c) case N3 correspond to the three tumor models ((a), (b), and (c),
respectively) illustrated in Figure 2.1.
38
In order to further evaluate the SAR and blood flow estimation techniques, an extreme
nonhomogeneous case was constructed (case N4) consisting of a highly perfused 1 cm
diameter tissue core (blood flow = 500 ml/100 g/min) surrounded by unperfused tissue
(blood flow = 0). The results of this simulation are presented in Figure 2.8, Figure 2.8(a)
shows the steady state temperature distribution just prior to the cool down measurements.
Figure 2.8(b) shows the errors in estimates of SAR, and Figure 2.8(c) is a plot of estimated
blood flow. The extremely high blood flow in region 4 of the model is enough to cause a no­
ticeable error (-37 per cent) in SAR prediction in the center of the tumor (Figure 2.8(b)), but
the estimated blood flow curve (Figure 2.8(c)) is smooth and does not reveal the presence of
this highly perfused region.
39
10
Blood Flow
500
0 ml/IOOgmIn
CaseN4;
Steady-slate
Tempeiature
8
6
Along
4
Along diagonal
2
0
DIslwice from Center (cm)
Case N4: Error In SAR estimations
Along diagonal
DIstanco from Center (cm)
Case N4: Estimated Blood Flow
Along axes
Along diagonal
1
2
Diatanco from Center (cm)
Figure 2.8. Data (along the x and y axes and along the diagonal) plotted as a function of
distance from the tumor center for the extremely nonhomogeneous test case N4: (a) steady
state temperature, (b) error in SAR estimate, and (c) estimated blood flow. The central 0.5
cm radius (region 4) of this model had a blood flow of 500 ml/100 g/min and the entire
surrounding area (regions 1, 2, and 3) was assigned a blood flow of zero.
40
0.8-
Case N4: Node adjacent to antenna
True BF = 0 ml/IOOg min
Slope for blood flow estimation
0.6
&
0.4-
Calculation
Window
0.2-
Est. BF = 9.06 ml/100g mm
400
Time After Power-off (sec)
2.0
Case N4: Node in tumor center
True BF = 500 ml/1OOg min
Slope for blood flow estimatioi
|T
o
Calculation
Window
5
0.5-
Est. BF = 11.8 ml/1OOg min
0
100
200
300
400
Time After Power-off (sec)
Figure 2.9. Cool-down temperatures (log scale) at two locations (nodes) in the extreme
nonhomogeneous test case N4: (a) adjacent to the antenna (blood flow = 0), and (b) center of
the tumor (blood flow = 500 ml/100 ^min). Blood flow estimates were based on the slope of
these curves within the specified time window (120 to 240 s, inclusive). The temperature
data points used for these estimates are indicated by (+). When the temperature decay does
not fit a single exponential, the estimated blood flow is a function of the chosen calculation
window.
41
60
Case HM
Antenna
True BF
40
40 ml/IOOg min
Along axes
20
tA A,
—^
Along diagonal
0
0
1
2
4
3
Distance from Center (cm)
I
Case N4
True BF (ml/IOOg min) -
500
60-
S
s
o
E
5
40-
i
1
"
20-
111
0
1
2
3
4
Distance from Center (cm)
Figure 2.10. The effect of using an early time window (0 to 60 s) for calculating estimated
blood flow in a homogeneously perfused tissue (a), and in an extremely nonhomogeneously
perfused tumor (b). These plots differ from Figures 2.5(b) and 2.8(c), respectively, only in
that the earlier calculation time window was used.
42
2.5
DISCUSSION
The techniques of using heat-up temperature transients to estimate SAR distributions
and using cool-down temperatures to estimate blood flow are well established and convenient
to employ clinically. When the tissue is at a steady state temperature distribution, the
conduction term is relatively constant and changes in regional blood flow are reflected by
changes in the input power required to maintain the steady state temperatures. Roemer et al.
(1985) have made a detailed comparison between this steady state technique of estimating
blood flow with the well known cool down technique. Using experiments on dog thigh and a
one-dimensional inhomogeneous computer model, they concluded that the time-dependent
conduction effects make a significant contribution to the cool-down temperatures. They point
out two major consequences of conduction: (1) it interferes with the ability of thermal
clearance techniques to measure blood flow, and (2) it causes discrepancies between the
steady state and transient methods of estimating perfusion. They describe the thermal decay
in terms of a single parameter Wg, which combines the effects of conduction and blood flow.
Our results reinforce these findings, and further illustrates how conduction severely limits the
ability of thermal clearance methods to resolve blood flow heterogenieties. The validity of
thermal clearance methods for estimating blood flow has also been questioned by Sandhu
(1986), who based his findings on a homogeneous spherical tumor model. He concluded that
thermal decay is representative of blood flow in only very limited situations. Our results
indicate that cool-down measurements may be representative of true blood flow in homoge­
neously perfused tissue if the blood flow is sufficiently high. However, we demonstrate that
tissue inhomogeneities significantly affect the accuracy of the estimations.
In this simulation, SAR values were predicted using heat-up temperatures with
reasonable accuracy throughout most of the tumor volume in both homogeneous and
nonhomogeneously perfused tumors, provided that measurements were made approximately
0.5 cm from the antennas. As blood flow was increased, the error in SAR estimation was
shifted downward (Figure 2.4) as expected. In general, however, this figure shows that
43
blood flow does not have a large effect on SAR estimates. Except for the region near the
antennas, SAR estimates were within 6 percent of the true SAR value for all homogeneous
blood flow cases. Although SAR values were underestimated in regions of high blood flow
in the nonhomogeneous cases, the resulting error did not exceed -10 per cent for blood flow
values up to 150 ml/100 g/min (cases Nl, N2, and N3). The extremely high blood flow (500
ml/100 g/min) in case N4 resulted in an underestimation of -37 per cent at the tumor center.
In all cases, the maximum errors in SAR estimation occurred near the antenna catheter, where
they approached ICX) percent These discrepancies in SAR estimation are more clearly seen in
Figure 2.11(a), which shows the estimated and true SAR values near the antenna catheter.
The differences are probably due, at least in part, to the fact that the SAR estimates based on
temperature changes are not able to track the rapid variations in SAR encountered near the
antennas. Another contributing factor is the fact that the formulation for V^T is not valid at
the surface of the antenna catheter due to the no-flux boundary condition. Therefore, we
conclude that the relatively large errors in SAR predictions near the antennas are due to two
factors: (1) the steep SAR gradients in this region, and (2) contributions from the numerical
methods.
Figure 2.11(b) illustrates a comparison of two methods for estimating SAR in the
extremely nonhomogeneously perfused case N4. In this example, the parabolic curve fit used
in this study is compared to a simple linear regression applied to temperatures obtained at Is
intervals for 4 s after power-on. This graph demonstrates that the SAR estimates could be
further improved if reliable temperatures could be obtained clinically at 1 s intervals.
It should be emphasized that the present model assumes an electrically homogeneous
tissue. In practice, step discontinuities between electrically dissimilar tissues, such as those
occurring between muscle and fat, would cause large gradients in the SAR distribution and
hence adversely affect SAR estimates.
44
1.0-
Antenna
Catheter
True SAR
0.8-
I
Estimated SAR
0-6-
0.2-
Blood Flow = 0
0.8
1.0
1.2
1.4
1.8
1.6
2.0
Distance from origin along diagonal (cm)
+150
Blood Flow
(ml/1OOg min)
500
0
+100-
0>
CO
E
1
+50-
Quadratic (0-40 sec)
cc
. Linear (0-4 sec)
c
2
111
Case N4
-50
0
1
2
3
4
Distance from center along diagonal (cm)
Figure 2.11. (a) A comparison between the theoretical SAR and estimated SAR (based on
heat-up temperatures) in the region of the antenna. No-flux boundary conditions were
assumed at the antenna catheter surface. There is zero blood flow throughout the region in
this example, (b) Comparison of two techniques for estimating SAR from heatup
temperatures in the extremely nonhomogeneously perfused case N4. Comparison is made
between the quadratic fit used in this study (11 points, 4 s intervals for 40 s) and a simple
linear regression (5 points, 1 s intervals for 4 s). Although the linear technique is more
accurate, it is presently not practical in the clinic.
45
The results of this two-dimensional analysis for the homogeneous blood flow situation
show that blood flow can be estimated quite accurately throughout the treatment volume if the
blood flow is high; in this simulation, a blood flow of 100 ml/100 g/min was predicted using
cool-down calculations with an error of less than 10 per cent over most of the region. As
expected, the blood flow estimates were poorer as the true perfusion values were decreased.
At low blood perfusion rates, errors greater than 100 per cent were not unusual. Strohbehn et
al. (1982) define a normalized blood flow rate,
= tq/ty, where tq and ty are time constants
for conduction and blood flow, respectively:
tb = 1 / ( Pb m), and
to =
piCt / k, where Rq is the radius of the antenna array.
If Vjsf < 1, thermal conduction can be expected to dominate, and if
> 1, blood flow should
prevail in determining the thermal transient response. For this model, homogeneous blood
flows of 5, 40, and 100 ml/100 g/min yield values of 0.93, 7.41, and 18.52 for v^,
respectively. When blood flow is 5 ml/100 g/min,
is nearly unity, implying that
conduction contributes about the same as blood flow to the thermal decay. Therefore, we
expect to find errors on the order of 100 per cent when attributing a single thermal decay
constant to blood flow alone (Figure 2.5(a)). As the blood flow is increased,
also
increases, and the contribution of the conduction term becomes less significant.
For the case of an inhomogeneous tumor model, it was found that the actual blood
perfusion pattern could not be resolved using single-point cool-down temperatures. Blood
flow estimates were found to be very smooth throughout the tumor volume even in the
presence of extremely heterogeneous blood flow patterns. Thus, the blood flow estimated by
thermal clearance at a given point in nonhomogeneous tissue may or may not be related to the
actual blood flow at that point. It is evident from Figure 2.7 that the thermal clearance
technique, when applied 2-4 minutes after power-off was unable to detect blood flow
heterogenieties in any of the tumor models, including the extreme case N4. These findings
suggest that blood flow estimates made in nonhomogeneous tissues yield, at best, an average
46
value of tissue cooling and do not necessarily represent true local blood flow values. This can
be explained by the fact that conduction tends to smooth out the thermal distributions, so that
the temperatures do not change sharply with distance in the tissue even in the face of large
step changes in blood flow (Figure 2.8(a)). This smoothing effect is enhanced further after
power is turned off, as the tissue tends to approach a single uniform temperature.
Cool-down temperatures are plotted on a log scale as a function of time in Figure 2.9 at
two points (nodes) in the extremely nonhomogeneous case N4. One point (Figure 2.9(a))
was the control node adjacent to the antenna catheter (x=y=0.922 cm, blood flow = 0) and
the second node (Figure 2.9(b)) was located at the center of the tumor, (x=y=0 cm, blood
flow = 500 ml/100 g/min). The "+" marks identify the data points used in estimating the
blood flow from the cool down temperatures; in this case the usual points between 120 and
240 s after power was turned off were used for analysis. This time interval was chosen to
reduce the contribution of conduction in the blood flow estimates due to initially high thermal
gradients. When the IMAAH system is used clinically, some of the temperature probes are
typically located in the antenna catheter alongside each antenna (Coughlin et al., 1985;
Roberts et al., 1986). In this situation, the early part of the cool-down curve is further
distorted by the adjacent highly conductive antenna. The plots in Figure 2.9 suggest that the
spatial resolution of the blood flow could be improved by using an earlier time interval for the
estimation. Figure 2.10 illustrates estimated blood flow distributions for the homogeneous
case HM (Figure 2.10(a)) and the extremely nonhomogeneous case N4 (Figure 2.10(b))
when blood flow estimates were based on the first 60 s of cool down temperatures after
power was discontinued. Figures 2.10(a) and 2.10(b) are based on the same temperature data
as Figures 2.5(b) and 2.8(c), respectively; the only difference was in the time periods used in
the blood flow estimations. In comparing Figure 2.10(a) with 2.5(b) (homogeneous blood
flow), it can be seen that blood flow estimation is more dependent on the initial local
conduction effects and is less accurate, especially near the antenna, when earlier cool-down
temperatures are used. On the other hand, comparison of Figure 2.10(b) to Figure 2.8(c)
47
shows the improved spatial resolution obtained in the nonhomogeneous case when the earlier
times were used for calculating the blood flow estimates. Nevertheless, the errors are still
extremely large.
According to the bioheat transfer equation, the thermal decay should be described by a
single time constant if conduction is negligible. Therefore, it seems reasonable to use this fact
as a criterion for separating valid from invalid cool-down results. The question is posed: if
the thermal washout temperatures decay exponentially, can accurate blood flow data then be
extracted? In order to answer this question, the variance (square of the standard error,
Beyer, 1980) of the linear regression of the ln(T(t,x,y)/ro(x,y)) versus time curve was used
as a measure of how well the temperature data fit a single time constant exponential decay.
This was compared to the blood flow estimate based on the cool-down temperatures. The
variance of the linear regression (for cool-down temperatures observed from 0 to 400 s after
power-off) is plotted as a function of the per cent error in the blood flow estimate (absolute
value) in Figure 2.12. The data here are from the moderately perfused homogeneous case
(HM), and includes all points within a 4 cm radius of the tumor center (829 nodes). The
correlation coefficient for this data is 0.573, indicating some degree of correlation. This plot
illustrates that there are nodes where blood flow estimation is poor (high per cent error), yet
have low variance. On the other hand, there are no points with high variance where the blood
flow estimate was accurate. This suggests that low variance (i.e. a single time constant) is a
necessary but not sufficient condition for accurate blood flow predictions from cool-down
measurements.
48
40
Case:HM (blood flow = 40 ml/1GOgmin)
Correlation Coefficient = 0.573
20
30
Error I in BF Estimate (%
Figure 2.12. Variance of the linear regression used to estimate blood flow is plotted as a
function of the per cent error (absolute value) in the resulting blood flow estimate. This plot
was generated using all nodes within 4 cm of the tumor center (829 nodes). The tissue had a
homogeneous blood flow of 40 ml/100 g/min (case HM), and all sampled cool-down
temperatures (calculation window of 0 to 400 s after power-off) were used to estimate blood
flow and calculate variance.
49
The simulations performed in this study probably represent best case situations, which
cannot realistically be expected in the hyperthermia clinic. This study assumed that the
bioheat transfer equation was valid and that blood flow did not change with time or
temperature. It was also assumed that temperatures could be measured with great precision
(not limited to 0.1 °C as with most clinical thermometry systems); the results of this study did
not account for discretization error. In this model, it was also known that the cool-down
temperatures would approach the original baseline temperature at t=«». This baseline
temperature is not as clearly defined in clinical and in animal thermal clearance studies (Lyons
et al, 1989; Samulski et al., 1989). Lastly, it was assumed that a two dimensional analysis
would allow valid conclusions to be drawn. Whereas the original proposed tumor models
(Figure 2.1) represent spherical tumors, this model represents the tumor as a cylinder
extending perpendicular to the data plane, with all vasculature removing heat in this direction.
None of these assumptions are valid in the clinic, but we believe that they are generally
consistent with a best case situation, and that removal of any of these assumptions would
result in a more pessimistic outcome.
Finally, it should be noted that this study is based on a model which is very challenging
for transient temperature studies. The 4-antenna interstitial microwave array used here
delivers localized heating to an area approximately 3 cm in diameter. In addition, the tumor
model includes some wide blood flow variations over relatively small distances (as small as 3
mm).
50
2.6
CONCLUSIONS
A theoretical two-dimensional model was used to simulate hyperthermia treatment of a
nonhomogeneously perfused, electrically homogeneous tumor by an interstitial microwave
antenna array. It was found that heat-up temperatures could be used to estimate SAR
throughout most of the heated region in both homogeneous and nonhomogeneous blood flow
cases. It was also found that cool-down measurements generally could not be expected to
yield a reliable estimate of blood flow; these blood flow estimates were reliable only when
made in highly perfused, homogeneous tissue. In general, the temperature decay after poweroff had a time constant that varied with time. If the initial time constant was used to estimate
blood flow, inhomogeneities could be resolved better in the blood flow distribution, but the
blood flow estimates were more sensitive to conduction effects caused by thermal gradients
present in the initial steady-state temperature distribution. If a later time constant was used in
the blood flow estimate, blood flow estimates were less susceptible to errors caused by the
initial thermal gradients, but could not detect blood flow heterogenieties due to subsequent
conductive smoothing of the temperature distribution. The implication of these findings is
that for the inhomogeneously perfused tumor model there is no cool-down time window
which will allow resolution of the thermal decay into separate perfusion and conduction
components.
The results of this study suggest that the simple exponential thermal clearance technique
applied at single points cannot reliably predict true blood flow in realistic clinical situations.
In addition, we present evidence that the reliability of blood flow estimates may not be
accurately predicted by how closely the thermal decay curve fits a single exponential time
constant. Several techniques have been proposed to isolate the blood flow information from
the thermal decay curve (Milligan et al., 1983; Milligan and Panjehpour, 1985; Samulski et
al., 1989). These methods require either empirical assumptions based on experiments or
accurate modeling of the hyperthermia applicator. In either case, assumptions are made about
the contribution of the conduction component. These assumptions are applicator-specific and
51
do not account for modifications in the temperature distribution caused by blood flow.
Assumptions about the conductive component are readily made in homogeneous tissue, but
cannot be made for nonhomogeneously perfused tissue without knowledge of the actual
blood flow pattern. Another possibility which warrants further study is the estimation of V^T
using temperature data adjacent to the actual measurement point. This estimation may be
possible in the clinic using a single mutiple-sensor probe, providing that it is oriented along
the direction of the maximum temperature change.
SAR estimates are usually made when the temperature gradients are relatively small, and
are therefore less susceptible to errors caused by local blood flow and conduction. Thus,
SAR estimates are more accurate over a wide range of blood flow values and are not
influenced greatly by blood flow inhomogeneities. It should be emphasized, however, that
the tumor model used in this study was assumed to be electrically and thermally
homogeneous.
3.0 COAXIAL MICROWAVE ANTENNAS
WITH CHOKE
FOR HYPERTHERMIA
submitted to International Journal of Hyperthermia
52
53
3.1 ABSTRACT
Hyperthermia, the elevation of tissue temperature to 42°C - 50°C, is being studied
clinically as an adjunct to radiation therapy for the treatment of cancer. One heating
technique is the interstitial microwave antenna array hyperthermia system (IMAAH), in
which an array of miniature coaxial microwave antennas are inserted into flexible plastic
catheters which have been implanted into the tumor for interstitial brachytherapy. One of
the major problems of the miniature coaxial dipole antennas used for interstitial
hyperthermia is the fact that the impedance and power deposition pattern of these antennas
changes with insertion depth. In this paper, we examine the coaxial choke as a possible
solution to this problem. The choke was not the ideal quarter-wavelength type, because its
length was determined by the resonance half-length length of the antenna. Several choked
antennas were constructed and compared with the conventional antenna. In order to test the
performance of the antennas, complex impedance was measured at 10 different insertion
depths in muscle-equivalent phantom and the standard deviations of the measurements used
to measure the degree to which impedance varied with insertion depth. A theoretical model
was also developed to describe the performance of the choked antennas, and the predicted
behavior was found to compare favorably with the experimental results. Based on these
theoretical and experimental results, conclusions are made regarding practical aspects of
choke design, construction, and performance.
54
3.2
INTRODUCTION
Hyperthermia, the elevation of tissue temperature to 42°C to 50°C, is currently being
studied as an adjunctive treatment to radiation therapy for the treatment of cancer. One
heating technique that has been developed is the interstitial microwave antenna array
hyperthermia (IMAAH) system. Using this technique, miniature coaxial microwave
antennas are used to deliver hyperthermia through an array of 2.2 mm. O.D. nylon
catheters which have been implanted into the tumor for brachytherapy (Coughlin et al.
1983,1985). Typically, the antennas are spaced 1.5 to 2 cm. apart, and are driven at 433
MHz, 915 MHz, or 2450 MHz. The design and behavior of these miniature coaxial
antennas has been described in detail experimentally (DeSieyes et al., 1981; Jones et al.,
1986; Wong etal., 1986) and theoretically (King etal., 1983; Mechling, 1989; Strohbehn
et al., 1986; Trembly, 1982).
One of the major problems facing this technique is the fact that the performance of the
microwave antennas is a function of insertion depth, an effect which has been observed
experimentally (Chan et al., 1989; Jones et al., 1988) and explored theoretically (James et
al., 1989; Mechling, 1989). A typical situation is shown in Figure 3.1. For optimum
radiation, the antenna should be inserted such that section lengths h^ and hg are equal, each
section corresponding to a quarter wavelength in the composite tissue/catheter medium; in
practice, insertion depth, and hence the length of section B, are determined by the clinical
situation. When insertion depth is not ideal, several problems may be encountered. First,
the antenna performance may be reduced due to increased reflected power at the antenna
junction. This in turn results in increased power requirements and ohmic heating of the
antenna feedline. Clinically, this often results in pain for the patient, particularly at the
entrance site of the catheter. Second, as insertion depth increases the resistance seen along
section B increases, causing it to radiate preferentially and displacing the heating pattern
away from the antenna tip. This may result in the unintended heating of normal tissues and
subtherapeutic temperatures in the tumor.
55
Coaxial Microwave Antenna
feedline
— hg
1 r—
D,
(insertion depth)
Figure 3.1. Conventional coaxial microwave antenna. Section A is designed to be a quarter
wavelength resonant length in the surrounding catheter / tissue medium. The section B
length is determined by the clinical situation and is equal to the depth of insertion. Junction
impedance is the sum of ZA and Zg.
The microwave antenna operates by means of a voltage which is applied at the antenna
junction. The impedance seen at the antenna junction Zjct is the sum of the two section
input impedances ZA and Zg. If these impedances are not equal, then the section having the
higher input resistance will radiate preferentially. For the conventional antenna shown in
Figure 3.1, ZA is independent of insertion depth (ZAT IS assumed to be infinite), but hg
and therefore ZB is a function of insertion depth.
The objective of the choke design in Figure 3.2 is to eliminate this problem. If the
choke in Figure 3.2 is perfect, then the termination impedance for section B, ZBT =
and
ZB = ZA at all insertion depths. In the practical design proposed here, however, the choke
is not ideal and thus ZBT is finite. Our goal was to characterize this practical choke
experimentally and theoretically. Such a design would be easily implemented in larger
antennas designed for intracavitary use, but may be feasible in future interstitial antenna
designs as well.
56
Coaxial Microwave Antenna with Choke
coaxial
feedline
(choke thickness)
(choke length)
For 915 MHz Antenna:
Ideal section lengths h^= hg= 3-3.5 cm
X/4 choke length = 6.4 cm
Detail of choke section
^BT Z f + Z c
Zf
Feedline dielectric
_
Section B
Section dielectric
I
Choice dielectric |
Figure 3.2. Microwave antenna with coaxial choke. Section length hg is now independent
of insertion depth. However, the ideal choke length is about twice as long as the ideal
antenna section length, so that the practical choke design is not ideal. Enlarged detail of the
choke shows that the termination impedance ZBT = Zf + ZC.
57
3.3 MATERIALS AND METHODS
Experimental antennas
In this study, we constructed several prototype antennas designed to operate at 915
MHz and we examined the performance of the antennas over a range of frequencies. A
microwave antenna equipped with a choke is shown schematically in Figure 3.2. The outer
conductor of the feedline is "folded over" a dielectric layer to form the radiating element of
the antenna, section B. For the choke to perform ideally, hg must be a quarter-wavelength
in the choke dielectric; for a typical plastic (such as Teflon) having a dielectric constant of
1.7, this corresponds to a length of 6.4 cm. at 915 MHz. On the other hand, we require hs
to be equal to hA and want the antenna to approximate a quarter-wave dipole in the
composite tissue/catheter dielectric medium. For 915 MHz, this corresponds to a length of
3 to 3.5 cm. Since a section length of 6.4 cm. would radiate poorly and be clinically
impractical, the antennas were made with hg = 3 to 3.5 cm. Therefore, the choke was nonideal at 915 MHz, and one of the objectives of this study was to determine the frequency
range over which the antenna/choke combination did have the desired effect.
Section B was constructed by applying heat-shrink tubing over the coaxial feedline to
form the choke dielectric layer. Antenna section B was formed by a length of braid slipped
over this heat-shrink layer and soldered to the feedline shield circumferentially. Section A
was formed by applying braid over a similar "dummy" length of feedline covered by a
similar layer of heat-shrink, so that both sections had the same diameter. The prototype
antennas were insulated with a layer of heatshrink tubing, and open end of the tube sealed
with silicone sealant.
Some of the important parameters of the prototype antennas studied are listed in Table
3.1. One antenna (RG58-2) was of a conventional non-choked design, but was larger than
our standard 1.0 to 1.6 mm O.D. interstitial antenna. Three antennas (RG174-2, RG174-3,
RG174-4) had a choke dielectric consisting of one layer (-0.3 mm) of heat-shrink tubing;
we call these "thin" chokes. Two antennas had a choke dielectric which we considered
"thick" (>1 mm). One antenna was made with a choke consisting of four heat-shrink tubing
layers (RG178-12); the second antenna having a thick choke was a larger antenna designed
for intracavitary hyperthermia. With one exception, all antennas were designed for
operation at 915 MHz and had section lengths, h^ and he, of about 3 cm. The exception
was one of the thin-choked antennas (RG174-3), which was constructed with hg = 6.4 cm
to form an ideal quarter-wavelength choke at 915 MHz. The thick-choke antennas had a
smaller-diameter feedline chosen so that the overall outer antenna diameters would be
nearly equal.
Table 3.1. Antenna parameters.
RG58.2 RG174-2 RG174.3
RG174.4 RG178.12 F12-2
Choke
None
Thickness,mm
Thin
0.30
Thin
0.32
Thin
0.33
Thick
1.11
Thick
1.49
hA, cm
he, cm
3.20
3.30
3.20
6.40
3.20
3.45
3.20
3.25
2.60
3.00
1.70
1.29
0.99
1.70
1.31
0.99
1.70
1.32
0.99
1.70
1.76
0.65
1.70
2.14
0.65
1.70
1.82
1.43
1.70
1.83
1.40
1.70
1.87
1.44
1.70
2.20
1.88
1.70
2.67
2.33
1.70
1.74
0.99
1.70
1.74
0.99
1.70
1.74
0.99
1.70
2.11
0.65
1.70
2.42
0.65
3.20
CHOKE DIELECTRIC
Gc
Cc, mm
ac, mm
ANTENNA INSULATION
Es
Cg, mm
as, mm
1.70
2.09
1.77
FEEDLINE INSULATION
Ef
Cf, mm
af, mm
1.70
2.09
1.77
59
Impedance measurements
Impedance was measured using the system shown in Figure 3.3. A digitally-controlled
sweep oscillator (HP 8350A) was used to drive the antenna through a dual-directional
coupler (HP 778D), and a vector voltmeter (HP 8508A) measured the forward and
reflected voltages through phase-matched cables (HP 11851B). For each antenna, a
reference feedline was constructed using an identical length of cable, but with a short circuit
soldered at the location of the antenna junction. This enabled the impedance seen at the
antenna junction, Zjct, to be directly referenced to a short circuit, and eliminated the effect
of feedline parameters. Software was developed on an HP87 microcomputer to control the
sweep oscillator, acquire the impedance data from the vector voltmeter, perform the
required calculations, and store the data on floppy disk.
Impedance measurements were made with the antennas embedded in muscleequivalent phantom (Guy, 1978). Zjct was determined for insertion depths (dj in Figures
3.1 and 3.2) of 3.5, 4, 5, 6, 7, 8, 9, 10, 11, and 12 cm. At each insertion depth,
impedance was measured at 20 MHz increments from 200 MHz to 2000 MHz. This
enabled us to observe the impedance variation with depth as a function of frequency. This
variation was quantified by calculating the standard deviation of the measured impedances
at the different depths for each frequency.
60
Measurement of Z at antenna junction
Data acquisition
and storage
HP 87 Microcomputer System
E
HP-IB
, HP-IB
HP 8508A
Vector Voltmeter
HP 8350A
Sweep Oscillator
1
A
Reference SC
Phasematched |REF
Dual-directional coupler
{
Antenna
Figure 3.3. Vector voltmeter system used to measure complex impedance. An HP 87
microcomputer controlled the sweep oscillator and stored the data from the vector
voltmeter.
Theoretical calculations
The antenna input impedance, Zjct, was calculated using insulated antenna theory
(King et al., 1981) to calculate ZA and Zg. Conceptually, the antenna section is a central
conductor of a coaxial transmission line with the catheter (region 2) serving as the dielectric
and the surrounding tissue (region 4) forming a lossy outer conductor, and terminated by
an impedance Zh. Wavenumbers k2 and Iq. in the two regions can be obtained from the
general expression for region n:
61
Note that the wavenumber, k, is related to the propagation constant, y, by the following
relationship:
k = -jY, where j = -i.
It was assumed that |i.2 =
= permeability of free space. For all of the catheter dielectric
materials (region 2), it was assumed that 02 = 0 and e2r (relative permittivity) = eq/eq = 1.7,
where Eq is the permittivity of free space. The tissue dielectric constant and conductivity, £4
and C4, are functions of frequency. The measured data reported by Schwan and Foster
(1980) and Guy et al. (1974) for the relative dielectric constant (e4r = 84/eo) and
conductivity (04, mho/m) of muscle were pooled to generate the following relationships:
e4r(/) = 74.128 -7.8787 logio(/)
04(f) = 0.99214 + (5.6815x10-4) f + (3.0982x10-8) /2 mho/m, where
/ = frequency in MHz, 200 <y < 2450.
Insulated antenna theory (King et al., 1981) yields the following relationships for
calculating the input impedance for an antenna section Zin{c, a, h, k4, k2, Zh}, where:
c
= outer radius of dielectric layer (region 2)
a
= inner radius of dielectric layer (region 2)
h
= length of antenna section
k4{f,£4,114,04} = complex wavenumber of surrounding tissue (region 4)
k2{f,£2.M'2.<72} = complex wavenumber of dielectric layer (region 2), and
Zh = termination impedance at the end of the antenna section.
The input impedance to the antenna section is given by a transmission line equation:
Zin = i Zo cot(KLh + i0h),
where ZQ is the characteristic impedance of the transmission line, KL is the line
wavenumber, and Qh is the complex terminal function:
62
Zo — Haf
^
kj k4cH^/\k4c).
1/2
ln| + F(k4c)
KL = k2
lng + ^F(k4c)
^4
0h =
In the above expressions, Hq \z) and Hj^\z) are Hankel functions of the first kind, orders 0
and 1, respectively, and
4^\k4C)
F(k4c) = k4C
\k4c)
In order to calculate ZB, it is necessary to find the termination impedance ZBT. which
is the sum of the feedline impedance Zf and the choke section impedance Zc (see detail in
Figure 3.2). Zf was calculated by again applying insulated antenna theory to the feedline
section beginning the end of the choke section and terminating at the insertion point of the
antenna, where the impedance was assumed to be infinite. For calculating Zc{cc, ac, h,
Ecr), an ideal lossless transmission line was assumed to be terminated with a short circuit,
and the relative permittivity of the choke dielectric Ecr was assumed to be 1.7:
VE^
"\ac/
tan
2%fie^
Cv hfi)
where Cy is the velocity of light in free space.
Calculations were implemented on a VAX 11/785 computer using the parameters
summarized in Table 3.2. Note that each theoretical calculation of Zjct required three
applications of insulated antenna theory and one transmission line calculation for the choke
63
impedance. The theoretical model also enabled us to explore the termination impedance ZBT
in terms of its components Zf and 2!c.
Table 3.2. Parameters used for theoretical calculation of impedances.
Ouantitv Calculation
Line Dielectric
Length
O.D. LD,
Termination Zh
ZA
lA
Antenna insulation Ssr
hA
2CS
2AS
ZAT = OO
ZB
lA
Antenna insulation E.
HB
2CS
2AS
ZBT = Zf + ZC
Zf
lA
Feedline insulation Cfr
di-hg
2CF
2AF
oo
Zc
TL
Choke dielectric
hB
2CC
2AC
0
ea
NOTES:
lA = Insulated antenna theory
TL = Lossless transmission line
62r = Relative dielectric constant of line (region 2)
64
3.4
RESULTS
Impedance measurements
Figure 3.4 shows the measured real and reactive components of the junction
impedance Zjct for three antennas as a function of frequency and insertion depth. For each
of the antennas, Zjct looks reasonably close to the ideal (50+j0)fll impedance at the design
frequency of 915 MHz. At each frequency, there are 10 data points each representing an
impedance value at a particular insertion depth. Therefore, the more scattered the data
points are at a given frequency, the greater is the impedance variation with insertion depth
at that frequency. In general, impedance can be seen to be strongly dependent on insertion
depth at lower frequencies, but this dependence decreases with increasing frequency.
Above around 1500 MHz, the antennas are electrically long, so that the input impedance for
all three antennas is essentially independent of insertion depth (note that the minimum
insertion depth is 3.5 cm). Figure 3.4(a) shows data for an antenna without a choke
(RG58-2), Figure 3.4(b) shows data for a thin-choked antenna (RG174-2), and Figure
3.4(c) illustrates results from a thick-choked antenna (RG178-12). Visually, it appears that
there is little difference in the "scatter" between Figures 3.4(a) and 3.4(b), but that the data
for the thick-choked antenna are clearly tighter at lower frequencies. This appears to be the
case for both real and imaginary components of the impedance. From these data, it would
appear that a reasonable way to quantify these results would be to look at the standard
deviation of the ten impedance values at each frequency. This is done in Figure 3.5, where
the standard deviation of the magnitude of the impedance is shown as a function of
frequency for 5 different antennas. In addition to the three antennas shown in Figure 3.4,
data from another thin-choked (RG174-4) and thick-choked (F12-2) antenna were added.
Antenna RG174-4 was intended to represent a replicate of RG174-2, and antenna F12-2
was a large antenna designed for transurethral hyperthermia of the prostate- it offers a more
extreme case of a thick choked antenna.
65
Junction impedance, antenna RG58-2 (no choke)
n
3.5 cm
•
4 cm
•
5 cm
•
6 cm
•
°
7cm
8 cm
A
9 cm
A
10 cm
• 11 cm
+ 12 cm
-i
200
500
r
800
1100 1400
Frequency (MHz)
1700
2000
Junction impedance, antenna RG 58-2 (no choke)
•
#
H
O
•
•
A
A
200
500
800
1100 1400
Frequency (MHz)
1700
2000
Figure 3.4(a). Real and imaginary components of the junction impedance from antennas
having no choke (RG58-2). At each frequency, there are 10 data points corresponding to
measurements made at different insertion depths.
66
Junction impedance, antenna RG174-2 (thin choke)
90
Q
3.5 cm
(0
E
a>
•
o
c
A
•f(0
OT
'35
CD
QC
A
•
+
200
500
800
1100 1400
Frequency (MHz)
1700
2000
Junction innpedance, antenna RG174-2 (thin choke)
20
•
(0
*
E
SI
a
o
o
•
0)
•
o
c
A
(0
A
o
CO
a>
(£
200
500
800
1100 1400
Frequency (MHz)
1700
2000
Figure 3.4(b). Real and imaginary components of the junction impedance from antennas
having a thin choke (RG174-2). At each frequency, there are 10 data points corresponding
to measurements made at different insertion depths.
67
Junction impedance, antenna RG 178-12 (thick choke)
80
.
h.
¥
200
500
800
1100 1400
Frequency (MHz)
1700
Q
3.5 cm
*
4 cm
•
5 cm
*
•
<3
6cm
7 cm
8 cm
A
9 cm
* 10 cm
• 11 cm
+ 12 cm
2000
Junction impedance, antenna RG178-12 (thick choke)
Q
3.5 cm
A
10 cm
+
12 cm
(0
E
.c
o
S
-15-
c
(0
*-
u
CO
<D
CC
-45
-60
200
500
800
1100 1400
Frequency (MHz)
1700
2000
Figure 3.4(c). Real and imaginary components of the junction impedance from antennas
having a thick choke (RG178-12). At each frequency, there are 10 data points
corresponding to measurements made at different insertion depths.
68
The data in Figure 3.5 suggest that the antennas with thin chokes do not behave
significantly differently from the antenna with no choke. However, the thick-choked
antennas clearly show a reduction in standard deviation (and hence dependence on insertion
depth) at the lower frequencies. This implies that although the choke was not designed to
operate at the antenna design frequency of 915 MHz, a thick choke may in practice reduce
the effect of insertion depth at frequencies below its expected operating frequency (a choke
of 3.2 cm would be expected to operate at about 1800 MHz).
Figure 3.5(c) shows the variation with insertion depth of a thin choked antenna having
a section length he = 6.4 cm to produce a choke with a design frequency of 915 MHz. The
two thin-choked antennas in Figure 3.5(a) are reproduced in the figure for comparison.
Clearly, this antenna shows little impedance variation with insertion depth at 800-1000
MHz, and the thin choke is effective if its design length is chosen to be more ideal.
However, a section length of hg = 6.4 cm would result in a poor radiation pattern, as the
section input resistance of section B (Rg, the real component of Zg), would be several
times larger than RA (the real component of ZA) and would therefore radiate preferentially,
drawing power deposition away from the antenna tip.
69
Effect of choke thickness, experiment
No choke (RGS8-2)
«
Thin choke (RG174-2)
Thin choke (RG174-4)
10"
Thick choke (RG178-12)
Thick choke (F12-2)
200
400
600
800
1000
1200
Frequency (MHz)
Effect of choke thickness, theory
-
No choke (RGS8-2)
-A—
Thin choke (RG174-2)
Hi—
Thick choke (RG178-12)
O 10
200
500
800
1100
1400
1700
2000
Frequency (MHz)
Figure 3.5(a),(b). Standard deviation data from the impedance data. At each frequency, the
standard deviation was obtained of the impedance at ten different insertion depths, a)
experimental results from the antennas in table 2, and b) theoretical results for
conventional, thin choke, and thick choke antennas.
70
Effect of choke length, experiment
hB = 3.30cm(RG174-2)
hB = 3.45cm(RG174-4)
hB = 6.40ctn(RG174-3)
12-
(0
E
N
>
O
Q
•o
w
3-
200
500
800
1100
1400
1700
2000
Frequency (MHz)
Figure 3.5(c). Standard deviation data from the impedance data. At each frequency, the
standard deviation was obtained of the impedance at ten different insertion depths.
Comparison of performance when choke length is optimized for 915 MHz (hg = 6.4 cm).
Theoretical results
Three antennas: RG58-2 (no choke), RG174-2 (thin choke), and RG178 (thick choke)
were analyzed theoretically. The theoretical results are compared with the experimental
measurements in Figure 3.6. For each of the three antennas, the real and reactive
components of the impedance are plotted as a function of frequency at a shallow (4 cm) and
deep (12 cm) junction depth. It can be seen that although the absolute accuracy of the
theoretical calculations seem to be only fair, the qualitative agreement between theory and
experiment is acceptable.
The performance of the choke ultimately depends on the termination impedance ZBT.
which is in turn the sum of Zf and Z^. Zf will vary with insertion depth and frequency.
71
while Zeis a function of frequency only and will be purely reactive due to the assumption
that the choke dielectric is lossless. Figure 3.7 shows the theoretical behavior of Zc as a
function of fi"equency for the thin choke and thick choke antennas. As seen in the figure,
the frequency at which the choke operates ideally is not the antenna design frequency of
915 MHz, but is at about 1800 MHz. Although it cannot be seen clearly in the plot, Zc is
higher for the antenna having the thick choke by about j30£2 at 915 MHz. Because of its
higher characteristic impedance (larger Cc/ac ratio), the thicker choke has a broader
resonance peak than the thin-choked antenna and can have an effect over a wider frequency
range. Note that the choke reactance (Zc) is positive at frequencies below its resonance peak
and negative at frequencies above this peak (Figure 3.7).
Figure 3.8 shows the real and reactive components of Zf as a function of frequency for
the thin (3.8(a)) and thick (3.8(b)) choke antennas. The data from 10 different insertion
depths are combined. The thick choked antenna has a larger Cf/af ratio resulting in a larger
characteristic impedance; consequently, it can be seen that Zf is generally larger for antenna
RG178-12. Thus, both Zf and Zc are generally larger in magnitude for the thicker choked
antenna.
Since ZBT is the sum of Zf and Zic, where Zf represents the termination impedance that
an unchoked antenna would have, we are most interested in the relative contribution of 2^
to ZBT- Figure 3.9 shows the theoretical relationship between Zf and Zc at 433 MHz, 915
MHz, and 2450 MHz for the thin choke (RG174-2) and thick choke (RG178-12) antennas.
The labels on the data points represent junction insertion depths in cm. Since Zf is a
transmission line terminated at the tissue surface (an open circuit), its initial value will be
(O-joo)Q at small insertion depths, and it will form a spiral with increasing insertion depth,
converging around its line characteristic impedance. Since Zc is purely reactive and
independent of insertion depth, the termination impedance ZBT IS represented by the Zf
spiral shifted along the jX axis by a fixed value. This is clearly illustrated in Figure 3.9. At
all three frequencies, the thick choke causes a greater positive shift of the feedline
72
impedance. At 433 MHz, Figure 3.9(a) shows that the choke contributes relatively little for
either antenna. At 915 MHz, the thick choke is seen to contribute more significantly than
the thin choke in Figure 3.9(b). At 2450 MHz, Zc dominates and the entire spiral is shifted
(Figure 3.9(c); note that this is above the choke resonant frequency, so that the Zf spiral is
shifted by a negative reactance. Again, it should be reemphasized that the choke design
frequency is about 18(X) MHz, at which point the Zf spiral would be shifted by an infinite
reactance value.
73
Junction impedance, antenna RG58-2 (no choke)
o Theory, depth=4cm
A Theory, deplh=12cm
Experiment, depths4cm
A Experiment, depth=12cin
200
500
800
1100
1700
1400
2000
Frequency (MHz)
Junction impedance, antenna 58-2 (no choke)
20
(0
0
E
®
-20
o
o
m -40
o
o
A
Theory, depth=4cm
Theory, depth=12cm
• Experiment, depth=4cm
• Experiment, depth=12cm
•m °
(0
£ -60 "
a
.•
-80
11
200
500
1
800
I
1
1100
'
1
1400
•
»
•
1700
2000
Frequency (MHz)
Figure 3.6(a). Comparison between experiment and theoretical junction impedance for
shallow (4 cm) and deep (12 cm) insertion depth, antenna RG58-2 (no choke).
74
Junction impedance, antenna RG174-2 (thin choke)
CO
E
o
c
(0
-+-•
Experiment, deplh=4cm
A Experiment, dep*h=12cm
o Theory, depth=4cm
OT
W
O
QC
A
200
500
Theoiy,depth=12cm
800
1100
1700
1400
2000
Frequency (MHz)
Junction impedance, antenna RG174-2 (thin choke)
—
20
<n
0
I -
2
0
I
o
. -40
0)
o
C
•j
CO
-60
•
•
A
Q
A
a
°
'80'
_
'
Experiment, depth=4cm
Experiment, depth=12cm
Theory, depth=4cm
Theory, depth=12cm
-100"
.o
"120 "4
200
•
»
500
•
I
800
'
I
1100
*
I
1400
•
I
•
1700
2000
Frequency (MHz)
Figure 3.6(b). Comparison between experiment and theoretical junction impedance for
shallow (4 cm) and deep (12 cm) insertion depth, antenna RG174-2 (thin choke).
75
Junction impedance, antenna RG178-12 (tliick choke)
• Experiment, deplh=4cm
A Experiment, deplh=12cm
o Theory, depth=4cm
A Theory, depthzl2cm
10
200
•
I
500
'
I
800
•
I
1100
'
I
1400
'
I
1700
2000
Frequency (MHz)
Junction impedance, antenna RG178-12 (thick choke)
g -40• Experiment, depth=4cm
A Experiment, depthsl2cm
o Theory, depth=4cm
A Theoiy, depth=12cm
-100
200
500
800
1100
1400
1700
2000
Frequency (MHz)
Figure 3.6(c). Comparison between experiment and theoretical junction impedance for
shallow (4 cm) and deep (12 cm) insertion depth, antenna RG178-12 (thick choke).
76
Choke reactance Zc
6000
4000-
Thin choke (RG174-2)
to
•C
o
Thick choke (RG178-12)
2000 -
8
C
(0
4—
o
(0
0)
oc
-2000 -
-4000 -6000
200
500
800
1100
1400
Frequency (MHz)
1700
2000
Figure 3.7. Choke impedance Zc for thin and thick chokes, section lengths 3.25 - 3.3 cm.
77
Zf, antenna RG174-2, 3.5 ^ depth < 12 cm
120
• 3.5 cm
n
(0
E
•
£
O
A
A
QC
200
500
800
1100 1400
Frequency (MHz)
1700
2000
Zf, antenna RG174-2, 3.5 ^ depth :S 12 cm
•
•
•
A
A
X -1200"
-1600"
-2000
200
500
800
1100 1400
Frequency (MHz)
1700
2000
Figure 3.8(a). Feedline impedance Zf for thin (RG174-2) choked antennas at various
junction depths.
78
Zf, antenna RG178-12, 3.5 ^ depth ^ 12 cm
o
«
3.5 cm
90•
A
A
200
500
800
1100 1400
Frequency (MHz)
1700
2000
Zf, antenna RG178-12, 3.5 < depth < 12 cm
a
-400'
•
A
>/'1200 •
A
-1600
-2000
200
-I
500
r
800
1100 1400
Frequency (MHz)
1700
2000
Figure 3.8(b). Feedline impedance Zf for thick (RG178-12) choked antennas at various
junction depths.
79
Antenna RG174-2: contribution of Zc, 433 MHz
tn
E
xr
o
>c
o Zf
• Zf + Zc
20
R, ohms
Antenna RG 1 78-12; contribution of Zc, 433 MHz
-80 -
-120-
Zf + Zc
-160-
R, ohms
Figure 3.9(a). Theoretical feedline impedance (Zf, open circles) at various insertion depths.
The numbers labelling the data points are antenna (junction) insertion depths in cm. The filled
circles represent the total termination impedance, ZBT. when the choke impedance (ZC) is
added. Data are shown for thin (RG174-2) and thick (RG178-12) choked antennas at 433
MHz.
80
Antenna RG 174-2: contribution of Zc, 915 MHz
50
250
CO
# -25.
x"
-7500
0
40
R, ohms
20
80
60
Antenna RG178-12: contribution of Zc, 915 MHz
00
500 - -
9-
-50X
00-
50-200
0
40
80
R, ohms
120
Figure 3.9(b). Theoretical feedline impedance (ZF, open circles) at various insertion depths.
The numbers labelling the data points are antenna function) insertion depths in cm. The filled
circles represent the total termination impedance, ZBT. when the choke impedance (ZC) is
added. Data are shown for thin (RG174-2) and thick (RG178-12) choked antennas at 915
MHz.
81
Antenna RG 174-2: contribution of Zc, 2450 MHz
0.
-10-1
o
Zf
e
Zf + Zc
E -20-1
sz
o
xT "30 •r—^
-40.
-50
T
0
0
20
R, ohms
"T
30
40
Antenna RG i 78-12: contribution of Zc, 2450 MHz
-40-80-
60-
-200
0
20
40
R, ohms
60
80
Figure 3.9(c). Theoretical feedline impedance (ZF, open circles) at various insertion depths.
The numbers labelling the data points are antenna (junction) insertion depths in cm. The filled
circles represent the total termination impedance, ZBT. when the choke impedance (ZC) is
added. Data are shown for thin (RG174-2) and thick (RG178-12) choked antennas at 2450
MHz.
82
3.5
DISCUSSION
In this study, a 915 MHz interstitial microwave antenna design was modified to
include a choke. Because the wavenumber in the choke dielectric was very different from
that formed by the composite media surrounding the antenna, the choice of section length
he had to be a compromise between optimum antenna radiation and optimum choke
operation. A section length he of 6.4 cm based on ideal choke design would have an input
resistance three to four time greater than the ideal 3.2 cm antenna section (King et al.,
1981). Such an antenna would radiate preferentially over the shorter section A, displacing
power deposition away from the tip. The identical situation occurs when a non-choked
antenna is inserted with its junction 6.4 cm deep into phantom, where the resulting poor
power deposition pattern has been demonstrated experimentally (Jones et al., 1988).
Therefore, we chose hg for optimum radiation in tissue (approximately 3.2 cm) at 915
MHz. This resulted in a choke with an actual design frequency of 1800 MHz. By
observing the performance of the choke over a range of firequencies, our objective was to
determine the extent to which such a practical, but non-ideal choke can function.
Using insulated antenna theory and conventional transmission line expressions, we
have presented a theoretical model for calculating the junction impedance of a choked
antenna. Several factors may have contributed to the differences seen between the
experimental and calculated results. First, the insulated antenna theory is valid under the
following conditions (Trembly, 1982):
1)
k4
k2
1
2)
c/a > 2 ,
3)
|k2Cp«l,and
4)
|k4c| <0.5 .
83
Although conditions 1,3, and 4 are met in all cases, the insulation thicknesses were
generally too thin for the theory to be stricdy applicable; for the antenna section calculations
c/a averaged 1.23, and for the feedline calculations c/a averaged 1.61 for the no choke and
thin choke antenna feedlines. It should be emphasized that each calculation of junction
impedance required the insulated antenna theory to be applied three times; therefore, any
errors encountered in its implementation will be amplified. Secondly, the construction of
the antennas and choke was not ideal: the antenna junction gap is not infinitesimal and the
chokes were made using a circumferential solder connection having finite dimensions.
Third, the reference short circuits used in the impedance measurement were made by
soldering the center conductor to the outer shield of the feedline and were not ideal. Fourth,
the dielectric properties of the Guy phantom used experimentally were probably somewhat
different from those calculated in the theoretical model. Finally, the measured dimensions
of the antenna were subject to variability; for example, the outer diameter of an antenna
section would measure differently at various points along the section or if the antenna was
rotated.
The experimental results clearly show a periodic impedance fluctuation which does not
appear in the theoretical result. These fluctuations did not appear when measurements were
repeated on a 50Q termination; therefore, we conclude that these undulations are due to
higher order effects not accounted for in this theoretical model. In spite of the many
possible sources of error, however, there is reasonable agreement between theory and
experiment (Figure 3.6), and the theoretical model is useful for predicting the general
behavior of choked antennas.
Measuring the junction impedance at 10 different insertion depths for each frequency
permitted reliable estimates to be made of the variation of Zjct with insertion depth, which
we quantified using the standard deviation. It was found that using either the real or
imaginary component of the impedance resulted in the same conclusions, so the magnitude
of the impedance was used in the subsequent antenna comparisons. In our experiments, we
84
compared a conventional antenna without a choke to two antennas having a thin choke and
two antennas having a thick choke dielectric. At high frequencies (>1500 MHz), Zjct was
independent of insertion depth for all of the antennas, presumably because the minimum
insertion depth of 3.5 cm was relatively long for these frequencies, and antenna impedance
had approached the characteristic impedance value!. Since the choke would be expected to
operate at a higher frequency (1800 MHz), the effect of the choke is not readily apparent in
Figures 3.4(a)-(c) and 3.5(a). In addition, we found that the thin choke antennas did not
behave significantly different from the unchoked antenna at lower frequencies. From these
experiments, the thin choke contributed little near the frequency of interest, 915 MHz.
Figure 3.5(b) demonstrates that the thin choke would function as expected if it could
be allowed to operate near its ideal frequency. Such a choke could be designed if the
relative dielectric constant of the choke dielectric layer were chosen to make the section
length, hg = 3.2 cm equal to a quarter-wavelength. A theoretical thin-choked antenna was
postulated which was identical to antenna RG174-2, except that the ideal dielectric was
chosen for the choke layer (ecr = 6.49). The theoretical standard deviation curve obtained
from this idealized antenna is shown in Figure 3.10. As predicted, this antenna would have
constant impedance with insertion depth at 915 MHz.
Our experimental results demonstrate that the thick choke (>1 mm) can significantly
improve the performance of the non-ideal choke (Figure 3.5(a)) at lower frequencies. The
theoretical model also predicts this outcome, but the effect is not as pronounced. From
Figure 3.7, it can be seen that the thick choke has a broader impedance peak due to its
larger characteristic impedance (greater Cc/ac ratio); thus, Zc has some effect even an octave
away from the resonant choke frequency. Figure 3.9 shows that the choke contributes
relatively more to the total terminal impedance as frequencies get closer to its resonant
frequency (433 MHz, 915 MHz, and 2450 MHz are about 2,1, and 0.5 octave away). The
theory also tells us that the feedline impedance Zf is also higher for the thick-choked
antennas, similarly by virtue of a higher Cf/af ratio. From the data shown in Figure 3.9, it
85
can be seen that adding the choke impedance Zc to the feedline impedance Zf does not
always result in a higher termination impedance ZBT. because ZC can either add or subract
from the reactive component of Zf. However, the addition of Zc to Zf will always make the
total termination impedance ZBT relatively less dependent on insertion depth, because ZC is
constant. Further investigation is needed, perhaps with SAR experiments, to determine
whether a high termination impedance is required, or if a lower but constant impedance is
adequate.
RG174 antennas, effect of choke dielectric
e= 1.70
15-
e = 6.49
OT
E
!1 0 N
Q
(0
200
500
800
1100
1400
1700
2000
Frequency (MHz)
Figure 3.10. Theoretical performance of a choked antenna with an ideal choke dielectric. The
antenna design is optimum both in terms of radiation properties and choke performance.
86
Finally, the theory suggests possible design strategies to improve the choke
performance. The goal of the choke design is to maximize ZBT = ZC + ZF. First, the choke
would be ideal if the choke dielectric could be designed to match the section length
requirement of the antenna; in this case, the ideal Ecr = 6.49. Even if this is not attainable,
increasing the dielectric constant toward this ideal value would move the peak in Figure 3.7
to lower frequencies, increasing Zc and improving choke performance at 915 MHz.
Secondly, a thicker choke layer increases Zc and improves performance when the antenna
is operating away from the choke resonant frequency; it broadens the peak as shown in
Figure 3.7. A thin choke layer is effective when operating at the quarter wavelength
resonant frequency, but a thin layer would result in a very narrow-band device because the
characteristic impedance of the choke would be small. Third, Zf can be increased by
making Cf/af large. Thus, performance of the choked antenna may be improved by using a
small diameter coaxial cable under a thick layer of insulation to feed the antenna elements.
87
3.6
CONCLUSIONS
The challenge of clinical hyperthermia is to deliver a controlled and well-circumscribed
temperature distribution to the tumor volume, without damaging the surrounding normal
tissues. Interstitial microwave techniques have been used to deliver hyperthermia to a wide
variety of tumors, ranging from readily accessible superficial tumors to deep-seated tumors
requiring intraoperative placement of the antenna catheters. One of the significant problems
of the flexible coaxial antennas used for these procedures is the fact that the performance of
the antenna changes with insertion depth. In this paper, we consider a choke as one
possible solution. If the antenna is designed for optimum radiation in tissue and constructed
using conventional plastic dielectric materials, the choke design cannot be ideal and will not
be operating near its resonant frequency. Using antenna impedance as a function of depth,
we have investigated this non-ideal choke experimentally and developed a theoretical model
which supports the experimental results and accurately predicts the general behavior of
these choked antennas.
Our theoretical and experimental results show that the performance of the non-ideal
choke can be improved by 1) choosing a choke dielectric having a relative dielectric
constant closer to the theoretical ideal value (in this case Ecr = 6.49), 2) increasing the
thickness of the choke dielectric layer, and 3) using a small coaxial cable to feed the
antenna elements with a relatively thick insulation layer. Although the analysis of antenna
impedance has allowed some of the important design parameters to be elucidated, the
ultimate performance of the choke will need to be determined using SAR measurements.
The implication of these design considerations is that it may be difficult to develop a
practical coaxial microwave antenna equipped with a choke having dimensions small
enough to be used for interstitial hyperthermia (< ~1.5-2mm O.D.). One possible design
option would be to employ a dielectric film having the ideal dielectric constant for the choke
layer. This would result in ideal choke performance at a particular driving frequency, but
the antenna dimensions and input frequency might need to be very precise, because the
88
very thin choke layer would result in an even narrower reactance spike than that shown in
Figure 3.7 (due to the fact that ^ would be very small); it may even be necessary to "tune"
such an antenna. Using a small coaxial cable to feed the antenna may result in significant
power loss along the line, so the length of this feedline should be minimized and
connection to the generator should be made by connecting to a larger feeder cable.
Finally, the choke design is easily incorporated into larger coaxial antennas, such as
those designed for intracavitary hyperthermia. The ability to use a thick choke layer and a
thick layer of insulation over the feedline enables the choke to function adequately far from
its resonant frequency. When properly implemented, a choke can greatly improve the
consistency of microwave antenna by making its performance less dependent on insertion
depth. This improves the predictability of the resulting temperature distribution, reduces
patient discomfort, and gives the clinician confidence when planning and delivering
hyperthermia treatments.
4.0 A COAXIAL MICROWAVE APPLICATOR
FOR TRANSURETHRAL HYPERTHERMIA OF
THE PROSTATE
submitted to International Journal of Radiation Oncology,
Biology, and Physics
89
90
4.1 ABSTRACT
Benign prostatic hyperplasia (BPH) is a common disease of elderly men. The current
definitive treatment for urinary obstruction caused by this disease is surgery (transurethral
resection of the prostate, or TURP). Recent evidence suggests that hyperthermia may be
useful non-surgical altermative for treatment of symptomatic BPH. A transurethral
microwave applicator has been designed around a Foley catheter for delivery of local
hyperthermia to the prostate. The Foley balloon is used to maintain the antenna position
within the prostatic urethra. The Foley catheter also serves as a choke to confine power
deposition to the intended region. The antenna is a coaxial dipole designed to operate at
915 MHz, but has low reflected power at 433 MHz and 2450 MHz. Qualitative and
quantitative SAR patterns are shown for this antenna at these three frequencies. In vivo
experiments in dog demonstrated that temperatures >42°C could be obtained while
maintaining a maximum urethral temperature of 47°C to 48°C. Histology obtained acutely
after the hyperthermia treatments showed minimal damage to the periurethral tissues. This
transurethral applicator can be used alone to treat BPH, or in combination with coaxial
rectal antennas for treating prostatic carcinoma.
91
4.2 INTRODUCTION
Benign prostatic hyperplasia (BPH) is a common disease of elderly men, occurring in
the majority of men over the age of 60 (Geller, 1989), and in over 90% of men over the age
of 80 (Sagalowsky and Wilson,1987). BPH is the most common cause of urinary
obstruction in men, and 10%-20% of men will require prostatic surgery at some time in
their lives to relieve obstructive symptoms (Sagalowsky and Wilson, 1987; Lepor, 1989).
Currentiy, surgical resection is the only accepted treatment for BPH, and approximately
300,000 operations are performed annually (Geller, 1989). Morbidity and mortality for
transurethral resection of the prostate (TURP) are 17% and 1%, respectively (Geller,
1989). The fact that many patients suffering from BPH are elderly and may not be
candidates for surgery has resulted in numerous attempts for non-surgical management of
this disease. Alternatives to surgery have included medications (a-adrenergic antagonists,
5-a reductase blockers, hormones) and mechanical dilatation (Lepor, 1989). Recently,
there has been evidence that hyperthermia may be a useful modality for the management of
symptomatic BPH (Lindner et al., 1987; Sapozink et al., 1989; Servadio et al., 1986,
1987; Yerushalmi etal., 1986).
92
4.3 MATERIALS AND METHODS
Theoretical Considerations
The transurethral coaxial microwave applicator is based on the insulated antenna
design which we have used for an interstitial microwave antenna array hyperthermia system
and has been described previously (Coughlin et al., 1983, 1985). These small coaxial
antennas range from 0.9 to 1.6 mm diameter and have been studied in detail experimentally
(Wong et al., 1986; Jones, 1986, 1988) and theoretically (King et al., 1981, 1983;
Mechling, 1989; Trembly, 1982; Tumeh and Iskander, 1989). The transurethral antenna is
different from these interstitial antennas in two ways. First, the antenna is much larger in
diameter (5.3 mm). Secondly, section B of the antenna is separated from the incoming
antenna feedline by a dielectric layer formed by the Foley catheter. This dielectric layer is
1.49 mm thick and forms a choke which reduces the antenna current which would
ordinarily flow all the way up the antenna feedline to the insertion point. An ideal choke
would have a length of a quarter-wavelength in the dielectric separating the antenna section
and the feedline, or approximately twice the present length. However, a previous study has
demonstrated that a non-ideal choke having a dielectric layer of diis thickness can make the
input impedance and radiation pattern of the antenna independent of insertion depth (Wong
and Trembly, 1990).
Transurethral microwave antenna
The transurethral microwave applicator was constructed around a standard 12 Fr Foley
urethral drainage catheter (C.R. Bard, Covington, GA). The Foley catheter features a
balloon which is inflated within the bladder to prevent it from being pulled out of the
urethra (Figure 4.1). The balloon also ensures that the heating elements of the hyperthermia
antenna are located within the prostatic urethra. The antenna elements, sections A and B,
are equal in length to form a symmetrical coaxial dipole and are located so as to radiate from
the proximal margin of the balloon over a length of 4 to 6 cm. The elements are formed
93
from a fine braid which was custom manufactured (New England Electric Wire, Lisbon,
NH) for maximum flexibility. The feedline is RG-178 coaxial cable with the outer jacket
removed. The coaxial feedline is passed through the drainage channel of the Foley catheter
and connections to the antenna sections are made through an exit hole made at the junction
region through the catheter. Section A is electrically connected to the inner conductor of the
feedline, and section B is connected to the shield of the feedline. The present antenna is
designed to operate at 915 MHz. The section lengths, hA and hg were chosen for maximum
radiating efficiency at this frequency (i.e. the resonant length which corresponds to a
quarter-wavelength in the composite tissue/catheter medium). The resonant length was
determined by theoretical calculation using the insulated antenna theory developed by King
(King et al., 1981), and verified using reflected power measurements to determine the
frequency at which reflected power was minimized. Actual section lengths as measured on
the antenna were hA = 2.6 cm and he = 3.0 cm.
The Foley antenna was insulated up to the proximal edge of the balloon with a layer of
polyolefin heat-shrink tubing for maximum flexibility (VFP-876, 3M Corporation, Austin,
TX). The end of the heat shrink layer at the antenna terminal end of section A and the
urinary drainage opening of the Foley catheter was sealed with silicone sealant. This
ensured that the antenna elements and feedline would be fully insulated. A 19 gauge
thermometry catheter was attached along the outside of the antenna also using silicone
sealant. The overall diameter of the applicator was nominally 5.3 mm, with a maximum
diameter of 7.0 mm due to the attached thermometry catheter.
94
(not to scale)
Antenna Elements
Foley Balloon
Insulating Layer
Thermometry Catheter
Foley Catheter
Antenna elements:
Section A connected to center conductor of feedline
Section B connected to shield of feedline
(a)
Cross-section
(not to scale)
Insulation Layer
Foley Catheter
(b)
Antenna Elements
A Balloon channel
B Urinary drainage channel containing coaxial feedline
C Thermometry catheter
Figure 4.1. Diagram of transurethral hyperthermia applicator: a) longitudinal view
showing antenna sections A and B, b) cross-sectional view showing feedline within the
urinary drainage channel and partially external thermomedy catheter.
95
Impedance measurements
The antenna junction impedance was measured using a vector-voltmeter (HP 8508A)
and a digitally controlled sweep oscillator (HP 8350A). These devices were controlled by
an HP 87 microcomputer, which carried out the necessary calculations and stored the data.
A reference cable was constructed which was identical to the feedline of the antenna, except
that a short circuit was soldered at the location corresponding to the junction of the antenna.
Impedance measurements were made using this short circuit as a reference, so that the
impedance at the antenna junction was determined directiy and was independent of the
feedline parameters. Impedance was measured at ten different insertion depths ranging
from 3.5 cm to 12 cm (tissue surface to antenna junction) in Guy muscle-equivalent
phantom (Guy, 1978). Impedance was measured at the 915 MHz design frequency. In
addition, measurements were made at other frequencies of interest for microwave
hyperthermia: 433 MHz and 2000 MHz. The latter represented the upper limit of the
measurement equipment, but should approximate the antenna impedance at 2450 MHz,
because the antenna is electrically long at these high frequencies and the measured
impedance approximates the characteristic impedance of an infinitely long antenna at both
of these frequencies.
SAR measurements
The SAR (specific absorption rate, W/kg) distributions produced by the transurethral
antenna were evaluated qualitatively using liquid crystal sheets and quantitatively using a
computer controlled system. All SAR experiments were performed with the antenna placed
in a transparent HEC (hydroxy-ethyl-cellulose) muscle-equivalent phantom (Hartsgrove et
al., 1987).
Qualitative SAR patterns were produced using rectangular sections of liquid crystal
paper (30-35°C range) which were placed on either side of the antenna and secured with
suture. The phantom container was placed on a photographic copy stand in a temperature
96
controlled water bath which was maintained at 29.0°C for 6 to 8 hours prior to the
experiment to allow the entire phantom to equilibrate. This maximized the sensitivity and
reproducibility of the SAR patterns determined by this technique. After the phantom had
equilibrated, 20W of forward power was applied to the antenna for 15 seconds, at which
time a photograph was taken to record the resulting temperature pattern as an approximation
of the SAR.
Quantitative SAR measurements were made using a computer controlled system which
has been described previously (Wong et al., 1986). In this system, the computer pulses the
antenna with power for 60 seconds and calculates the initial rate of temperature rise using a
quadratic least-squares fit to the temperature data. SAR is then determined from this rate of
temperature rise:
SAR = c
lt=0+ ,
where c is the heat capacity of the tissue-equivalent phantom. Quantitative SAR
measurements were made along the antenna's built-in thermometry catheter and along the
antenna at a distance of 5 mm. For the latter measurements, the SAR data from four
surrounding thermometry catheters were averaged. SAR measurements were also made
transversely across the antenna to evaluate the radial power deposition. SAR patterns were
measured from the antenna at the three common clinical microwave frequencies: 433 MHz,
915 MHz, and 2450 MHz. For all SAR calculations, the net forward power at the antenna
junction was calculated to account for the reflected power and the losses in the feedline, and
results were plotted as SAR per watt applied at the antenna junction.
In vivo experiments
The transurethral antenna was tested in three mongrel dogs. The dogs were
anesthetized using an appropriate induction and maintenance dose of pentobarbital. The
dogs were placed in a supine position, and a perineal urethrostomy was performed to allow
introduction of the transurethral applicator. The symphysis was exposed and the prostate
97
bluntly dissected from underneath the the upper edge of the symphysis pubis. Two plastic
18 gauge pointed closed-end thermometry catheters were inserted into the prostate
perpendicular to the urethra in an anteroposterior direction. The transurethral antenna was
lubricated generously with K-Y jelly and easily inserted into the bladder. The Foley balloon
was inflated with 5 cc water and pulled gently to the bladder neck. The position of the
antenna in the prostate was verified by palpating the balloon in the bladder neck. Baseline
temperatures were measured along all thermometry catheters prior to each hyperthermia
application. Microwave power was then applied to the antenna, and a preliminary thermal
map obtained along the catheter to locate the maximum temperature (which occurred at the
antenna junction). This reference point was maintained at a predetermined control
temperature (45°C-48°C) for at least 20 minutes, after which steady-state temperature maps
were made along the antenna and at locations within the prostate at 433 MHz, 915 MHz,
and 2450 MHz.
Microwave power was provided by an American Microwave Technology model 1120
generator for 915 MHz experiments. 433 MHz power was provided by a power source
which was custom-designed by M/A Com, and 2450 MHz power was provided by a
Cheung Labs MPS2450-300 system. Forward and reflected power was measured
simultaneously by two HP 435 power meters and appropriate dual-directional couplers. In
all cases, power was controlled manually and required only minor adjustments to maintain
the desired temperatures.
An 8-channel Luxtron 3000 fiberoptic thermometry system was used to measure
temperatures; the thermometry was calibrated at 42.0°C in a temperature controlled water
bath just prior to each experiment. Temperature data were recorded and stored by an IBMbased data-acquisition system. Temperature maps were generated by an automatic stepper
motor system controlled by a separate IBM AT computer (Ryan et al., 1990) This system
permitted temperature distributions from all thermometry catheters to be obtained
simultaneously and displayed in real time during the experiment. Data were obtained at 2
98
mm intervals along the TMA and along the plastic 18 gauge pointed closed-end catheters
implanted in the prostate.
After completion of the experiments, the dogs were euthanized in a humane manner
using a standard solution. The superior and inferior pubic rami were divided using a bone
cutter, permitting the entire symphysis to be removed. This yielded excellent exposure, and
the bladder, prostate, and urethra (up to the perineum) were removed en bloc. The
thermometry catheters were clipped but not removed, and the complete block of tissue was
immediately placed in a formalin solution.
99
4.4
RESULTS
Impedance measurements
In Figure 4.2, impedance at the antenna junction is shown as a function of insertion
depth (junction depth in tissue) at 433 MHz, 915 MHz, and 2000 MHz. As expected, the
impedance at 915 MHz most closely approximates the ideal impedance of (50+0j) ohms.
There is a reactive component of about 20 ohms present, which may be due to the fact that
the antenna junction gap is not infinitesimal. Based on these junction impedances, we
would expect power transmission to be 95% at 915 MHz, 89% at 433 MHz, and 90% at
2000 MHz. Therefore, the antenna would be expected to radiate efficiently at all three of
these frequencies. As previously mentioned, we were not able to measure impedance at
2450 MHz, but we would not expect a large impedance difference between 2000 MHz and
2450 MHz.
The impedance is seen to vary by less than 5 ohms over all insertion depths at 915
MHz and 2000 MHz. At 433 MHz, the impedance varied by about 10 ohms over the range
of insertion depths tested, which suggests that the antenna performance at this frequency
may change with insertion depth. At 915 MHz and 2000 MHz, there is little change in
impedance with insertion depth which suggests that insertion depth will not affect antenna
performance at these frequencies.
100
55
504540° 30
oc
25
433 MHz
915 MHz
2000 MHz
20
3
4
5
6
7
8
9
10
12
Junction depth (cm)
30
20-
433 MHz
915 MHz
2000 MHz
(/)
E
xz
o
x"
-20
3
4
5
8
9
Junction depth (cm)
6
7
10
12
Figure 4.2. Real (a) and reactive (b) components of impedance at the antenna junction as a
function of insertion depth. Note that at the design frequency of 915 MHz, the antenna
impedance is independent of insertion depth.
101
SAR measurements
Temperature distributions photographed 15 seconds after power was applied were
used to approximate the SAR distribution qualitatively. Data are shown in Figure 4.3 for
driving frequencies of 433 MHz, 915 MHz, and 2450 MHz. In all cases, the forward
power to the antenna feedline was 20 W, and reflected power was less than 10%, so that
the net forward power at the antenna junction was roughly comparable at the three
frequencies. At 433 MHz, the SAR distribution is uniform longitudinally, but has poor
radial extension. Note that the power deposition is confined to the length of the antenna
elements. At the design frequency of 915 MHz, the SAR pattern is similar to that of the
smaller interstitial dipoles, with the maximum at the junction and decreasing over the length
of each antenna section. Finally, the SAR distribution at 2450 MHz is a more confined to
the junction region of the antenna with good radial extent.
Quantitative SAR measurements were made along the antenna thermometry catheter
and longitudinally at a distance of 5 mm from the antenna. These results are shown in
Figure 4.4, and are consistent with the qualitative patterns in Figure 4.3. Note that the peak
SAR at the junction is highest for 2450 MHz and lowest for 433 MHz, with the 915 MHz
design frequency being intermediate in both maximum SAR and longitudinal distribution.
The radial distribution is shown in Figure 4.5 in the plane of the antenna junction. The
distance shown included a geometrical correction to account for the antenna radius. Note
that the peak SAR measured transversely in Figure 4.5 is much larger than the SAR
measured along the longitudinal catheter. The explanation for this discrepancy is the fact
that the longitudinal catheter is mounted on the antenna, and is almost completely
surrounded by the low-conductivity catheter material; therefore, the SAR measurements
will be proportionally lower. In support of this, the radial and longitudinal data are similar
at 5 mm. The SAR data points would be expected to decay as
in the radial direction,
where a is the reciprocal of the penetration depth. When the radial SAR data were fit to this
102
function, the penetration depths were found to be 11.6 mm, 8.0 mm, and 4.3 mm for 433
MHz, 915 MHz, and 2450 MHz, respectively. These values are higher than the respective
penetration depths expected from bare antenna theory (5.9 mm, 5.2 mm, and 3.9 mm) but
generally lower than predicted from insulated antenna theory (11.5 mm, 10.8 mm, and 6.2
mm). Details of the comparison with theory are provided as an Appendix to this thesis.
#15 MHz
2450 MHz
——L
cm
1
1
Antenna
\
Tip
Figure 4.3. Liquid crystal temperature distributions, 15 seconds after power application to
represent a qualitative SAR distribution, a) 433 MHz, b) 915 MHz, and c) 2450 MHz.
104
SAR along transurethral catheter
50
2450 MHz
40915 MHz
20.
433 MHz
Section A
0
15 N
20
Section B
40
60
Distance (mm)
80
100
SAR 5mm from catheter, avg of 4 catheters
30-
2450 MHz
255: 20•
O)
V
1 5-
Antenna
Tip
915 MHz
CE 10.
<
(Ti
5-
060
80
100
Distance (mm)
Figure 4.4. a) Longitudinal SAR a determined along the thermometry channel of the
antenna, and b) longitudinal measurements parallel to the antenna at a distance of 5 mm.
Data are shown for 433 MHz, 915 MHz, and 2450 MHz.
105
200O
A
•
50-
433 MHz
915 MHz
2450 MHz
O)
00.
cc
,
M
50.
9
12
3
6
Distance from antenna surface (mm)
0
15
50
402450 MHz
O
433 MHz
A
915 MHz
°
2450 MHz
433 MHz
915 MHz
tr
<
CD
0
3
6
9
12
Distance from antenna surface (mm)
15
Figure 4.5. Radial SAR distribution in the plane of the antenna junction (a) and in a plane
2 cm away from the junction (b) for the three test frequencies. In the junction plane, 2450
MHz results in the largest peak SAR, but at 2 cm away from the junction, 433 MHz has the
highest SAR at the antenna surface, followed by 2450 MHz and finally 915 MHz.
106
In vivo experiments
The placement of the thermometry catheters in the dog prostate is shown
diagrammatically in Figure 4.6. Transurethral temperature distributions were obtained for
all three dogs at 915 MHz, and at 433 MHz and 2450 MHz for two of the dogs, and are
shown in Figure 4.7. The first dog had a small prostate, measuring 2 cm in length and 1.5
cm in diameter, so that the junction of the antenna was not contained within the prostatic
volume. The temperatures measured in the prostate were at a distance of approximately dL
= DR = 5 mm from the antenna surface, and were located ZL = 1.43 cm and ZR = 1.93 cm
from the junction along section A of the antenna. The second dog had an extremely small
prostate which was impossible to implant with temperature probes, so that only
intraurethral temperatures were measured in this animal. The third dog had a large prostate
measuring 4 cm long by 3.3 cm in diameter, and thermometry catheters were implanted in
the junction plane (ZL = ZR = 0) and at distances DL = 0.5 cm and DR = 1.1 cm.
Temperature distributions at 915 MHz within the prostate are shown in Figure 4.8(a)
for dog #1. This shows that temperatures of 41°C to 43°C are obtained in the prostate at a
longitudinal distance 1.5 to 2 cm away from the antenna junction, with a maximum
temperature of 47°C at the applicator surface. In Figure 4.8(b), the effect of antenna driving
frequency on the temperature distribution 1 cm away from the antenna surface is illustrated
in dog #3, and shows that frequency has relatively minor effect on the temperatures at this
distance. The temperature dip at 25 mm was a reproducible finding, and most likely due to
a thermally significant blood vessel or a local region of higher blood flow.
Histologically, the prostatic urethra showed foe ally extensive loss of the urothelium
with a generally intact underlying basement membrane. The urothelium in peripheral
urethral folds was generally intact. In some sections mild focal hemorrhage was seen
within the urethra in the urethral submucosa and in the prostate. The prostatic hemorrhage
was not more severe or consistent in the tissues nearer the urethra (which would have
107
received the highest thermal dose). One prostate (dog #2) showed focal degeneration of the
urethral submucosa and adjacent musculature in the tissue near the antenna junction.
Another prostate (dog #3), with spontaneous cystic hyperplasia had mild focal
intraglandular hemorrhage. There were also focal aggregates of degenerative lymphocytes,
neutrophils, hemosiderin containing macrophages, and red blood cells in the prostatic
urethral submucosa of this dog.
108
Balloon
Bladder
Antenna tip
Section A
d = distance from antenna catheter
surface to thermometry catheter
Z = distance from antenna junction to
thermometry catheter
Antenna Junction
Section B
Urethra
Figure 4.6. Schematic diagram of distances determined for temperature catheters on dog
prostate. The bladder, urethra, prostate, and catheter were removed en block to measure
these parameters.
109
915 MHz
Section A
Section B
100
Distance (mm)
433 MHz
Dog '2
Dog *3
Section A
Section B
40
too
60
Distance (mm)
50
2450 MHz
—
Dog *2
—
Dog *3
u
a>
ro
L.
a>
CL
E
<u
"
Section A
Section B
38
20
40
60
80
100
Distance (mm)
Figure 4.7. In vivo transurethral temperature distributions in dog, a) 915 MHz (design
frequency), b) 433 MHz, and c) 2450 MHz.
110
43
Dog #1
-
42-
Left catheter
Right catheter
u
O)
S-
% 40(U
Q.
E 39cu
I-
Prostate
38
0
5
10
15
20
25
30
35
40
Distance (mm)
Dog #3
915 MHz
433 MHz
2450 MHz
Prostate
T
20
30
40
Distance (mm)
Figure 4.8. In vivo temperature distributions in prostate: a) dog #1, dL = dR = 5 mm, ZL =
1.43 cm, and ZR = 1.93 cm. b) dog #3, right catheter, DR = 1.1 mm, ZR = 0.
Ill
4.5
DISCUSSION
Hyperthermia has been used in treating the prostate for cancer (Servadio and Leib,
1984; Szmigielski et al., 1988; Yerushalmi et al., 1982, 1986a,b), BPH (Lindner gf a/.,
1987; Sapozink gf a/., 1989; Servadio er a/., 1986,1987; Yerushalmi etal., 1985,1986a),
and chronic nonbacterial prostatitis (Servadio et ah, 1986, 1987). These studies have
demonstrated that hyperthermia may be a valuable nonsurgical treatment for obstructive
symptoms caused by these prostatic diseases, and that hyperthermia can be effectively
delivered to the prostate using either a transurethral or transrectal approach.
Lindner et al. (1987) treated 6 patients with symptomatic BPH requiring an indwelling
catheter. Previous attempts to wean these patients from their catheters had failed (following
a 5-day course of phenoxybenzamine). The patients were given 5 to 10 hyperthermia
treatments (1-2 treatments per week) using a microwave (915 MHz) water-cooled
transrectal applicator, and were followed for 6 months. Five of the six patients showed
subjective and objective improvement of their symptoms, and were able to be relieved of
their indwelling catheters. Four of these patients were weaned off their catheters one week
following the hyperthermia treatment course. The only treatment failure was a patient with a
large, tender prostate which could not be treated effectively. No complications were
observed in this study.
The largest series of patients undergoing prostate hyperthermia is currently under
study by two investigators in Israel. To date, Servadio and Yerushalmi have used a watercooled transrectal microwave applicator to deliver 500 hyperthermia treatments to 74
patients with benign and malignant prostatic diseases (Servadio et al., 1987). Yerushalmi et
al. (1985) summarize their results for the treatment of BPH. In this study, 29 patients with
severe symptoms and for whom surgery was contraindicated were given an average of 14
hyperthermia treatments on a twice weekly schedule. Eleven of the patients had chronic
indwelling catheters. The patients were followed for up to 26 months and evaluated using a
scoring scale for symptoms of frequency, nocturia, urgency, and hesitancy. The authors
112
noted that symptoms markedly improved after 6-8 treatments, but they felt that a total of
12-15 treatments was optimum. All of the patients who did not have an indwelling catheter
showed symptomatic improvement, and eight of eleven (73%) patients who had indwelling
catheters resumed normal voiding, with a post-void residual of <60 ml. At 18 months
follow-up, none of the previously catheterized patients had developed recurrence of severe
obstructive symptoms or urinary retention. No side effects were noted in this study, and
the transrectal hyperthermia was found to cause no damage to the rectal mucosa.
Astrahan and Sapozink et al. have developed a transurethral microwave hyperthermia
applicator (Astrahan etal., 1989a,b; Sapozink etal., 1989) for the treatment of BPH. This
applicator is also built around a Foley catheter, but uses 3 interstitial antennas mounted on
the periphery of the catheter. They have treated 21 men with a total of 177 hyperthermia
treatments and a mean follow-up of 10 months. In terms of objective response (residual
urine volume and urine flow rate), 17/21 (81%) of the patients showed improvement. In
terms of subjective parameters (frequency and stream force), 15/21 (71%) of the patients
showed improvement.
For treating BPH and for palliation of obstructive symptoms, the transurethral
approach would have the advantage of concentrating the thermal dose in the area
immediately surrounding the constriction. On the other hand, the transurethral antenna
would not be able to produce therapeutic temperatures to cover large lesions, such as
prostatic carcinoma. To treat these diseases, one or two transrectal antennas could be used
in combination with the transurethral antenna to form a linear or triangular array.
From a design standpoint, the applicator resembles an enlarged version of the insulated
interstitial microwave antennas that have been used by our group and others. Because of its
larger dimensions, the theoretical description of this antenna is not straightforward. For
insulated antenna theory to be applicable, the antenna must be electrically thin in both the
catheter dielectric and the surrounding tissue. Mathematically, the required conditions are
(King etal., 1981):
113
and
I k2 c 1^ « 1
(1)
I k4 c I < 0.5
(2).
where c is the outer radius of the antenna and k2 and k4 are wavenumbers in the catheter
and tissue, respectively:
Here, f is frequency and |i, e, and a are the permeability, permittivity, and conductivity of
the medium. Whereas the first condition is met at the frequencies of interest, the second
condition is only marginally satisfied at 915 MHz and not satisfied at 2450 MHz. Another
requirement of the insulated antenna theory is that the insulation must be sufficiently thick.
Specifically, the requirement is that the ratio c/a > 2, where c and a are the outer and inner
radii of the antenna insulation layer, respectively. In the case of this antenna, the ratio is
only 1.14. In summary, we cannot accurately describe the behavior of this antenna using
insulated antenna theory because of its electrically large dimensions and the fact that it is not
adequately insulated. Nevertheless, we would expect the general behavior of the antenna to
be similar to that of a typical insulated antenna, and our experimental results demonstrate
that this is the case.
Although the antenna was designed for operation at 915 MHz, reflected power was
also relatively low at 433 MHz and 2450 MHz. Both quantitative and qualitative SAR
measurements demonstrated a typical elliptical pattern centered at the antenna junction. At
433 MHz, the longitudinal distribution was uniform, but the radiation radially was very
poor, as would be expected since the antenna section is much less than a quarter
wavelength at this frequency. At 2450 MHz, the longitudinal extent of the SAR pattern was
shortened and the peak SAR at the junction increased. This general result would be
expected, although the peak SAR measured at 2450 MHz was three times higher than that
measured at 915 MHz.
114
The antenna design has an advantage over the conventional interstitial dipole antennas,
because section B is separated from the feedline by a dielectric layer serving as a choke. If
this choke operates perfectly, then no current flows from section B of the antenna to the
feedline. However, this choke is not perfect, because the section length of 3 cm
corresponds to a choke operating frequency of about 1800 MHz (where 3 cm is a quarterwavelength in the choke dielectric layer). Impedance measurements showed that there was
some variation with insertion depth at 433 MHz, suggesting that section B is not
completely isolated from the feedline, and that antenna performance will be a function of
insertion depth at this frequency. On the other hand, the impedances measured at 915 MHz
and 2450 MHz were constant at all insertion depths. The choke operates better at these
frequencies, which are closer its resonant frequency.
Although the antenna design was optimized for operation at 915 MHz, SAR
measurements demonstrate that the choked antenna design is able to confine the antenna
radiation to the prostatic tissue and may permit other frequencies to be used. SAR
measurements demonstrate that the antenna does not radiate efficiently at 433 MHz, and
impedance measurements suggest that antenna performance varies with insertion depth at
this frequency. The antenna sections are much shorter than a quarter-wavelength at this
frequency; for the antenna to function well at 433 MHz, the sections A and B would need
to be much longer (~5 to 6 cm). Nonetheless, this frequency and 2450 MHz may be useful
in certain situations. For example, 2450 MHz may be useful for treating smaller lesions
(shorter section lengths could be used) and 433 MHz yields a more uniform periurethral
SAR distribution along the entire length of the antenna. The present antenna was designed
for operation at 915 MHz, but the design can be optimized for other frequencies by
adjusting the antenna section lengths to correspond to the resonant quarter wavelength.
In vivo studies demonstrated that the antenna could effectively heat the prostate to
therapeutic temperatures (> 43°C) at 1 cm from the antenna surface when the maximum
urethral temperature was 47°C to 48°C. The longitudinal temperature distributions were
115
consistent with the SAR measurements made at each frequency along the antenna. These
temperature measurements confirmed that the heating pattern was confined to the length of
the antenna elements. With the maximum temperature maintained at 47°C to 48°C, the
average length heated along the urethra to > 43°C was 62 mm, 54 mm, and 41 mm for 433
MHz, 915 MHz, and 2450 MHz, respectively. It was interesting that the temperature
distributions in the prostate were very similar at all three frequencies, in view of the fact
that the calculated penetration depths were very different at the three frequencies. This
suggests that the actual radial temperature distribution in tissue may not be a very strong
function of frequency, and that the frequency could be selected based solely on the
longitudinal treatment requirements.
Overall, the tissue effects from hyperthermia treatments were focal, mild and confined
primarily to the tissue immediately adjacent to the urethra. The loss of the urothelium is
most consistent with mechanical denudation, since the urothelium in urethral folds which
was not in contact with the antenna was generally unaffected. The mild focal hemorrhage
observed in the prostatic parenchyma was likely due to palpation or physical manipulation
of the prostate during the procedure, since the hemorrhage was not more severe near the
antenna (region of highest thermal dose). None of the observed tissue effects seen at this
time period and thermal dose appears to be dose limiting for prostatic hyperthermia.
An advantage of the coaxial dipole design is that fact that the treatment volume can be
greatly increased by using arrays. One possibility would be to add a rectal applicator to treat
bulky prostates or prostatic cancer. Once such an antenna array is defined, it is possible to
control the temperature distribution within the array through manipulation of amplitude and
phase (Trembly et al., 1986, 1988, 1990b) The volume heated can be further increased
through the use of air cooling (Eppert et al., 1990; Tremby et al., 1990a).
Like some other transurethral applicator designs, the antenna is built around a Foley
catheter. The Foley balloon ensures accurate and reproducible placement in the prostatic
urethra, and may obviate the need for radiographic confirmation of antenna position. We
116
found that the 12 Fr size to be the most appropriate size, because the antenna elements and
thermometry catheter add to the overall outer diameter. Based on our experience with the
dog model, the smaller Foley catheters were too flexible and therefore more difficult to
guide through the urethra.
117
4.6
CONCLUSIONS
A new transurethral microwave applicator for hyperthermia has been described. The
antenna is designed as a coaxial dipole and incorporates a choke structure which allows the
antenna to behave independently of insertion depth at the design frequency of 915 MHz.
The insulated antenna theory is not adequate for describing this antenna because of its
relatively large size and relatively thin insulation layer. Impedance measurements
demonstrated that the antenna could operate at 433 MHz, 915 MHz, and 2450 MHz with
low reflected power, and that impedance was independent of insertion depth at the two
higher frequencies. SAR measurements show good antenna performance at 915 MHz, and
a narrower longitudinal profile at 2450 MHz. The antenna delivered low SARs at 433
MHz, and its performance at this frequency may change with insertion depth. We would
expect improved performance at 433 MHz if longer section lengths were used.
The in vivo temperature measurements demonstrated that therapeutic temperatures
could be obtained in the prostate at a distance of at least 1 cm from the antenna surface with
the maximum urethral temperature held at 47°C to 48°C. Histologically, these temperatures
resulted in minimal acute damage. Although long-term effects were not evaluated in this
study, the basement membrane remained essentially intact, and no irreversible changes
would be expected to result from these treatments. Clinically, it may be possible to use
higher temperatures, because the antenna would lie in the tissue which would ordinarily be
removed by transurethral resection. Temperature distributions in the prostate showed only
slight changes with frequency. This suggests that the frequency can be chosen based
mainly on the desired treatment length. Antennas could be designed having the appropriate
section lengths. The 915 MHz design is useful for heating lengths of 4 to 6 cm. We would
expect this length to be approximately doubled at 433 MHz and halved at 2450 MHz.
Currently, plans are being made to improve the catheter design by developing ways to
make the catheter surface more smooth and to restore the drainage function of the catheter
(the existing channel is used to house the antenna feedline). A clinical trial has been initiated
118
at our institution for transurethral hyperthermia for BPH, and our ultimate goal is to
develop transrectal antennas to be used in combination with this transurethral applicator for
treating bulky prostatic cancer.
CONCLUSIONS
119
120
5.1 TRANSIENT TEMPERATURE MEASUREMENTS
In Chapter 2, a two-dimensional time-dependent finite element model was used to
predict the accuracy of using transient temperature data taken after power-on and power-off
to estimate SAR and blood flow, respectively. In this simulation, a 2 cm square 915 MHz
interstitial antenna array was used to heat a 4 cm diameter tumor, and the Pennes bioheat
transfer equation was used to calculate the resulting temperature distribution.
Representative blood flow values were assigned within the tumor, and the applied SAR
distribution was based on insulated antenna theory. SAR values were estimated from the
power-on transient temperatures, and blood flow values were estimated using a single time
constant decay model to calculate blood flow (i.e. ignoring thermal conduction). SAR and
blood flow predictions were then compared to the known "true" values throughout the
treatment region. The conclusions from this study were:
1) Thermal clearance can yield reasonably accurate estimates of blood flow only if the
blood flow is relatively high (> 40 ml/100g*min) and the regional perfusion is
homogeneous.
2) Thermal clearance estimates were generally poor in the nonhomogeneously perfused
model, and the technique is not able to resolve local differences in blood flow.
3) The thermal decay after power-off would be expected to have one time constant if
thermal conduction is negligible. However, this was found to be a necessary but not
sufficient condition for accurate estimation of blood flow.
4) SARs could be estimated accurately during the power-on phase of heating, even in
the presence of relatively high blood flow (150 ml/100g*min). However, it should
be noted that this model was dielectrically homogeneous.
Although this paper was not intended to pertain to intracavitary hyperthermia, the
overall conclusion is applicable. One of the major problems of intracavitary hyperthermia is
adequate thermometry. Often, the only practical location is along the applicator itself. If it is
possible to estimate the blood flow near the applicator, then a better prediction could be
121
made regarding the radial temperatures away from the applicator. Unfortunately, these
results show that a more sophisticated thermal model is apparently necessary.
5.2 INVESTIGATION OF A MICROWAVE CHOKE
One problem with the interstitial microwave system presently used for hyperthermia is
the fact that the performance of the antennas changes with insertion depth. A possible
solution to this problem is a choke. However, the practical choke length does not
correspond to the desired resonant quarter-wavelength of the insulated antenna, and may
not operate adequately at the driving frequency of the antenna. This problem was
investigated in Chapter 3 by comparing several antennas having various choke dimensions
and observing the performance of the choke in terms of changing the variation of
impedance with insertion depth over a range of frequencies. The results of this work are the
following:
1) An automated impedance measurement system has been developed which enables
antennas to be tested conveniendy and rapidly over a wide range of frequencies.
2) A model based on insulated antenna theory has been developed for calculating the
impedance of a choked antenna. This should be particularly useful for the design of
improved interstitial antennas.
3) Based on theoretical considerations and experimental results, the following
recommendations can be made to improve the choke design: 1) find a more ideal
dielectric material (relative permittivity of 6 to 7), 2) use a small diameter feedline
with relatively thick insulation, and 3) a thick choke layer improves performance.
5.3 DESIGN OF A TRANSURETHRAL APPLICATOR
A coaxial microwave antenna has been designed for transurethral hyperthermia of the
prostate to treat BPH (Chapter 4). This application has many advantages from the
standpoint of applicator design. The maximum temperatures, which are measured at the
122
applicator surface, are located in the tissue to be treated. There is little risk of overheating
the prostatic urethra, since this is the tissue that would ordinarily be removed surgically.
The most important design consideration is the limitation of power deposition to the
prostatic urethra. Anatomically, the bladder provides a reference point adjacent to the
prostate, which allows the applicator to be positioned with a foley balloon. Tests on the
applicator have led to the following conclusions:
1) The foley catheter provides a choke layer which confines power depostion to the
length of the antenna elements. In-vivo studies demonstrate that therapeutic
temperatures can be obtained over distances greater than 1 cm from the antenna
surface in the prostate while causing minimal acute histological damage. Although
the applicator is intended to treat BPH, the heating volume could be extended
posteriorly by the addition of a transrectal applicator for treating cancer.
2) The size of the antenna and relatively thin insulation layer make theoretical
characterization of this antenna difficult. Comparison of transverse SAR
measurements with theoretical predictions suggest that the antenna has a penetration
depth intermediate between that predicted by bare antenna theory and insulated
antenna theory.
5.4 FUTURE DIRECTIONS
The major goal of this work has been to investigate several aspects of the coaxial
microwave antenna, and to extend its application to intracavitary hyperthermia. Design
considerations have been suggested for adding a choke to interstitial antennas; however,
SAR experiments will be needed to determine the extent to which impedance changes
influence antenna performance. It would be worh searching for suitable higher dielectric
materials for the choke, or exploring ways of increasing the dielectric constant of the
present material.
123
A clinical protocol for treating BPH using the transurethral applicator has been
approved at the Dartmouth-Hitchcock Medical Center (Appendix 6). In addition, it will be
worth establishing a relationship with a catheter company to refine the present design.
Improvements would include adding a urinary drainage channel, embedding the antenna
elements in the catheter wall, designing a disposable catheter, and improving the attachment
of the thermometry catheter. A rectal antenna (air or water cooled) design would be a next
logical step to be added to the transurethral antenna to treat prostate cancer.
The transurethral antenna is approximately 5 mm in diameter, and intracavitary
antennas could be constructed for other sites (e.g. esophagus) using a very similar design.
Finally, the concept of using hyperthermia to treat a very common benign disease has been
introduced. This may have a significant impact on the future of this field.
REFERENCES
124
125
RF .FRRF.NrFS
Astrahan MA, Sapozink MD, Cohen D, Luxton G, Kampp TD, Boyd S, and Petrovich Z,
1989a, Microwave applicator for transurethral hyperthermia of benign prostatic
hyperplasia. Int. J. Hyperthermia 5(3): 283-296.
Astrahan MA, Sapozink MD, Jozsef G, Cohen D, Luxton G, Boyd S, Kampp TD, and
Petrovich Z., 1989b, Microwave applicators for transurethral hyperthermia of benign
prostatic hyperplasia. 9th Annual Meeting, North American Hyperthermia Group, March
18-23, (Abstract).
Beyer WH, 1980, Handbook of Mathematical Sciences, 5th ed. CRC Press, Boca Raton, p.
731.
Bowman HF, 1982, Thermodynamics of tissue heating; modeling and measurements for
temperature distributions. Physical Aspects of Hyperthermia, edited by P.H. Nussbaum
(New York: American Institute of Physics), pp. 511-548.
Brezovich lA, Atkinson WJ, and Chakraborty DP, 1984, Temperature distributions in tumor
models heated by self-regulating nickel-copper alloy thermoseeds. Medical Physics, 11(2),
145-152.
Brezovich lA, Young JH, and Wang MT, 1983, Temperature distributions in hyperthermia
by electromagnetic induction: A theoretical model for the thorax. Medical Physics, 10 (1),
57-65.
Broschat SL, Chou CK, Luk KH, Guy AW, and Ishimaru A., 1988, An insulated dipole
applicator for intracavitary hyperthermia. IEEE Trans. Biomed. Engin. 35(3): 173-178.
Castaneda F, Lund G, Larson BW, Limas E, Umess M, Reddy P, Wasserman N, Hulbert
J, Hunter DW, Castaneda-Zuniga WR, and Amplatz K., 1987, Prostatic urethra:
experimental dilatation in dogs. Radiology 163: 645-648.
Chan KW, Chou CK, McDougall JA, Luk KH, Vora NL, and Forell BW., 1989,
Changes in heating patterns of interstitial microwave antenna arrays at different insertion
depths. International Journal of Hyperthermia 5(4): 499-507.
Clegg ST, Roemer RB, and Cetas TC, 1985, Estimation of complete temperature fields from
measured transient temperatures. International Journal of Hyperthermia, 1 (3), 265-286.
Coughlin CT, Douple EB, Strohbehn JW, Eaton WL, Trembly BS, and Wong TZ., 1983,
Interstitial hyperthermia in combination with brachytherapy. Radiology 148(1): 285-288.
Coughlin CT and Wong TZ, 1986, IMMAH-induced hyperthermia for the treatment of
biliary cancers. 6th Annual Meeting, North American Hyperthermia Group, Las Vegas,
NV, April 12-17 (Abstract).
Coughlin CT, Wong TZ, Strohbehn JW, Colacchio TA, Sutton JE, Belch RZ, and Douple
EB, 1985, Intra-operative interstitial microwave-induced hyperthermia and brachytherapy.
International Journal of Radiation Oncology, Biology, Physics, 11,1673-1678.
126
De Sieyes DC, Douple EB, Strohbehn JW, and Trembly BS, 1981, Some aspects of
optimization of an invasive microwave antenna for local hyperthermia treatment of cancer.
Medical Physics 8:179-183.
Divrik AM, Roemer RB, and Cetas TC, 1984, Inference of complete tissue temperature
fields from a few measured temperatures: an unconstrained optimization method. IEEE
Transactions on Biomedical Engineering, BME-31,150-160.
Eppert VE, Trembly BS, Richter HJ, 1990, Air cooling for an interstitial microwave
antenna: theory and experiment. IEEE Trans. Biomed. Engin., in press.
Geller J, 1989, Overview of benign prostatic hypertrophy. Urology (suppl.) 34(4): 5763.
Guy AW, 1978, Phantom models for muscle, brain, fat, and bone at microwave
frequencies. Written communication.
Guy AW, Lehmann JF, and Stonebridge JB, 1974, Therapeutic applications of
electromagnetic power. Proceedings IEEE, 62, 55-75.
Hartsgrove G, Kraszewski A, and Surowiec A., 1987, Simulated biological materials for
electromagnetic absorption studies. Bioelectromagnetics 8: 29-36.
James BJ, Strohbehn JW, Mechling JA, and Trembly BS, 1989, The effect of insertion
depth on the theoretical SAR patterns of 915 MHz dipole antenna arrays for hyperthermia.
International Journal of Hyperthermia 5(6): 733-747.
Jones KM, Mechling JA, Trembly BS, and Strohbehn JW, 1988, SAR distributions for
915 MHz interstitial microwave antennas used in hyperthermia for cancer therapy. IEEE
Trans. Biomed Engineering 35(10): 851-857.
Jones KM, Wong TZ, Mechling JA, Trembly BS, and Strohbehn JW, 1986, SAR
patterns from an interstitial microwave antenna array hyperthermia system. Proc. I2th
Annual Northeast Bioengineering Conf., Orphanoudakis SC, ed., IEEE, New York, pp.
209-212.
King RWP and Harrison CW, 1969, Antennas and Waves: A Modern Approach, MIT
Press, Cambridge MA.
King RWP, Shen LC, and Wu TT, 1981, Embedded insulated antennas for
communication and heating. Electromagnetics 10: 551-572.
King RWP, Trembly BS, and Strohbehn JW, 1983, The electromagnetic field of an insulated
antenna in a conducting or dielectric medium. IEEE Transactions on Microwave Theory and
Techniques, MTT-31, 574-583.
Lepor H, 1989, Nonoperative management of benign prostatic hyperplasia. J. Urology
141: 1283-1289.
Leybovich L, Emami B, Straube W, Schlessinger E, Nussbaum GH, Devineni DVR, and
Sessions D, 1987, Intracavitary hyperthermia: a newly designed applicator for tracheal
tumors. Endocurie. Hypertherm. Oncol. 3: 23-29.
127
Li DJ, Luk KH, Jiang HB, Chou CK, and Hwang GZ, 1984, Design and thermometry of
an intracavitary microwave applicator suitable for treatment of some vaginal and rectal
cancers, nternational Journal of Radiation Oncology, Biology, Physics 10: 2155-2162
(1984).
Lindner A, Golomb J, Siegel Y, and Lev A, 1987, Local hyperthermia of the prostate
gland for the treatment of benign prostatic hypertrophy and urinary retention- a preliminary
report. British Journal of Urology 60: 567-571.
Luk KH, Jiang HB, and Chou CK, 1985, SAR patterns of a helical microwave
intracavitary applicator, in Hyperthermic Oncology 1984, vol. 1, Overgaard J, ed., pp.
591-594.
Lyons BE, Samulski TV, Cox RS, and Fessenden P, 1989, Heat loss and blood flow during
hyperthermia in normal canine brain. I: Empirical study and analysis. International Journal
of Hyperthermia 5(2): 225-247.
Mechling JA, 1989, Treatment planning software for the Dartmouth IMAAH system,
Ph.D. Thesis, Dartmouth College, Hanover, NH.
Mechling JA and Strohbehn JW, 1986, A theoretical comparison of the temperature
distributions produced by three interstitial hyperthermia systems. International Journal of
Radiation Oncology, Biology, Physics, 12: 2137-2149.
Mendecki J, Friedenthal E, Botstein C, Paglione R, and Sterzer F, 1980, Microwave
applicators for localized hyperthermia treatment of cancer of the prostate. International
Journal of Radiation Oncology, Biology, Physics. 6(11): 1583-1588.
Milligan AJ and Panjehpour M, 1985, Canine normal and tissue estimated blood flow during
fractionated hyperthermia. International Journal of Radiation Oncology, Biology, Physics,
11, 1679-1684.
Milligan AJ, Conran PB, Ropar MA, McCulloch HA, Ahuja RK, and Dobelbower RR,
1983, Predictions of blood flow from thermal clearance during regional hyperthermia.
International Journal of Radiation Oncology, Biology, Physics, 9: 1335-1343.
Paulsen KD, Strohbehn JW, Hill SC, Lynch DR, and Kennedy FE, 1984, Theoretical
temperature profiles for concentric coil induction heating devices in a two-dimensional, axiasymmetric, inhomogeneous patient model. International Journal of Radiation Oncology,
Biology, Physics, 10, 1095-1107.
Pence DM, Song CW, 1986, Effects of heat on blood flow. In Hyperthermia in Cancer
Treatment, vol. II, Anghileri LJ and Robert J, eds. CRC Press, Boca Raton, pp. 1-16.
Petrowicz O, Scheiblich J, Crucius A, Erhardt W, Wriedt-Liibbe I, Ultsch B, and BlUmel
G, 1982, High-frequency transmitter for the localized heat treatment of the prostate gland.
Natl. Cancer Inst. Monogr. 61: 473-476.
Robbins SL, Cotran RS, and Kumar V, 1984, Pathologic Basis of Disease, 3rd ed.,
W.B. Saunders, Philadelphia, pp. 1101-1104.
Roberts DW, Coughlin CT, Wong TZ, Fratkin JD, Douple EB, and Strohbehn JW, 1986,
Interstitial hyperthermia and iridium brachytherapy in the treatment of malignant glioma: a
phase I clinical trial. Journal of Neurosurgery 64: 581-587 (1986).
128
Roemer R., Fletcher AM, and Cetas TC, 1985, Obtaining local SAR and blood perfusion
data from temperature measurements: steady state and transient techniques compared.
International Journal of Radiation Oncology, Biology, Physics, 11,1539-1550.
Roos DIF, 1988, Interstitial and intracavitary microwave applicators for hyperthermia
treatment of cancer, Ph.D. Thesis, Chalmers Institute of Technology, Goteborg, Sweden.
Rubin P, and Casarett G, 1966, Microcirculation of tumors, part I: anatomy, function, and
necrosis. Clinical Radiology, 17,220-229.
Ryan TP, Wikoff RP, and Hoopes PJ, 1990, An automated temperature mapping system
for use in ultrasound or microwave hyperthermia. International Journal of Hypertherrriia,
submitted.
Ryan TP, 1990, Comparison of commercially available microwave antennas for
hyperthermia treatment of cancer: SAR results for single antennas and arrays. International
Journal of Hyperthermia, submitted.
Sagalowsky AI and Wilson JD, in Braunwald E, Isseibacher KJ, Petersdorf RG, Wilson
JD, Martin JB, and Fauci AS, eds., 1987, Harrison's Principles of Internal Medicine,
11th ed., McGraw-Hill, New York, pp. 1582-1585.
Samulski TV, Cox RS, Lyons BE, and Fessenden P, 1989, Heat loss and blood flow during
hyperthermia in normal canine brain. II: Mathematical model. International Journal of
Hyperthermia 5(2): 249-263.
Sandhu TS, 1986, Measurement of blood flow using temperature decay: effect of thermal
conduction. International Journal of Radiation Oncology, Biology, Physics, 12,373-378.
Santhiaseelan V, Iskander MF, Howard GCW, and Bleehen NM, 1986, Theoretical analysis
and clinical demonstration of the effect of power pattern control using the annular phasedarray hyperthermia system. IEEE Transactions on Microwave Theory and Techniques,
MTT-34 (5), 514-519.
Sapozink M, Boyd S, Astrahan M, Jozsef G, Nourparvar S, and Petrovich Z, 1989,
Transurethral hyperthermia for benign prostatic hyperplasia - preliminary results. 9th
Annual Meeting, North American Hyperthermia Group, March 18-23, (ABSTRACT).
Scheiblich, J. and Petrowicz, O., 1982, Radiofrequency-induced hyperthermia in the
prostate. Journal of Microwave Power 17(3): 203-209.
Schwan HP and Foster KR, 1980, RF-field interactions with biological systems: electrical
properties and biophysical mechanisms. Proc. IEEE 68(1): 104-113.
Servadio C, and Leib Z, 1984, Hyperthermia in the treatment of prostate cancer. Prostate
5(2): 205-211.
Servadio C, Leib Z, and Lev A, 1987, Diseases of prostate treated by local microwave
hyperthermia. Urology 30(2): 97-99.
Servadio C, Leib Z, and Lev A, 1986, Further observations on the use of local
hyperthermia for the treatment of diseases of the prostate in man. European Urology 12:
38-40.
129
Strohbehn JW, and Mechling JA, 1986, Interstitial techniques for clinical hyperthermia.
Physical Techniques in Clinical Hyperthermia, Hand JW and James JR, eds.. Research
Studies Press Ltd., Letchworth, pp. 210-287.
Strohbehn JW, and Roemer RB, 1984, A survey of computer simulations of hyperthermia
treatments. IEEE Transactions on Biomedical Engineering, BME-31,136-149.
Strohbehn JW, Trembly BS, and Douple EB, 1982, Blood flow effects on the temperature
distributions from an invasive microwave antenna array used in cancer therapy. IEEE
Transactions on Biomedical Engineering, BME-29,649-661.
Szmigielski S, Zielinski H, Stawarz B, Gil J, Sobczynski J, Sokolska G, Jeljaszewicz J,
and Pulverer G, 1988, Local microwave hyperthermia in treatment of advanced prostatic
adenocarcinoma. Urology Research 16(1): 1-7.
Trembly BS, Douple EB, and Hoopes PJ, 1990a, The effect of air cooling on the radial
temperature distribution of a single microwave hyperthermia antenna in vivo. International
Journal of Hyperthermia, submitted.
Trembly BS, Douple EB, Ryan TP, and Hoopes PJ, 1990b, The effect of phase
modulation on the temperature distribution of a microwave hyperthermia antenna array in
vivo. International Journal of Hyperthermia, submitted.
Trembly BS, Wilson AH, Havard JM, Sabatakakis KG, and Strohbehn JW, 1988,
Comparison of power deposition by in-phase 433 MHz and phase-modulated 915 MHz
interstitial antenna array hyperthermia systems. IEEE Trans. Microwave Theory Techn.,
36(5): 908-916.
Trembly BS, Wilson AH, Sullivan MJ, Stein AD, Wong TZ, and Strohbehn JW, 1986,
Control of the SAR pattern within an interstitial microwave antenna array through variation
of antenna driving phase. IEEE Trans. Microwave Theory Techn. MTT-34(5): 568-571.
Trembly BS, 1982, The electric field of an insulated, linear antenna embedded in an
electrically dense medium, Ph.D. thesis, Dartmouth College, Hanover, NH.
Trembly BS, 1985, The effects of driving frequency and antenna length on power deposition
within a microwave antenna array used for hyperthermia. IEEE Transactions on Biomedical
Engineering, BME-32(2), 152-157.
Tumeh AM and Iskander MP, 1989, Performance comparison of available interstitial
antennas for microwave hyperthermia. IEEE Trans. Microw. Theor. Tech. 37(7): 11261133.
Urano M and Douple EB, eds., 1988, Hyperthermia and Oncology, vol. 1: Thermal effects
on cells and tissues, VSP, Ultrecht, the Netherlands.
Urano M and Douple EB, eds., 1989, Hyperthermia and Oncology, vol. 2: Biology of
thermal potentiation of radiotherapy, VSP, Ultrecht, the Netherlands.
van den Berg PM, de Hoop AT, Segal A, Praagman N, 1983, A computational model of the
electromagnetic heating of biological tissue with application to hyperthermic cancer therapy.
IEEE Transactions on Biomedical Engineering, BME-30 (12), 797-805.
130
Waterman FM, Nerlinger RE, Moylan DJ, and Leeper DB, 1987, Response of human tumor
blood flow to local hyperthermia. International Journal of Radiation Oncology, Biology,
Physics, 13, 75-82.
Wong TZ, Mechling JA, Jones EL, and Strohbehn JW, 1988, Transient finite element
analysis of thermal methods used to estimate SAR and blood flow in homogeneously and
nonhomogeneously perfused tumour models. InternationalJournal of Hyperthermia 4(6);
571-592.
Wong TZ and Trembly BS, 1990, Linear coaxial microwave antennas with choke for
hyperthermia. International Journal of Hyperthermia, submitted.
Wong TZ, Strohbehn JW, and Douple EB, 1985, Automated measurement of power
deposition patterns from interstitial microwave antennas used in hyperthermia. Proceedings
of the 11th Annual Northeast Bioengineering Conference, edited by W.S. Kuklinski and
W.J. Ohley (New York: IEEE), pp. 58-61.
Wong TZ, Strohbehn JW, Jones KM, Mechling JA, and Trembly BS, 1986, SAR patterns
from an interstitial microwave antenna array hyperthermia system. IEEE Transactions on
Microwave Theory and Techniques, MTT-34, 560-567.
Yerushalmi A, Fishelovitz Y, Singer D, Reiner I, Arielly J, Abramovici Y, Catsenelson R,
Levy E, and Shani A, 1985, Localized deep microwave hyperthermia in the treatment of
poor operative risk patients with benign prostatic hyperplasia. Journal of Urology 133:
873-876.
Yerushalmi A, Servadio C, Leib Z, Fishelovitz Y, Rokowsky E, and Stein JA, 1982,
Local hyperthermia for treatment of carcinoma of the prostate: a preliminary report. Prostate
3(6): 623-30.
Yerushalmi A, Shani A, Fishelovitz Y, Arielly J, Singer D, Levy E, Katsnelson R,
Rakowsky E, and Stein JA, 1986b, Local microwave hyperthermia in the treatment of
carcinoma of the prostate. Oncology 43:299-305.
Yerushalmi A, 1986a, Hyperthermia treatment of prostate tumors. In Hyperthermia in
Cancer Treatment, vol. Ill, Anghileri LJ and Robert J, eds. CRC Press, Boca Raton, pp.
89-101.
Zienkiewicz OC, 1971, The Finite Element Method in Engineering Science. (New York:
McGraw Hill).
7.0 APPENDICES
131
132
ANTENNA RG58-2
DATE OF CONSTRUCTION: June 13,1989
DESCRIPTION: Antenna made from RG-58 coax for greater diameter. It is intended to
be a larger unchoked antenna having a diameter more similar to that of the choked
antennas.
MATERIALS:
RG-58 coax cable with jacket
3/16" heat shrink tubing (Radio Shack, Heatrax #8514)
SMA male connector (larger than our usual, Omni Spectra 2031-5002-00)
ASSEMBLY TECHNIOUE: Began with 59.5cm of RG-58 cable, from which 15cm of
jacket was removed. The junction end of the cable was prepared by stripping both the
braid and dielectric to expose approximately 4mm of inner conductor. The braid was
trilnmed and smoothed to be flush with the end of the dielectric. This end of the braid
(junction region of section B) was soldered circumferentially with a thin layer of solder
for stability and strength, and sanded with #600 paper. The inner conductor was tinned,
and section A prepared from an excess length of RG-58 (with jacket removed). Section A
was soldered to the inner conductor in the usual manner, and the junction region
smoothed using the #600 sandpaper. The antenna was cleaned with an alcohol wipe and
covered with a 21cm length of 3/16" heat shrink tubing. A glob of silicone sealant was
squeezed into the tip to make it air-tight before heating the shrink tubing. Finally, a new
SMA connector (larger version) was attached to the feedline.
MEASURED DIMENSIONS:
Section A
Section B
Length
Overall O.D.
Braid (element) O.D.
3.2 cm
0.164" (4.17mm)
0.139" (3.53mm)
0.164" (4.17mm)
0.139" (3.53mm)
Junction to connector:
Junction size:
58.5cm
1mm
133
Antenna RG174-2
CONSTRUCTION DATE: June 16, 1989
DESCRIPTION: Choked antenna using RG174 coax feedline for comparison with
antennas #12 and RG58-2. The choke dielectric consists of a single layer of heat
shrink, compared to four layers used in antenna #12 and no choke in antenna RG58-2.
Section lengths of all three antennas were intended to be 3.2cm.
MATERIALS:
RG-174 coaxial cable (with jacket), standard (small size) SMA connector (previously
unused).
Heatrax heat shrink tubing: 3/32" (#8512), 3/16" (#8514)
Braid: 24-7-44, .084" nom. O.D., New England Electric Wire (from Stu's 4conductor cable).
ASSEMBLY TECHNIQUE: Began with 60cm of RG174 cable, from which 7.4cm of
jacket was removed. The junction end of the feedline was prepared by stripping the
coax to expose 6mm of inner conductor and about 1mm of dielectric. The inner
conductor was tinned. The feedline braid was cleaned with an alcohol wipe, and
6.55cm of 3/32" heat shrink tubing was applied over this shield to form the choke
dielectric layer. The heat shrink was very slightly short, so that there was a small
(0.5mm) gap between the remaining feedline jacket and the proximal end of the heat
shrink layer. This layer was wiped clean with an alcohol rub, and the 24-7-44 braid
stretched over the choke dielectric and trimmed to the desired length. The choke section
was soldered to the feedline braid circumferentially in the usual manner. Section A was
prepared using a jacket-less piece of RG-174 cable and adding the same heat shrink
layer as used in section B. The inner conductor of the feeline was trimmed, allowing a
"hook" to be formed for a solid connection to section A. The junction end of section A
was tapered, allowing a symmetrical "hemispherical" solder joint to the feedline inner
conductor. Section A was then attached by circumferentially soldering the connection.
The antenna was cleaned with an alcohol wipe, and covered with 18.3cm of 3/16" heat
shrink tubing. A new SMA connector (the same size as used for RG178 feedlines) was
attached in the usual manner.
MEASURED DIMENSIONS:
Section length
Overall antenna O.D.
Antenna element (braid) O.D.
Junction size: 1mm
Connector to junction: 59.3cm
Section A
Section B
3.2cm
0.143" (3.63mm)
0.112" (2.84mm)
3.3cm
0.144" (3.66mm)
0.113" (2.87mm)
134
Antenna RG174-3
CONSTRUCTION DATE: October 8,1989
DESCRIPTION: Choked antenna similar to RG174-2, except that the choke length is
designed based on the wavelength in the choke dielectric. Assuming that the relative
dielectric constant of the choke dielectric is Er = 1.7, the velocity c in this medium will be
divided by the square root of Er. Thus, the quarter wavelength in the choke is
approximately 6.3 cm at 915 MHz, and this is the target length for section hg. As with
antenna RG174-2, the choke layer is a single layer of Heatrax heat-shrink tubing.
Cvacuum
MATERIALS:
RG174 coaxial cable (with jacket), standard SMA connector (small, Omni Spectra 20315003-00, previously unused), Heatrax heat shrink tubing: 3/32" (#8512), 3/16 (#8514),
Braid: New England Electric Wire 24-7-44, .084" nom OD, from Stu's 4-cond cable.
ASSEMBLY TECHNIOUE: Began with excess RG174 cable, 76cm. 13.5cm of jacket
was removed using a scalpel. The exposed braid was cleaned with an alcohol wipe, then a
13cm length of 3/32" heatshrink was applied over it. In retrospect, this did not leave
enough exposed end to strip the coax and expose the inner conductor, so 3mm of the heat
shrink was carefully removed using strippers and scalpel. The jet end of the cable was
prepared by exposing 6mm of inner conductor, 1mm of dielectric, and 1mm of cable
shield. The exposed shield and inner conductor were tinned. The heatshrink on the antenna
was marked at 6.3cm from the exposed braid to mark the limit of the choke, and an excess
of the New England braid slipped over the heat shrink to this point to form section B. This
was trimmed to the desired length and soldered to the feedline braid.
Section A was prepared by removing the jacket from another length of RG174 cable and
heat shrinking an excess of 3/32" tubing. This was covered by an excess length of New
England braid and soldered to the inner conductor of the feedline. Before soldering on
section A, it was necessary to trim the exposed inner conductor to 4mm. After connection,
section A was trimmed to 3.2cm.
The antenna was covered with approximately 25cm of 3/16" heat shrink. Before shrinking,
a glob of siHcone sealant was squeezed into ±e distal tip to form a seal.
The antenna was taped side-by-side alongside antenna RG174-2, and die coax prepared for
the SMA connector so as to have the exact same connector-to-junction distance. The SMA
connector was soldered to the feedline in the usual manner.
MEASURED DIMENSIONS:
Section length
Overall antenna O.D. (2c)
Antenna element O.D. (2a)
Section A
Section B
3.2 cm
0.143" (3.63mm)
0.110" (2.79mm)
6.4cm
0.145" (3.68mm)
0.110" (2.79mm)
Junction width: 1mm (section A connection slightly tapered)
Connector to junction: 59.3cm (not measured directly, but identical to RG174-2)
135
Antenna RG174-4
CONSTRUCTION DATE: March 30,1990
DESCRIPTION: Choked antenna using RG174 coaxial feedline, a duplicate of antenna
RG174-2 in terms of construction and feedline length. Intended section lengths are 3.2
cm.
MATERIALS:
RG174 coaxial cable with jacket, the usual miniature SMA connector (Omni Spectra
2031-5003-00), Heatrax heatshrink tubing (Radio Shack, 3/32" (#8512) and 3/16"
(#8514)). Custom braid from Stu's 4-conductor cable (New England Electric Wire 247-44, 0.084" nom. O.D.)
ASSEMBLY: A new SMA connector was installed on a length of RG174 in the usual
manner. The cable was cut to be slighdy longer than the connector-to-junction length of
antenna RG 174-2. The two connectors were aligned and the feedline taped alongside
antenna RG174-2 so that the junction could be placed at the same distance. The
unprepared feedline extended about 1.5 cm beyond the junction of antenna RG174-2.
87 mm of jacket was removed from the feedline using a scalpel. The feedline braid was
wiped with alcohol and covered with a 7.0 cm length of 3/32" heatshrink tubing to
form the choke dielecuic layer. O.D. of choke layer was 0.1(M" (2.64 mm).A length of
3/32" heatshrink was shrunk over a ~4.5 cm length of RG174 (without jacket) to serve
as a mandrel for measuring braid length for section B, and subsequenty become section
A. A 3.3 cm length of braid was prepared using tiiis mandrel, and the braid was slipped
over the feedline to form section B. The shield of the RG174 feedline was trimmed so
that it extended about 1-2 mm from the choke dielectric, and the coax dielectric was
stripped, exposing about 1.5 cm of inner conductor The feedline shield was tinned and
then soldered to the braid, forming section B. It was noted that the attachment of
section B left very littie coax dielectric exposed (<0.5 mm.), so antenna bending may
cause this antenna to short circuit. The inner conductor was tinned and the braid slipped
over the mandrel for section A. The antenna was wiped with alcohol, and
approximately 25 cm of 3/16" heatshrink tubing was cut and shrunk over the completed
antenna. A glob of silicone sealant was placed in the end of the tube prior to
heatshrinking to seal the antenna end. Excess heatshrink was trimmed off the distal tip
of the antenna. It was noted during the final measurements that section B varied
somewhat in length from 3.4 to 3.5 cm.
MEASURED DIMENSIONS!
Section Length
Overall antenna O.D.
Antenna element braid O.D.
Section A
Section B
3.20cm
0.147" (3.73mm)
0.113" (2.87mm)
3.45cm (avg.)
0.147" (3.73mm)
0.113" (2.87mm)
Junction size (length): 1mm
Connector to junction length: 59.3 cm (matched to antenna RG174-2)
136
Antenna RG178-12
CONSTRUCTION DATE: June 15,1989
DESCRIPTION: Antenna based on RG-178 feedline with four layers of heat shrink to
form choke dielectric. Represents the "thick choke" antenna for comparison with
antennas RG58-2 and RG174-2.
MATERIALS:
RG-178 coax feedline (with jacket) with standard SMA connector (new)
Heatrax shrink tubing (Radio Shack): 1/16" (#8511), 3/32" (#8512), 1/8" (#8513), 3/16"
(#8514)
New England Electric Wire braid, 24-7-44, .084" nom. O.D, (from Stu's 4-conductor
cable)
ASSEMBLY TECHNIQUE: 7.2cm of jacket was removed at the junction end of a 60cm
length of RG-178 coaxial feedline. The junction area was prepared by removing 8mm of
braid and stripping the dielectric to expose about 4mm of inner conductor, which was
tinned. There was approximately 0.5mm of dielectric protruding from under the feedline
braid. The four layers of heat shrink tubing listed below were applied over the feedline
braid, leaving approximately 1mm of braid exposed at the junction for connection to the
choke. The heatshrink layers were flush at the junction end. efore each layer was applied,
the underlying surface was wiped with an alcohol wipe and allowed to d^. Next, the 247-44 braid was stretched over the heat shrink layers to form the choke. The desired
length of the braid was determined (to make hg 3.2cm) and the excess trimmed. The
junction end of the braid was twisted to make contact with the feedline braid and soldered
circumferentially.
Section A was made using an excess of RG-178 coax (jacket removed) and applying
the same four layers of heat shrink as used for section B. The braid was slipped over
these layers, and "tapered" at one end to form a hemispherical end to which the central
conductor of the feedline could be soldered. The central conductor was bent to form a
hook before soldering on the finished section.
A 16cm length of 3/16" heat shrink was used to cover the entire antenna. A plug of
silicone sealant was used to seal the tip before heat shrinking this final layer. A new SMA
connector (standard size) was attached in the usual manner.
CONSTRUCTION SUMMARY / DIMENSIONS:
Section A
Section B
Choke layer 1 (innermost)
Choke layer 2
Choke layer 3
Choke layer 4 (outermost)
1/16" HS
1/16" HS
3/32" HS
1/8" HS
6.35cm of 1/16" HS
10.0cm of 1/16" HS
9.9cm of 3/32" HS
10.0cm of 1/8" HS
Section Length
Braid (radiating element) O.D.
Overall insulated antenna O.D.
3.20cm
0.148" (3.76mm)
0.173" (4.39mm)
3.25cm
0.148 (3.76mm)
0.173"(4.39mm)
Junction Size: 1mm
Connector to junction: 59cm
137
Antenna F12-2
Construction transcript: Started with approximately 82.5 cm. of RG178 with jacket
removed for the feedline. The Foley balloon was inflated and a mark made 3.5 cm.
proximal to this mark: a hole was made for the feedline at this point. Section B was
prepared from 5 cm. of the New England Electric Wire braid (24-7-44,0.084" nom. O.D.
from Stu's custom 4-conductor cable). After section B was slipped on the Foley, the
junction end of the feedline was prepared and slipped through the caAeter from the junction
feedline hole. Section A was made from another 5 cm.length of braid and slipped over the
Foley catheter. Sections A and B were soldered to the feedline, again maximizing contact
area by using a generous length of feedline conductor and spreading the individual wires.
Diameters of the conductive elements A and B areas follows:
Section A
Section B
0.187" (4.75 mm)
0.180" (4.57 mm)
The exit point of the feedline at the junction was sealed with a drop of silicone sealant and
allowed to dry for 2 hours. 25 cm. of 3M clear polyolefin heat shrink tubing (.25" dia.,
VFP 876) was slipped over the catheter. Silicone sealant was placed circumferentially
around the distal part of the catheter underneath the proximal edge of the balloon, and the
heat shrink was shrunk starting at the distal end of the catheter. The distal part of the heat
shrink reached to about 2 mm. below the proximal edge of the balloon, leaving room for
placement of the thermometry catheter. The thermometry catheter consisted of AWG20
light wall (TFL20 natural, Atlantic Tubing Co.) tubing. The terminal end of the
thermometry catheter was plugged using a piece of wire insulation that snugly fit in,
secured by super glue. The thermometry catheter was secured at the distal end of the Foley
antenna using a piece of 1/4" Heatrax heat shrink tubing (#8515, Radio Shack) and the
remainder of the thermometry tubing was tacked alongside the antenna using small pieces
of surgical tape spaced approximately 3 cm. A wooden tongue depressor was used to apply
a small amount of silicone sealant to attach the thermometry catheter to the Foley antenna. A
standard SMA connector (Omni Spectra 2031-5003-00) was attached in the usual manner.
The silicone sealant was allowed to dry for an hour or two, at which time the tape was
removed and sealant applied at the remaining points along the antenna. Finally, some
silicone sealant was injected using a tuberculin syringe (without needle) into the drainage
port of the Foley catheter. Silicone was also applied circumferentially around this area and a
cm length of 3/16" Heatrax heat shrink tubing (#8514) was shrunk to totally seal these
drainage holes. The balloon was then inflated and it was verified that none of the heat
shrink on either side of the balloon interfered with the inflation. Finally, approximately 1
cm. of 1/4" Heatrax heat shrink tubing was used to secure the proximal end of the
thermometry catheter. Silicone sealant was also applied underneath this layer.
138
T HAYER S CHOOL OF ENGINEERING
R
T
M
O
U
T
H
C
O
L
L
E
G
E
Terence Z. Wong, M.S.
(603) 646-8248
TECHNICAL MEMORANDUM
TO:
Microwave hyperthermia group
FROM: Terry Wong
DATE: March 30,1990
SUBJECT: HP software for HP vector voltmeter
Software has been developed on the HP87 computer to automatically measure complex
impedance at any number of specified frequencies. Basically, the software accomplishes
the same normalization on the vector voltmeter as the "SAVE REF ' function, but enables
multiple frequencies to be run without connecting/disconnecting the reference short each
time. A brief description of this software and its operation is given here.
REQUIRED MATERIALS:
HP 8508A Vector voltmeter
HP 8350A Sweep oscillator
Dual-directional coupler (covering appropriate frequency range)
Phase-matched cables (2 required)
Test load (antenna)
Reference short- Identical feedline (same length, cable type, and connector as
antenna) terminated in a short circuit
Data disk (5.25" DS,DD)
METHOD OF OPERATION:
The basic setup is shown below in figure 1, and the following process is repeated at each
frequency. First, the reference short circuit is connected to the dual-directional coupler
output and the resulting impedance, Zsc, measured. For a shorted line, the input impedance
is:
Zgc ~ Zo tanh yl
where 1 is the line length, ZQ is the characteristic impedance of the line (50 Q), and y is the
complex propagation constant of the line. The short circuit is then replaced by the unknown
load ZL, which is the impedance at the antenna junction. The resulting impedance measured
by the vector voltmeter is Zm'.
ZL + ZQ tanh ji.
^ ^Zo + ZttanhTl
HP 87 Software for impedance measurements
139
From the first measurement, we know that tanh yl = Zgc/Zo- Therefore, we can calculate
the impedance at the reference point (antenna junction):
Zl-ZO
[Zsc " Zm]
ZmZis c
-Zo
. Zo
This process must be repeated at each frequency. The computer program enables
measurements to be made at multiple frequencies using two passes. First, the reference
short circuit is connected and the resulting data stored on a floppy disc. Then, the unknown
load is connected and the previously obtained short circuit data are used to calculate the
actual impedance at each frequency. Thus, there are three types of data files used by the
computer: 1) FREQuFIL contains a list of frequencies for which measurements are desired,
2) SC_FIL contains the frequencies and corresponding reference short circuit data, and 3)
Z_F1L contains the frequencies and calculated impedance at the reference point (antenna
junction).
USING THE SOFTWARE:
By convention, the program disk should go into the LEFT drive and the data disk into the
RIGHT drive. The disk drive and other peripherals should be powered up before turning
on the HP87. The program is menu-driven by "ZMENU", which automatically starts up
when the computer is turned on.
1)
Create a FREQ FIL, the file containing desired measurement frequencies. This is
done by running CR_FREQ. This program allows you to key in the desired
frequencies (in MHz) for subsequent measurements. When finished, type zero as your
last data entry.
2)
Create SC FIL, the data file containing short circuit data. Run CR_SCFIL with the
reference short circuit attached to the dual-directional coupler output.
3)
Create Z FIL, the file containing frequencies and the desired impedance data. Run
CR_ZFIL with the load under test connected to the dual-directional coupler output.
Note that the values displayed on the vector voltmeter will not correspond to the
calculated values seen on the HP 87 computer screen.
4) To print data, mn LISTDATA. This program will give you the option of displaying
or printing the data. This program can read any of the three types of data files created
above.
Note that you need not repeat all of these steps each time. For example, if the setup is
unchanged the short circuit data file is good for any number of runs.
Since there is no printer at Thayer School, there is a duplicate of the program disc at the
hospital, so that you can print out your data there. Please advise me of any problems or
suggestions that you have with this software.
HP 87 Software for impedance measurements
140
Data acquisition
and storage
HP 87 Microcc ;uter System
HP-IB^
HP 8350A
Sweep Oscillator
H
HP-IB
HP 8508A
Vector Voltmeter
i
A
Phasematched [REF
Reference SC
Dual-directional coupler
Antenna
FIGURE 1. Diagram of sytem for impedance measurements. The vector voltmeter is
connected to the dual-directional coupler via phase-matched cables. The forward power
should be connected to input A of the vector voltmeter and the reflected power connected to
input B.
HP 87 Software for impedance measurements
141
SOFTWARE LISTING
'Oa • CR_FREQ
150
200
Program allows user to create list of frequencies for impedance
measurements.
250
300
February 1990
350
Terence Z Wong
400
450
500 DIM FREQ(300),FREQFIL$[8]
550 !
600 CLEAR ® DISP
650 DISP "Program creates file containing frequencies to be swept by HP"
700 DISP "Enter name of OUTPUT (Frequency data) file":
750 INPUT FREQFILS
800 CREATE FREQFIL$&":D701", 1 0
850 ASSIGN# 1 TO FREQFIL$&":D70I"
900 !
950 DISP
1000 DISP "Enter frequencies in MHz, 0 to stop"
1050 FOR i=1 TO 300
1100
DISP i!
1150
INPUT FREQ(i)
1200
IF FREQ(i)=0 THEN 1400
1250 NEXT i
1300 !
1400 FREQ<0>=i-1
1410 FOR j=i TO 300
1420
FREQ(j)=0
1430 NEXT j
1440 !
1500 PRINT# 1 ; FREQO
1550 ASSIGN# 1 TO *
1560 DISP ® DISP "Frequency file created"
1565 WAIT 2000
1570 CHAIN "ZMENU"
1600 END
HP 87 Software for impedance measurements
142
100 ! CR_SCFIL
150
Program obtains complex impedance from vector voltmeter at frequencies
200
specified by data file FREQFILS and records impedance with a reference
250
termination (short cct>. The data are recorded in an SCDATA file for
300
subsequent use by program CR_ZFIL. to calculate the actual impedance
325
values.
326
350
400
HARDWARE REQUIREMENTS:
HP 8350B
Sweep Oscillator
HPIB address 716
450
HP 8508A
Vector Voltmeter
HPIB address 708
500
550
GOO
Mar 1990
Terence Z Wong
650
700
750 DIM PS[70].T$[70],FREQFILS[8].SCFIL$[8]
800 DIM FREQOOO)
850 P$="DDDD.D.4X,SDDDD.ODD.2X,SDDDD.ODD,BX.SDDDD.ODD.3X,SDDD.D"
900 T$="FREQ(MHZ)
Zreal
Zimag
Zmag
Zphase"
950 CLEAR ® DISP
1000 DISP
1050 DISP "Enter data file containing frequencies-"
1100 INPUT FREQFILS
1150 ASSIGN# I TO FREQFIL$&":D701"
1200 READ# 1 : FREQO
1250 ASSIGN# 1 TO *
1255 DISP "OUTPUT file for short circuit impedance data-"
1260 INPUT SCFILS
1265 Nrec=FREQ<0)+1
1270 CREATE SCFIL$&":D701",Nrec,24
1275 ASSIGN# 2 TO SCFILSA":D701"
1280 PRINT# 2 : FREQFILS
1285 PRINT# 2 ; "*SCDATA*"
1288 0scaddr=718
1300 REMOTE Oscaddr
1350 OUTPUT Oscaddr ;"IP"
1400 OUTPUT Oscaddr ;"CW"
1450 OUTPUT Oscaddr ;"PL10"
1500 OUTPUT Oscaddr ;"M0"
1550 OUTPUT Oscaddr ;"*RST"
1600 !
1650
1700
1710
1750
1800
1850
1900
1950
2000
2050
2100
2150
2200
2250
2300
2350
2450
2500
2550
Z0=50
PRINT "Attach reference load (short cct) to output port"
PRINT "
<cont> when ready-"
PAUSE
PRINT
PRINT
PRINT T$
FOR i=1 TO FREQ(O)
FREQ$="eW="&VAL$ (FREQ(i))&"MZ"
OUTPUT Oscaddr ;FREQ$
WAIT 2000
OUTPUT 708 ;"FORMAT RECT"
OUTPUT 708 :"MEAS? Z"
OUTPUT 708 ;"FETCH?"
ENTER 708 ; Zre.Zim
PRINT# 2 ; FREQ<i),Zre.Zim
OUTPUT 708 -."FORMAT POLAR"
OUTPUT 708 :"MEAS? Z"
OUTPUT 708 •."FETCH?"
HP 87 Software for impedance measurements
2800
ENTER 708 : Zmag.Zpha
2650
PRINT USING P$ : FREQ<i),Zre.Zim,Zmag,Zpha
2700 NEXT i
2750 !
2800 ASSIGN# 2 TO *
5300 t
5310 DISP ® DISP "»»* Short circuit data recorded ***"
5320 WAIT 2000
5330 CHAIN "ZMENU"
5350 END
HP 87 Software for impedance measurements
144
1 0 0 ! CR_ZFIL
150
I
Program generates impedance data at predetermined frequencies.
Frequency and short circuit data are read from a data file
previously created by <CR_SCFIL>.
200
250
300
350
400
450
500
550
HARDWARE REQUIREMENTS:
HP 8350B
Sweep Oscillator
HP 8508A
Vector Voltmeter
HPIB address 716
HPIB address 708
600
550
Mar 1990
Terence Z Wong
700
750 DIM P$C70].T$[70].SCFIL$[8].FREQFIL$[8].DUM$[8],ZFIL$[8]
850 P$="DDDD.D.4X,SDDD.DDD,2X,SDDD.DDD,8X.SDDDD,DDD.3X.SDDD.D"
Zimag
Zmag
Zphase"
900 T$="FREQ<MHZ) Zreal
950 ?
1000 CLEAR ® DISP
!050 DISP "Enter file containing frequencies and short circuit data-";
1100 INPUT SCFILS
1150 ASSIGN# 1 TO SCFILSA":D701"
1160 IF TYP (1)= 1 THEN PRINT
WRONG DATA FILE TYPE
® GOTO 1000
1200 READ# 1 ; FREQFILS
1210 READ# I ; DUM$
1220 IF DUM$ <> "-SCDATA*" THEN DISP
NOT A SC DATA FILE
@ GOTO 1000
1230 ASSIGN# 2 TO FREQFILSA":D701"
1232 READ# 2 ; NFREQ
1234 ASSIGN# 2 TO *
1250 DISP "Input OUTPUT data file for impedance data-";
1255 INPUT ZFILS
1260 Nrec=NFREQ+1
1265 CREATE ZFIL$&":D701",Nrec,24
1270 ASSIGN# 3 TO ZFILSA";D701"
1290 0scaddr=718
1300 REMOTE Oscaddr
1350 OUTPUT Oscaddr "IP"
1400 OUTPUT Oscaddr "CM"
1450 OUTPUT Oscaddr "PL10'
1500 OUTPUT Oscaddr "MO"
1550 OUTPUT Oscaddr "•RSI1600 !
1650
1700
1750
1800
1850
I860
1865
1875
1900
1950
1980
2000
2010
2050
2100
2150
2200
2250
2300
2750
Z0=50
PRINT Attach test load to output port- CONT when ready-"
PAUSE
PRINT
PRINT
PRINT# 3
FREQFILS
PRINT# 3
SCFILS
!
PRINT TS
LOOP:
IF TYP (I >=3 THEN DONE
READ# 1 ; FREQ.ZSr.ZSi
FREQ$="CW="dVAL$ <FREO)&"MZ'
OUTPUT Oscaddr ;FREQ$
WAIT 2000
OUTPUT 708 :"FORMAT RECT"
OUTPUT 708 ;"MEAS? Z"
OUTPUT 708 i"FETCH?"
ENTER 708 ; Zre.Zim
!
HP 87 Software for impedance measurements
145
2800
2850
3550
3600
3650
3700
3750
3800
3850
3900
3950
4000
4050
4100
4120
4150
4200
4205
4210
4220
4250
4300
4350
4400
4420
4430
4450
4500
4550
4600
4650
4700
4750
4800
4850
4900
4950
5000
5050
5100
5150
5200
5250
5300
5350
ZNr=Z0*< ZSr-Zre)
ZNi=Z0*(ZSi-Zim)
ZDr=FNAxB_R<Zre,Zim,ZSr,ZSi>/ZO-ZO
ZDl=FNAxB_I<Zre.Zim,ZSr.ZSi)/ZO
f
ZLr=FNfioverB_R(ZNr,ZNi.ZDr,ZDi)
ZLi=FNAoverB_I(ZNr,ZNi,ZDr,ZDi)
Zmag=SQR <ZLr*ZLr+ZLi*ZLi>
IF ZLi=ZLr THEN Zpha=45 ® GOTO 4100
Zpha=ATN2 (ZLi.ZLr)
PRINT USING P$ : FREQ.ZLr.ZLi
PRINT# 3 : FREQ.ZLr.ZLi
GOTO LOOP
»
DONE:
ASSIGN# 1 TO *
ASSIGN# 3 TO *
DISP
DISP
DISP
MISSION
A C C O M P L I S H E D ***"
DISP
WAIT 2000
CHAIN "ZMENU"
Functions for complex multiplication and division:
Z = ZA * ZB
Z = ZA / ZB
DEF FNAxB_R<Ar.Ai.Br,Bi> = Ar*Br-Ai*Bi
REAL part of A » B
DEF FNAxB_I<Ar,Ai,Br,Bi) = Ai»Br+Ar*Bi
IMAG part of A * B
DEF FNAoverB_R(Ar,Ai,Br,Bi) = <Ar»Br+Ai*Bi>/<Br*Br+Bi*Bi>
REAL part of A / B
DEF FNAoverB_I<Ar,Ai.Br,Bi) = (Ai»Br-Ar*Bi>/<Br«Br+Bi*Bi>
IMAG part of A / B
END
HP 87 Software for impedance measurements
146
100
110
120
130
! LISTDATA
!
? Program to list/print data files from vector voltmeter impedance
! measurements.
140 !
150 ! Terence Z Wong
Mar 90
160 !
170 !
180 DIM P1$[70],P2$[70].FREQFIL$[8].SCFILSC8],DATAFIL$[8].PC$[1]
200 !
210 P1$="FREQ<MHz)
Zreal
Zimag"
220 P2$="2X.DDDD.5X.SDDDD.DDDD,5X.SDDDD.DDDD"
230 !
240 CLEAR @ DISP "PROGRAM LISTS DATA FROM IMPEDANCE DATA FILES-"
250 DISP
260 DISP "Data file name
270 INPUT DATAFILS
280 ASSIGN# 1 TO DATAFIL$&":D701"
282 IF TYP (1)=1 THEN GOTO 1000
290 !
320 !
330 DISP @ DISP "Output to <P>rinter or <C>RT (CRT is default)";
340 INPUT PCS
350 IF PC$="P" OR PC$="p" THEN PRINTER IS 701
352 PRINT @ PRINT "DATA FILE:";DATAFIL$
355 IF TYP <1)=1 THEN GOTO 1000
360 !
366 READ# 1 ; FREQFILS
368 READ# 1 ; SCFILS
390 PRINT @ PRINT "Frequency data from file =";FREQFIL$
400 IF SCFIL$="*SCDATA*" THEN PRINT "FILE CONTAINS SC DATA" ELSE PRINT "Shor
cuit data file=":SCFIL$
410 PRINT
420 PRINT P1$
430 PRINT
440 IF TYP <1)=3 THEN 900
450 READ# 1 : FREQ.Zre.Zim
460 PRINT USING P2$ ; FREQ,Zre,Zim
470 GOTO 440
480 ?
900 ASSIGN# 1 TO *
910 PRINT
920 PRINT
930 CHAIN "ZMENU"
940 !
1000 ! Frequency data file
1010 READ# 1 : Nfreq
1020 FOR i=1 TO Nfreq
1030
READ# 1 : freq
1040
PRINT freq
1050 NEXT i
1060 !
1100 ASSIGN# 1 TO *
1200 CHAIN "ZMENU"
2000 END
HP 87 Software for impedance measurements
147
100 ! PROGRAM: ZMENU
not
120
130
140
150
!
!
!
!
Menu driver for SAR programs.
March. 1990
Terence Z Wong
151 !
152 !
160 DIM HEADERSC46]
170 DIM PROG$(14)[10] !
Program names
180 DIM Pdesc$<14)[50] !
Program descriptions
190 !
200 ON ERROR GOTO BINFIX
210 LOADBIN "UTIL/1"
240 !
250 BEGINJPROG:
260 CLEAR @ PAGESIZE 24
270 HEADER$=" *** HP Vector Voltmeter: Z<frequency) •** "
280 HEADER$=HGL?$ <HEADERS,1)
290 !
300 AWRITE 2,10,HEADERS
310 !
320 N_prog=4
330 GOSUB PROGdata
340 !
350 FOR i=! TO N_prog
360
AWRITE i+3.5,"k"&VAL$ <i)
370
AWRITE i+3J0,PROG$<i)
380
AWRITE i+3,22,PdescS(i)
390 NEXT i
410 !
420 ON KEY# 1,"CR_FREQ" GOTO K1
430 ON KEY# 2,"CR_SCFIL" GOTO K2
440 ON KEY# 3,"CR_ZFIL" GOTO K3
450 ON KEY# 4,"LISTDATA" GOTO K4
460 ON KEY# 5 GOTO K5
470 ON KEY# 6 GOTO K6
480 ON KEY# 7 GOTO K7
490 ON KEY# 8 GOTO K8
500 ON KEY# 9 GOTO K9
505 ON KEY# 10 GOTO K10
510 ON KEY# 11 GOTO K11
520 ON KEY# 12 GOTO K12
540 ON KEY# 13 GOTO K13
550 ON KEY# 14." QUIT" GOTO K14
560 !
570 !
580 TAKE KEYBOARD
590 KEY LABEL
600 !
610
620
630
640
650
660
670
680
690
700
710
IDLE: GOTO IDLE
!
K1: RELEASE KEYBOARD
K2: RELEASE KEYBOARD
K3: RELEASE KEYBOARD
K4: RELEASE KEYBOARD
K5: GOTO IDLE
K6: GOTO IDLE
K7: GOTO IDLE
K8: GOTO IDLE
K9: GOTO IDLE
9
®
9
9
CHAIN
CHAIN
CHAIN
CHAIN
"CR_FREQ"
"CR_SCFIL"
"CR_ZFIL"
"LISTDATA"
HP 87 Software for impedance measurements
148
720 K10: GOTO IDLE
730 Kit: GOTO IDLE
740 K12: GOTO IDLE
750 K13: GOTO IDLE
760 K14:
770
RELEASE KEYBOARD
780
CLEAR
790
AWRITE 15.10,"Have a good day!"
810
STOP
820 !
830 !
840 !
850 PROGdata:
860 PROG$<t)="CR_FREQ" @ Pdesc$<I)="Creates list of frequencies for Z meas"
870 PR0G$(2>="CR_SCFIL" ® Pdesc$<2)="Z measurements with short ct termination"
880 PR0G$<3)="CR_ZFIL" ® Pdesc$<3>="Calculated complex Z at reference point"
890 PR0G$<4>="LISTDATA" ® Pdesc$<4)="Print data files"
900 PRGG$<5>="" @ Pdesc$<5>="Empty space"
910 PR0G$<6)="" ® Pdesc$(6>="Empty space"
920 PR0G$<7)="" 9 Pdesc$<7)="Empty space"
922 PR0G$<8)="" @ Pdesc$<8>="Empty space"
923 PR0G$<9)="" ® Pdesc$<9)="Empty space"
924 PROG$<10)="" @ Pdesc$<10)="Empty space"
925 PR0G$<11)="" @ Pdesc$<11)="Empty space"
926 PR0G$<12)="" @ Pdesc$<12)="Empty space"
928 PR0G$<13>="" @ Pdesc$<13)="Empty space"
930 PR0G$<14>="QUIT" ® Pdesc$<14)="Exit Z Programs"
940 !
950 !
960 RETURN
970 !
980 !
990 BINFIX:
1000 !
1010 IF ERRN =25 THEN BE6IN_PR0G
1020
DISP "*** Error ":ERRN ;" in line ":ERRL ;"
1030
RELEASE KEYBOARD
1040
STOP
1050 !
1060 !
1290 !
1300 END
149
Description of Software for calculation of Z
All calculations were based on insulated antenna theory as outlined in Tremby, 1982.
Unless otherwise stated, a data file consists of calculations for all frequencies bewteen 200
and 2000 MHz in 20 MHz increments.
CHOKEZ
Calculates Zc for section of coaxial line terminated by a short circuit.
CR_TERM
Creates a dummy file of a fixed termination impedance,Zj, assuming that
this does not change with frequency. This is used to assign an arbitrarily
high termination impedance to represent an open circuit or zero for a short.
IMPED
Calculates section impedance for a range of section lengths. Input is by
keyboard for frequency, antenna dimensions, termination impedance, and
range of section lengths.
IMPEDF
Calculates section impedance for a fixed section length but over a frequency
range of 200-2000 MHz. Required inputs include antenna dimensions by
keyboard and the data file containing termination impedances for each
Arequency. An output data file is created containing the impedances
calculated at each frequency.
READZ
Reads and lists any impedance data file.
ZSUM
Creates a new data file from the sum of two specified impedance data files,
i.e. sums the impedances at each frequency.
NOTES:
1) The values printed out by the above programs are R and X, where the complex
impedance Z is:
Z = R - iX, or
Z = R+jX.
Therefore, the printed values represent the negative of the imaginary part of Z.
2) For compilation the impedance calculations programs need to be linked to WONGLIB
and the IMSL library, e.g. LINK IMPED,WONGLIB,IMSLD/LIBR.
150
Steps for calculating the junction impedance
(Zjct)
of a choked antenna
1)
Create data file to represent infinite termination impedance [e.g. (1.0d20,0.d0)]; run
CR_TERM.
2)
Run IMPEDF to calculate section A impedance, ZA, using infinite termination
impedance file created in step 1 and appropriate antenna dimensions.
3)
Calculate choke section impedance, Zg, using CHOKEZ.
NOTE: steps 4-7 need to be repeated for each inseriton depth.
4)
Use IMPEDF to calculate feedline impedance, Zf, based on the depth of the
termination of section B (junction insertion depth minus section B length). Termination
impedance is infinite for no ground plane, zero with ground plane.
5)
Add impedances Zf and ZC using ZSUM to get ZBT-
6)
Calculate Zg using IMPEDF with appropriate section dimensions and termination
impedance ZBT-
7)
Add impedances ZA and Zg using ZSUM to obtain the antenna junction impedance,
Zjct-
151
Sample Session for Impedance calculation using IMPED and IMPEDF.
User input is shown in BOLD. Comments are BOLD and parenthesized
Each run of IMPEDF results in Z calculations for all frequencies between
200 and 2000 MHz in 20 MHz steps.
Tl> RUN IMPEDF
Program calculates complex impedance of interstitial
microwave antenna sections based on insulated antenna theoryAntenna O.D.(mm):
5.
Antenna element O.D.(mm):
3.
INPUT file containing termination impedancesOPENCT {data file containing arbitrarily high impedances,
created by CR TERM program}
Section length (cm), 0 to stop3.5
OUTPUT file for Zin data as a function of frequencyZSAMPLE
Data for frequencies 200MHz to 2000MHz by 20MHz:
{Data will be printed out in column form (not showh here) suitable for
pasting into MAC documents. The same data is written into the data file}
Section length (cm), 0 to stop- {Can repeat for other h values}
0
FORTRAN STOP
IMPED calculates section impedance for a range of section lengths at a
specified fixed frequency between 200 MHz and 2450 MHz (inclusive).
Tl> RUN IMPED
Program calculates complex impedance of interstitial
microwave antenna sections based on insulated antenna theoryAntenna O.D.(mm):
5.0
Antenna element O.D.(mm):
3.0
Enter frequency (MHz)915.
Sigma 2 = (O.OOOOOOOOOOOOOOOOE+00,O.OOOOOOOOOOOOOOOOE-i-00)
Sigma 4 = (1.486058354377747,0.OOOOOOOOOOOOOOOOE+00)
epsil 2 = (1.5O518O00OO00O00OE-ll,0.000000OOOOO000O0E+OO)
epsil 4 = (4.4975434911346436E-lO,O.0000000{)O0O00000E+OO)
frequency = 915000000.0000000
rad freq = (5749114556.82O00O,O.0OO000OOOOOOOOOOE+OO)
K4_2 = (18680.44255017681,10736.10351344179)
K2_2 = (625.1730210747031,0.OOOOOOOOOOOOOOOOE+00)
K4 = (141.8207918326051,37.85095039560174)
K2 = (25.00346018203687,0.0000000000000000E+00)
IK4/K2r2 = (34.46378956794884,0.OOOOOOOOOOOOOOOOE+00)
IK2cr2 = (3.9073313817168943E-03,O.00000O0000O00000E+OO)
IK4cl = (0.3669624592676120,0.0000000000000000E+00)
c/a = (1.666666666666667,0.OOOOOOOOOOOOOOOOE+00)
Bessel expansion (2nd kind):
YO = (-0.6701416421377425,0.1834337635364965)
Y1 = (-1.863019181597269,0.4309333693727323)
Number of terms in Bessel expansion (2nd kind):
Order 0:
7
Order 1:
7
KL= (41.08174676380713,13.92310201410609)
Line characteristic Impedance Zc= (40.79515797766810,15.08430938813257)
Complex Terminal Impedance Zh (COMPLEX*16)No gnd plane—>infinity, gnd plane—>0
(1.0d20,0.d0)
{complex, double-precision format}
Complex terminal function theta_h=
(4.0795157160097363E-19,1.508430908582978IE-19)
Calculation for a range of section lengths
Input min, max, and increment in cm
2.,4.,.5
{ Data in column form for MAC pasting}
***Section lengths (cm)***
2.000000000000000
2.500000000000000
3.000000000000000
3.500000000000000
4.000000000000000
***REAL Zin***
7.200663126697415
10.13023594950303
13.37060647109413
17.10497697111647
21.52457520432544
***IMAG Zin***
-40.21781118115799
-27.66108484851210
{ Note that these valures are X = -l*Im[Z] }
-18.45454382701078
-11.24050755168460
-5.456496501504663
{Data are also printed in tabular form for readability}
Section length (cm)
Zreal
Zimag
2.000000000000000
7.200663126697415
-40.21781118115799
2.500000000000000
10.13023594950303
-27.66108484851210
3.000000000000000
13.37060647109413
-18.45454382701078
3.500000000000000
17.10497697111647
-11.24050755168460
4.000000000000000
21.52457520432544
-5.456496501504663
FORTRAN CODE - THEORETICAL IMPEDANCE SOFTWARE
PROGRAM CHOKEZ
C
C Program calculates the input impedance of a choke section over a range
C of frequencies (200-2000 MHz).
C
C Terence Z Wong, Apr 14,1990
C
C
C0MPLEX*16 f,Zch,GAM_H,c,twopi,e_r,sq_eps,i,j,Zo,Zc(100)
REAL*8 FMHz(100),Zcr(100),Zci(100),b,a,h
CHARACTER*10 OUTFILE
C
c = (3.0D10,0.D0)
twopi = (6.28318531D0,0.D0)
j = (0.D0,-1.0D0)
i = (0.D0,1.0D0)
C
WRITE(*,*)
WRITE(*,*) 'Program calculates input Z of a choke section'
WRITE(*,*) 'Input section length (cm)-'
READ(*,*) h
WRITE(*,*) 'Choke dielectric OD (mm)-'
READ(*,*) b
WRITE(*,*) 'Choke dielectric ID (mm)-'
READ(*,*) a
WRITE(*,*) 'Choke relative dielectric constant-'
READ(*,*) e_r
C
sq_eps=SQRT(e_r)
Zo=(l 38.DO,O.DO)*LOG10(b/a)/sq_eps
k=0
DO 100 ifreq=200,2000,20
k=k+l
FMHz(k)=ifreq
f=ifreq*(l .OD06,O.DO)
GAM_H=j*twopi*sq_eps*f*h/c
Zc(k) = j*Zo*CDSIN(i*GAM_H)/CDCOS(i*GAM_H)
Zcr(k) = DREAL(Zc(k))
Zci(k) = -1.0*DIMAG(Zc(k))
100 CONTINUE
N=k
C
C
WRITE(*,*)
WRITE(*,*) 'Characteristic impedance of choke section =', Zo
C
WRITE(*,*)
WRITE(*,*) 'FREQUENCY (MHz)'
D0 200k=l,N
WRITE(*,*) FMHz(k)
200 CONTINUE
C
WRITE(*,*)
WRITE(*,•) REAL [Zchoke]'
D0 300k=l,N
WRITE(* *) Zcr(k)
300 CONTINUE
C
WRITE(*,*)
WRITE(*,*) IMAG [Zchoke]'
DO 400 k=l,N
WRITE(*,•) Zci(k)
400 CONTINUE
C
WR1TE(**)
WRITE(*,*) 'OUTPUT file for Z(frequency) data -'
20 FORMAT (A8)
READ(*,20) OUTFILE
OPEN (unit=1,file=OUTFILE,status-new')
WRITE(1 *) (Zc(k),k=l,100)
CLOSE (1)
C
END
PROGRAM CR_TERM
C
C
C
C
C
C
C
Program to create a dummy list of termination impedances, all having
a constant value (e.g. open or short cct).
Terence Z Wong, Apr 14 1990
C0MPLEX*16 Zt(100),Z
CHARACTER*10 FNAME
C
WRl l'E(*,*) 'Program creates dummy file of termination impedances'
WRITE(*,*) 'OUTPUT FILE-'
READ(*,10) FNAME
10 FORMAT(A8)
C
C
OPEN (UNIT=l,FILE=FNAME,STATUS='new')
C
WRITE(*,*) 'Input complex termination impedance (COMPLEX*16) -'
READ(*,*) Z
C
DO 100i=l,100
Zt(i)=Z
100 CONTINUE
C
WRITE(1,*) (Zt(k),k=l,100)
C
CLOSE (1)
C
END
C IMPED
C
C Program calculates the impedance of an antenna section based on the
C insulated antenna theory of Trembly. This version prints out the data in
C separate arrays so that they can be easily cut and pasted into other
C MAC programs such as Qicket Graph or Excel. Output is also provided in
C tabular form for convenience. Impedance is calculated for a range of
C section lengths with all other parameters fixed. The termination impedance
C of the antenna section as well as the range of section lengths is
C specified at run time.
C
C REQUIRED EXTERNAL ROUTINES:
C MMBZJN inlMSLD/LIBR
C BESL2K0 in WONGLIB
C BESL2K1 in WONGLIB
C EPSR in WONGLIB
C SIG in WONGLIB
C
C Version 02 Oct 19 1989
C Version 03 Apr 02 1990 - Frequency range 200-2450 MHz
C
C Terence Z Wong
C
C
COMPLEX*16 epv,ep2r,mu,rfreq,Zhn
COMPLEX*16 imag,jmag,pi,KL,K2,K4,K2_2,K4_2,eps2,eps4,sig2,sig4
COMPLEX*16K4C,C_A,LNCA,HO,Hl,YO,Yl.JO,Jl,BR(2),BI(2),F,Zc,Zin
COMPLEX*16 theta,XL,K42,K42_2,K4CM,K2C,K2C_2
REAL*8dia2,dia4,K4CR,K4CI,h,ZRin(50),ZIin(50),hL,hH
REAL*8 freq,dh,hcm(50)
C
WRITE(*,*) 'Program calculates complex impedance of interstitial'
WRITE(*,*) 'microwave antenna sections based on Tremblys'
WRITE(*,*) 'insulated antenna theory-'
100 WRITE(*,*)
WRITE(*,*) 'Antenna O.D.(mm):'
READ(*,*) dia4
WRITE(*,*) 'Antenna element O.D.(mm):'
READ(*,*) dia2
C
C Convert antenna diameters to meters
dia2 = dia2/1000.
dia4 = dia4/1000.
C
C Frequency in Hz
WWTE(*,*) 'Enter frequency (MHz)-'
READ (*,*) freq
freq=freq*1.0D06
C
C Constants
pi = (3.141592654D0,0.D0)
imag = (O.DO.l.ODO)
jmag = (0.D0,-1.0D0)
C
158
C Permeability of free space and regions 2,4 (H/m)
mu = 4.0*pi*1.0D-07
C
C Permittivity of vacuum (F/m)
epv = (8.854D-12,0.D0)
C
C
C Region 2 (catheter dielectric) parameters
ep2r = (1.7D0,0.D0)
C
C
C
rfreq = 2.*pi*freq
eps4 = EPSR(freq)*epv
eps2 = ep2r*epv
sig4 = SIG(freq)
sig2 = (O.DO,O.DO)
C
C Calculate wavenumbers (K4_2 and K2_2 are K4 and K2 squared)
K4_2 = (rfreq**2.)*mu*eps4 + imag*rfreq*mu*sig4
K2_2 = (rfreq**2.)*mu*eps2 + imag*rfreq*mu*sig2
C
K4 = K4_2**.5
K2 = K2_2**.5
C
C
WRITE(*,*) 'sigma 2 -,sig2
WRITE(*,*) 'sigma 4 =',sig4
WRITE(*,*) 'epsil 2 =',eps2
WRri'H(*,*) 'epsil 4 =',eps4
WRITE(*,*) 'frequency =',freq
WRITE(*,*) 'rad freq =',rfreq
WRITE(*,*) 'K4_2 =',K4_2
WRITE(*,*) 'K2_2 =',K2_2
WRrrE(**)
WRITE(*,*) 'K4 =',K4
WRITE(*,*) 'K2 =',K2
C
K42 = K4/K2
K42_2 = K42*DCONJG(K42)
C
K2C = dia4*K2/2.
K2C_2 = K2C*DCONJG(K2C)
C
K4C = dia4*K4/2.
K4CM = CDSQRT(K4C*DCONJG(K4C))
C
C_A = dia4/dia2
C
WRITE(*,•) 'IK4/K2r2 =",K42_2
WRITE(*,*) 'IK2cr2 =',K2C_2
WRITE(*,*) 'IK4cl =',K4CM
WRITE(*,*) 'c/a =',C_A
C
K4CR = DREAL(K4C)
K4CI = DIMAG(K4C)
N=2
C
C N=1 is Bessel function of the first kind, order 0
C N=2 is Bessel function of the first kind, order 1
C
CALL MMBZJN (K4CR,K4CI,N,BR,BI,ffiR)
C
JO = BR(l)+imag*BI(l)
Jl=BR(2)+imag*BI(2)
C
C Bessel functions of the second kind, orders 0 and 1
CALL BESL2K0 (K4C,Y0,ml)
CALLBESL2K1 (K4C,Yl,m2)
C
WRITE (*,*)
WRITE (*,*) 'Bessel expansion (2nd kind);'
WRITE (*,*)• YO=',YO
WRITE (*,*)' Y1=',Y1
WRITE (*,*)
WRITE (*,*) 'Number of terms in Bessel expansion (2nd kind):'
WRITE (*,*)' Order 0:',ml
WRITE (*,*)• Order l:',m2
WRITE (*,*)
C
C Hankel functions of orders 0 and 1
HO = JO + imag*YO
HI =J1 +imag*Yl
C
C See eqn(6) p89 of Trembly's thesis
C
F = H0/(H1*K4C)
LNCA = CDLOG(C_A)
C
KL = K2*( ((LNCA+F)/(LNCA+(K2_2/K4_2)*F) )**.5 )
C
WRITE(*,*) 'KL=',KL
C
Zc = mu*fi:eq*KL*(LNCA+(K2_2/K4_2)*F)/K2_2
C
WRITE(*,*) 'Line characteristic Impedance Zc=',Zc
C
C Calculate section input impedance based on section length
C
WRITE(*,*)
C
WRITE(*,*) 'Complex Terminal Impedance Zh (COMPLEX*16)-'
WRITE(*,*)' No gnd plane~>infinity, gnd plane~>0'
READ(*,*) Zh
Zhn = Zh/^
CALL ARCOTH (theta,Zhn,m)
WRITE(*,*) 'Complex terminal function theta_h=',theta
WRITE(*,*)
WRiih(*,*) 'Calculation for a range of section lengths'
WRITE(*,*) 'Input min, max, and increment in cm'
READ(*,*) hL,hH,dh
C
WRITE(**)
WRITER*)
k=l
hcm(k)=hL
500 h=hcm(k)/100.
C
XL = KL*h + imag*theta
C
Zin = imag*Zc*CDCOS(XL)/CDSIN(XL)
C
ZRin(k) = DREAL(Zin)
ZIin(k) = -1 •DIMAG(Zin)
C
hcm(k+l) = hcm(k)+dh
k=k+l
IF (hcm(k).LE.hH) GO TO 500
C
C
C PRINT in separate columns for easy pasting into CRICKET GRAPH
C
kmax=k-l
WRITE(**)
WRITE(*,*) '***Section lengths (cm)***'
DO 520k=l,kmax
520 WRITE(*,*) hcm(k)
C
WRITE(**)
WRITE(*,*) '***REAL Zin***'
DO 530k=l,kmax
530 WRITE(*,*) ZRin(k)
C
WRITE(*,*)
WRITE(*,*) '***IMAG Zin***'
DO 540 k=l,kmax
540 WRITE(*,*) ZIin(k)
C
C
C PRINT in tabular form for readable hardcopy
C
WRITE(*,1100)
1100 FORMAT (/20H Section length (cm),13X,5HZreal,20X,5HZimag)
DO 1000 k=l,kmax
WRITE(**)
WRITE(*,*) hcm(k),ZRin(k),ZIin(k)
1000 CONTINUE
END
161
C IMPEDF
C
C Program calculates the impedance of an antenna section based on the
C insulated antenna theory of Trembly. This version prints out the data in
C separate arrays so that they can be easily cut and pasted into other
C MAC programs such as (Mcket Graph or Excel. Output is also provided in
C tabular form for convenience. Impedance is calculated for a range of
C frequencies with all other parameters fixed. The termination impedance
C of the antenna section as a function of frequency is read in from a
C previously created data file. Currently, it is assumed that this data
C are for a frequency range of 200 MHz to 2000 MHz in 20 MHz steps and
C are in a COMPLEX*16 array.
C
C
C REQUIRED EXTERNAL ROUTINES:
C MMBZJN inlMSLD/LIBR
C BESL2K0 in WONGLIB
C BESL2K1 in WONGLIB
C EPSR in WONGLIB
C SIG in WONGLIB
C
C Version 02 Oct 19 1989
C Version 03 Apr 02 1990 - Tissue properties: frequency range 200-2450 MHz
C Version 04 Apr 14 1990 - Reads termination Z from external data file
C
C INPUT DATA RLE: Zh COMPLEX*16 array of termination impedances
C
values for freq = 200-2000MHz in 20MHz steps
C
C OUTPUT DATA FILE: Zin COMPLEX*16 array of section input impedances
C
values for freq = 200-2000MHz in 20MHz steps
C
C Terence Z Wong
C
C
COMPLEX*16 epv,ep2r,mu,rfreq,Zhn,Zh(100)
COMPLEX*16 imag,jmag,pi,KL,K2,K4,K2_2,K4_2,eps2,eps4,sig2,sig4
COMPLEX*16 K4C,C_A,LNCA,H0,H1,YO,Y1,J0,J1,BR(2),BI(2),F,Zc
COMPLEX*16theta,XL,K42,K42_2,K4CM,K2C,K2C_2,Zin(100)
REAL*8dia2,dia4,K4CR,K4CI,freq,ZRin(100),ZIin(100),h
CHARACTER*10 INFILE,OUTFILE
INTEGER k,kmax
C
WRITE(*,*) 'Program calculates complex impedance of interstitial'
WRITE(*,*) 'microwave antenna sections based on Tremblys'
WRITE(*,*) 'insulated antenna theory-'
100 WRITE(*,*)
WRITE(*,*) 'Antenna O.D.(mm):'
READ(*,*) dia4
WRITE(*,*) 'Antenna element O.D.(mm):'
READ(*,*) dia2
C
C Convert antenna dimensions to meters
dia2=dia2/1000.
dia4=dia4/1000.
c
C Constants
pi = (3.141592654D0,0.D0)
imag = (O.DO.l.ODO)
jmag = (0.D0,-1.0D0)
C
C Permeability of free space and regions 2,4 (H/m)
mu = 4.0*pi*l.dD-07
C
C Permittivity of vacuum (F/m)
epv = (8.854D-12,0.D0)
C
C Region 2 (catheter dielectric) rel permittivity
ep2r = (1.7D0,0.D0)
C
10 FORMAT (AlO)
WRITE(**)
WRITE(*,*) 'INPUT file containing termination impedances-'
READ(*,10) INFILE
OPEN (unit=l ,file=INFILE,status='old')
READ (1,*) (Zh(m),m=l,100)
CLOSE (1)
C
50 WRITE(*,*)
WRITE(*,*) 'Section length (cm), 0 to stop-'
READ(*,*) h
IF (h.EQ.O.) STOP
C
C Convert to meters
h=h/100.
C
WRITE(*,*)
WRITE(*,*) 'OUTPUT file for Zin data as a function of frequency-'
READ(*,10) OUTFILE
OPEN (unit=2,file=OUTFILE,status='new')
C
C
C MAIN LOOP- FREQUENCIES 200MHz to 2000MHz by 20MHz steps
C
k=0
DO 2000 ifreq=200,2000,20
C
k = k+1
freq = ifreq*1.0D06
rfreq = 2.*pi*freq
eps4 = EPSR(freq)*epv
eps2 = ep2r*epv
sig4 = SIG(freq)
sig2 = (O.DO,O.DO)
C
C Calculate wavenumbers (K4_2 and K2_2 are K4 and K2 squared)
K4_2 = (rfreq**2.)*mu*eps4 + imag*rfreq*mu*sig4
K2_2 = (rfreq**2.)*mu*eps2 + imag*rfreq*mu*sig2
C
K4 = K4_2**.5
K2 = K2_2**.5
C
K42 = K4/K2
K42_2 = K42*DCONJG(K42)
C
K2C = dia4*K2/2.
K2C_2 = K2C*DCONJG(K2C)
C
K4C = dia4*K4/2.
K4CM = CDSQRT(K4C*DCONJG(K4C))
C
C_A = dia4/dia2
C
C
K4CR = DREAL(K4C)
K4CI = DIMAG(K4C)
N=2
C
C N=1 is Bessel function of the first kind, order 0
C N=2 is Bessel function of the first kind, order 1
C
CALL MMBZJN (K4CR,K4CI,N,BR,BI,IER)
C
JO = BR(l) + imag*BI(l)
Jl=BR(2)+imag*BI(2)
C
C Bessel functions of the second kind, orders 0 and 1
CALL BESL2K0 (K4C,Y0,ml)
CALL BESL2K1 (K4C,Yl,m2)
C
C Hankel functions of orders 0 and 1
HO = JO + imag*YO
HI = J1 +imag*Yl
C
C See eqn(6) p89 of Trembly's thesis
C
F = H0/(H1*K4C)
LNCA = CDLOG(C_A)
C
KL = K2*( ((LNCA+F)/(LNCA+(K2_2/K4_2)*F) )**5 )
C
C
Zc = mu*freq*KL*(LNCA+(K2_2/K4_2)*F)/K2_2
C
C Calculate section input impedance based on section length
C
C
Zhn = Zh(k)/Zc
CALL ARCOTH (theta,Zhn,m)
C
C
XL = KL*h + imag*theta
C
Zin(k) = imag*Zc*CDCOS(XL)/CDSIN(XL)
C
ZRin(k) = DREAL(Zin(k))
Znn(k) = -1 •DIMAG(Zin(k))
C
2000 CONTINUE
C
C
C PRINT in separate columns for easy pasting into CRICKET GRAPH
C
kmax=k
WRITE(**)
WRITE(*,*) 'Data for frequencies 200MHz to 2000MHz by 20MHz:'
C
WRITE(*,*)
WRITE(*,*) •***REAL Zin***'
DO 530k=l,kmax
530 WRITE(*,*) ZRin(k)
C
WRrrE(**)
WRrrE(*,*) '***IMAG Zin***'
DO 540 k=l,kmax
540 WRITE(*,*) Zlin(k)
C
C
WRITE(2,*) (Zin(m),m=l,100)
C
CLOSE (2)
C
GOTO 50
C
END
PROGRAM READZ
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
Program to list out impedance data produced by the following programs:
CHOKEZ Z(f) of a line terminated by a short cct
IMPEDF Z(f) of a transmission line formed by a lossy medium
CR_TERM Z(f) = constant (eg open cct or short)
ZSUM Z(f) = sum of two previously-created Z(f) files
The underlying assumption is that complex impedance is
recorded at frequencies ranging from 200MHz to 2000MHz at 20MHz
intervals.
Data are listed in tabular form for ease of reading and as separate
columns for easy pasting into Macintosh applications.
Terence Z Wong, Apr 14 1990
COMPLEX*16 Z(IOO)
REAL*8 Zr(100),Zi(100),freq(100)
CHARACTER*12 ZFILE
C
WritE(**)
WRITE(*,*) 'Program lists Z(frequency) created by programs:'
WRITE(* *)' IMPEDE'
WRITE(*,*)'
CHOKEZ'
WRITE(* *)' CR_TERM'
WRITE(*,*)'
ZSUM'
WRITE(*,*) 'It is assumed that frequencies range from'
WRITE(*,*) '200MHz to 2000MHz in 20MHz steps'
5 WRITE(*,*)
WRITE(*,*) Z data file to be listed [type S to exit]'
READ(*,10) ZFILE
10 FORMAT(A10)
IF(ZFILE.EQ.'S'.OR.ZFILE.EQ.'s') STOP
C
OPEN (unit=1,file=ZFILE,status='old')
READ(1,*) (Z(i),i=l,100)
CLOSE (1)
C
k=0
DO 200 ifreq=200,2000,20
k=k+l
freq(k)=ifreq
Zr(k)=DREAL(Z(k))
Zi(k)=-l.*DIMAG(Z(k))
WRITE(*,*) ifreq,' ',Z(k)
200 CONTINUE
C
C
WRITE(*,*)
WRITE(*,*) '*** FREQUENCY (MHz) ***'
DO 300i=l,k
300 WRITER,*) freq(i)
C
WRrrE(**)
WRITE(*,*) •*** REALZ(freq) ***'
D0 310i=l,k
310 WRITE(*,*) Zr(i)
C
WRITE(**)
WRITE(* *) '*** MAG Z(freq) ***'
DO 320i=ljc
320 WRITE(*,*) Zi(i)
C
WRITE(*,*)
WRrrE(**)
GOTO 5
C
END
167
CWONGLIB
C
C Utility Functions and subroutines:
C
C BESL2K0 Bessel function of 2nd kind, order 0
C BESL2K1 Bessel function of 2nd kind, order 1
C HM (m) Calculates sum in Kreyszig, p132-133
C FAC(m) Factorial function (m positive)
C ARCOTH Arc-hyperbolic cotangent (complex argument)
C EPSr(f) Muscle rel dielectric const at frequency f
C SIG(f) Muscle electrical conductivity (mho/m) at frequency f
C
C
C Terence Z Wong
C
C
SUBROUTINE BESL2K0 (Z,YO,m)
C
C Subroutine calculates Bessel function of the second kind, order zero.
C Uses expansion described in Kreyszig E. in Advanced Engineering Mathematics,
C Wiley & Sons, New York, 1972, p.132.
C
C Z COMPLEX*16 Double precision complex argument for Bessel function.
C YO COMPLEX*16 Double precision complex result.
C m INTEGER Number of terms required in summation (see Kreyszig).
C
(max # of terms is currendy 1000)
C
C Other Routines Required:
C Hm(m) in WONGLIB
C FAC(m) in WONGLIB
C MMBZJN in IMSLD/libr
C
CTZW 10/18/89
C
COMPLEX*16 TT,SUM,imag,BO,Z,YO
REAL*8 pi,gamma,BR,BI,ZR,ZI,FI,FR,SUMR,SUMI,TOL
C
C
TOL=1.0E-15
pi=3.141592654
gamma=0.57721566490
imag=(0.D0,1.ODO)
SUM = (O.DO,O.DO)
DO 100 m=l,1000
isign=(-l)**(m-l)
TT = isign*HM(m)*( (((Z/2.)**m)/FAC(m))**2)
SUMR = DREAL(SUM)
SUMI = DIMAG(SUM)
168
C
SUM = SUM+TT
C
IF (SUMR.EQ.O..OR.SUMI.EQ.O.) GO TO 100
C
FR = ABS (DREAL(Tr)/SUMR)
FI = ABS (DIMAG(TD/SUMI)
IF (FR.LT.TOL.AND.FI.LT.TOL) GO TO 900
C
100 CONTINUE
C
C
C Get JO using IMSL routine
C
900 ZR = DREAL(Z)
ZI = DIMAG(Z)
N=1
C
CALL MMBZJN (ZR,ZI,N,BR,BI,IER)
BO = BR + imag*BI
C
C Calculate Yo using eq 6 on p 132 of Kreyszdg;
C
YO = (2./pi)*( BO*(CDLOG(Z/2.)+gamma) + SUM)
C
RETURN
END
SUBROUTINE BESL2K1 (Z,Yl,m)
C
C Subroutine calculates Bessel function of the second kind, order one.
C Uses expansion described in Kreyszig E. in Advanced Engineering Mathematics,
C Wiley & Sons, New York, 1972, p.133 (eq 8).
C
C Z COMPLEX*16 Double precision complex argument for Bessel function.
C Y1 COMPLEX*16 Double precision complex result.
C m INTEGER Number of terms required in summation (see Kreyszig).
C
(current max # terms is 1000)
C
C Other Routines Required:
C Hm(m) inWONGLB
C FAC(m) inWONGLIB
C MMBZJN inlMSLD/libr
C
CTZW 10/18/89
C
C
COMPLEX*16 T1,T2,Tr,SUM,imag,B1,Z,Y1
REAL*8 pi,gamma,BR(2),BI(2),ZR,ZI,FI,FR,TOL,SUMR,SUMI,dpi
C
C
TOL=1.0E-15
pi=3.141592654
dpi=2./pi
gamma=0.57721566490
imag=(0.D0,1.0D0)
C
SUM = (0.D0,0.D0)
DO 100 m=0,1000
isign=(-l)**(m-l)
ml=m+l
C
C Note that the summation term TT is not exactly as shown in eq 8 of Kreyszig,
C pi33. A factor of 1/2 has been taken outside the summation.
C
T1 = isign*(HM(m)+HM(ml)) / (FAC(m)*FAC(ml))
T2 = ((Z/2.)**m)**2
TT = T1*T2
C
SUMR = DREAL(SUM)
SUMI = DIMAG(SUM)
C
SUM = SUM+TT
C
IF (SUMR.EQ.O..OR.SUMI.EQ.O.) GO TO 100
C
FR = ABS (DREAL(TT)/SUMR)
FI = ABS (DIMAG(TT)/SUMI)
IF (FR.LT.TOL.AND.FI.LT.TOL) GO TO 900
C
100 CONTINUE
170
C
C
C Get J1 using IMSL routine
C
900 ZR = DREAL(Z)
71 = DIMAG(Z)
N=2
C
CALL MMBZJN (ZR,ZI,N,BR,BI,IER)
Bl=BR(2) + imag*BI(2)
C
C Calculate Y1 using eq 8 on p 133 of Kreyszig:
C
T1 =dpi*Bl*(CDL0G(Z/2.)+gamma)
T2 = -dpi/Z
Y1 = T1 + T2 + (Z*SUM)/(2.*pi)
C
RETURN
END
FUNCTION HM(m)
C
C
C
C
C
C
C
C
HM(tn) function: calculates the sum used in Bessel function expansionHM(m) = 1 + 1/2 + 1/3 +... + 1/m
(seepages 132-133ofKreyszig)
TZW 10/19/89
hval = 0.
IF(m.EQ.O)GOT0900
C
DO 100i=l,m
v=i
hval = hval + 1/v
100 CONTINUE
C
900 HM = hval
RETURN
END
C
C
FUNCTION FAC(m)
C
C FAC(m)- calculates m!, where m is a positive integer or zero.
C
C TZW 10/16/89
C
C
X=l.
IF (m.EQ.O) GO TO 1000
C
DO 200i=l,m
v=i
X=X*v
200 CONTINUE
C
1000 FAC = X
RETURN
END '
C
SUBROUTINE ARCOTH (ACOTH,Zd,in)
C
C
C
C
C
C
C
C
C
C
C
C
C
Subroutine returns double precision complex value of ARCHYPCOT of Zd,
also a COMPLEX*16 value. Note that for IZdl=l this function is undefined.
See Tremby's thesis p96, eqs (11) and (12) for series expansion used.
ACOTH COMPLEX*16 result: arc-hyperbolic cotangent of Zd
Zd COMPLEX*16 argument
m INTEGER number of terms in series expansion (max 1000)
TZW 10/20/89
COMPLEX*16 SUM,T,imag,Zd,ACOTH
REAL*8 MAG,SR,SI,pi,X,Y,TOL
C
MAG = DREAL( Zd*DCONJG(Zd))
TOL=1.0D-15
pi = 3.141592654
imag = (0.D0,1.0D0)
C
IF (MAG.GT.l.) THEN
SUM = (O.DO,O.DO)
DO 1000 i=l,1000
k=2*i-l
T = (1.0/k)*( (1.0/Zd)**k)
SUM = SUM + T
Y = DREAL(T)
X = DREAL(SUM)
C
C
IF (X.NE.O.) THEN
SR=ABS(Y/X)
ELSE
Case where X=0
IF (Y.EQ.O.) THEN
SR=0.
ELSE
SR=1.
END IF
END IF
C
Y = DIMAG(T)
X = DIMAG(SUM)
C
C
IF (X.NE.O.) THEN
SI=ABS(Y/X)
ELSE
Case where X=0
IF (Y.EQ.O.) THEN
SI=0.
ELSE
SI=1.
END IF
END IF
C
IF (SR.LT.TOL.AND.SI.LT.TOL) GO TO 1500
1000 CONTINUE
GOTO 1500
END IF
C
IF (MAG.LT.1.) THEN
SUM = -imag*pi/2.
DO 12001=1,1080
k=2*i-l
T = (1.0/k)*(Zd**k)
SUM = SUM + T
Y = DREAL(T)
X = DREAL(SUM)
C
IF (X.NE.O.) THEN
SR=ABS(Y/X)
ELSE
C
Case where X=0
IF (Y.EQ.O.) THEN
SR=0.
ELSE
SR=1.
END IF
END IF
C
SI = ABS( DIMAG(T)/DIMAG(SUM))
C
IF (SR.LT.TOL.AND.SLLT.TOL) GO TO 1500
1200 CONTINUE
ENDIF
C
1500 ACOTH = SUM
m=i
RETURN
END
SUBROUTINE COTAN (Z,COTZ,m)
C
C Subroutine calculates COT(Z), where Z is the double-precision complex
C argument and COTZ is the double-precision complex result, m is the integer
C representing the number of terms required in the series expansion.
C
C The expansion used is from HANDBOOK OF MATHEMATICAL FUNCTIONS.,
edited
C by M Abramowitz and I Stegun, National Bureau of Standards Applied
C Mathematics Series 55,1972, p75, eq 4.3.91.
C
C TZW Nov 2,1989
C
C
COMPLEX*16 Z,Z2,Zsq,SUM,TERM,C0TZ
REAL*8TOL,pi,pisq,ksq,SUMR,SUMI,FR,n
C
TOL=1.0E-10
pi = 3.141592654
pisq = pi*pi
Z2 = 2.0*Z
Zsq = Z*Z
C
SUM=1.0/Z
C
DO 400 k=l,10000
ksq = k*k*1.0
TERM = Z2 / (Zsq-ksq*pisq)
C
SUMR = DREAL(SUM)
SUMI = DIMAG(SUM)
C
SUM = SUM+TERM
C
FR = ABS( DREAL(TERM)/SUMR)
FI = ABS( DIMAG(TERM)/SUMI)
C
IF (FR.LT.TOL.AND.FI.LT.TOL) GO TO 450
400 CONTINUE
C
450 COTZ = SUM
m=k
C
RETURN
END
FUNCTION EPSR(freq)
C
C
C
C
C
C
C
Function returns relative dielectric constant at frequency = freq.
Present function is valid from 200 to 2000 MHz, and is based on
data from:
Guy et al, Proc IEEE 62(l):55-75,1974
Schwan and Foster, Proc IEEE 68(1):104-113,1980
freq = freq/1.0D06
CI =74.128
C2 = -7.8784
EPSR = CI + C2*LOG10(freq)
freq = freq*1.0D06
RETURN
END
C
C
FUNCTION SIG(freq)
C
C
C
C
C
C
Function returns electrical conductivity (mho/m) at frequency = freq.
Present function is valid from 200-2450 MHz, and is based on data from:
Guy et al, Proc IEEE 62(l):55-75,1974
Schwan and Foster, Proc IEEE 68(1):104-113,1980
CO = 0.99214
CI =5.6815e-04
C2 = -3.0982e-08
C
freq=freq/l .0D06
SIG = CO + (Cl+C2*freq)*freq
C
freq=freq*1.0D06
RETURN
END
PROGRAM ZSUM
C
C Program sums two complex impedance data files to create a third.
C This is useful to determine the total impedance over the 200MHz
C to 2000MHz fi-equency range when it is a sum of two other calculated
C impedances.
C
C Terence Z Wong, Apr 14 1990
C
C
CHARACTER*10 INFILEl,INFILE2,OUTFILE
COMPLEX*16 Zl(100),Z2(100),Zr(100)
C
WRITE(**)
WRITE(*,*) 'Program creates a new Z(frequency) file based on the'
WRITE(*,*) 'sum of two previously create Z files.'
WRITE(*,*)' Zsum(freq) = Zl(freq) + Z2(freq)'
WRITE(**)
10 FORMAT(A10)
WRITE(*,*) 'INPUT FILE #1 [Zl(freq)] -'
READ(*,10) INFILEl
OPEN (unit=1,file=INFILE1,status='old')
READ(1,*) (Zl(i),i=l,100)
CLOSE (1)
C
WRITE(*,*) 'INPUT FILE #2 [Z2(freq)]
READ(*,10) INFILE2
OPEN (unit=2,file=INFILE2,status='old')
READ(2 *) (Z2(i),i=l,100)
CLOSE (2)
C
DO 100i=l,100
ZT(i) = Zl(i) + Z2(i)
100 CONTINUE
C
WRITE(**)
WRITE(*,*) 'OUTPUT FILE for summed data to be written
READ(*,10) OUTFILE
OPEN (unit=3,file=0UTFILE,status='new')
WRITE(3 *) (ZT(i),i=l,100)
CLOSE (3)
C
END
Ill
SAR Measurements on Transurethral Microwave Antenna
THEORETICAL CONSIDERATIONS
Averaged measured dimensions for the prototype transurethral microwave antenna
used for the the experimental and theoretical calculations are as follows:
Section length
Antenna insulation O.D. (2c)
Antenna insulation I.D. (2a)
2.8 cm avg. (hA = 2.6 cm, hg = 3.0 cm)
5.33 mm
4.66 mm
The insulated antenna theory is valid if the following conditions are met (Trembly,
1982). In the table below, the required conditions are shown on the left, and the calculated
parameters for the prototype prostate antenna (F12-2) shown for die three clinically relevant
driving frequencies.
Parameter
Constraint
433 MHz
915 MHz
2450 MHz
Ik4/k2l^
»1
46.45
35.24
29.25
c/a
>2
1.14
1.14
1.14
Ik2c|2
«1
0.001
0.004
0.03
Ik4cl
^0.5
0.22
0.40
0.97
minr
>2c
5.33 mm
5.33 mm
5.33 mm
Since several of the conditions are not met, the insulation antenna theory may not be
valid in describing the behavior of this antenna. In particular, the relatively large diameter
of the antenna makes it electrically thick in tissue, and the insulation layer covering the
antenna is relatively thin, making the antenna appear more like a bare antenna.
In order to check the validity of the insulated antenna theory in this case, the antenna
junction impedance was calculated theoretically using the technique described in the choke
antenna paper (Chapter 3). The results were compared to the measured impedance at
insertion depths of 4 cm and 12 cm over a frequency range of 200 MHz to 2000 MHz. The
result is shown in Fig. 1. While there is some agreement at low frequencies near 200 MHz,
the theory clearly does not apply at the frequencies of interest (433 MHz, 915 MHz, and
2450 MHz). Therefore, we conlcude that we would not be able to accurately predict the
SAR patterns at these frequencies using insulated antenna theory.
Because the insulation ratio c/a is close to 1, it may be possible to describe the antenna
using bare antenna theory. However, this theory is also complicated and includes
conditions which are not met in this case; the antenna is assumed to be thin and tiie section
length an odd integral number of quarter wavelengths. Neither of these assumptions is fully
met here, so the results permit only a rough comparison. The bare antenna calculations
were programmed into an Excel macro by Prof. Trembly using equations 4.2.24 and
4.2.25 in King and Harrison (RWP King and CW Harrison, Antennas and Waves: A
Modern Approach. MIT Press, Cambridge MA, 1969, p.257). Calculations were
performed in the junction plane at frequencies of 433 MHz, 915 MHz, and 2450 MHz;
electrically, the antenna section was 0.329, 0.636, and 1.591 wavelengths long at these
respective frequencies (Fig. 5). The experimental and theoretical results for the radial
distribution in the junction plane are compared in Fig. 2. At all three frequencies, the theory
178
predicts a steeper fall in radial power deposition. This would be expected when comparing
an insulated antenna to a bare antenna.
At this point, we conclude that the antenna is neither fully insulated nor a truly bare
antenna, and that the antenna's large dimensions violate the size constriants of the two
theoretical models. The remainder of this appendix will serve to summarize the
experimental SAR measurements made on this antenna.
METHODS / RESULTS
For all experiments, the Hartsgrove HEC phantom was used (Hartsgrove, 1987).
Qualitative SAR experiments were made by placing the antenna between the edges of two
rectangular pieces of 30-35°C liquid crystal, so that all three were in the same plane. Suture
was used to hold the liquid crystal edges against the antenna surface. This assembly was
place in a "tupperware '-type plastic container such that the balloon edge was 9 cm deep;
this corresponck to a junction depth of 52 mm, and reproduced the depth from the first dog
experiment (11/9/89). The entrance point of the antenna was sealed with silicone sealant
and HEC phantom was poured into the container. The phantom was placed in a temperature
controlled circulating water bath at 29.0°C for 6-8 hours before experiments. The entire
setup was on a copy stand with photo-flood lamps. Microwave power (20W forward) was
applied for 15 seconds, at which time a photo was taken as an approximation of the SAR
distribution. This was done for 433 MHz, 915 MHz, and 2450 MHz; the photos are shown
in Figure 4.3 of this thesis and are not reproduced here.
Quantitative measurements were also done using the HEC phantom. The HP SAR
system was used to calculate transverse SARs at the junction plane (z=0 mm) and at 10 mm
intervals along the antenna (z=±10 mm, z=d20 mm, z=±30 mm). A geometric correction
was applied to correct for the diameter of the antenna. The abscissa on the transverse plots
is then the distance from the antenna surface.
Measurement
catheter
Temperature
Probe
Since the thermometry catheter is long compared to the antenna radius r, the distance from
the antenna surface is approximated closely by d = V r^+x^ - r , where x is the distance
along the catheter. Results are shown for 433 MHz, 915 MHz, and 2450 MHz in Fig. 3.
Symmetry was used to obtain either 2 or 4 data points for each radial distance. The data
were fit to 3rd order polynomial curves, so that the trends could be visualized more clearly.
Longitudinal measurements were also made along the antenna thermometry catheter
(Fig. 4a) and at a distance of 5 mm from the antenna surface (Fig. 4b). In the latter case, 4
measurement catheters surrounded the antenna at 90° intervals and the results were
averaged. For these experiments, the junction was also approximately 52 mm deep. The
axial SAR was also calculated using bare antenna theory along the catheter surface (radial
distance of 2.66 mm. Fig. 4c) and at a distance of 5 mm from the antenna surface (radial
distance of 7.66 mm, Fig. 4d). There are some end effects seen in Fig. 4c, but the axial
measurements near the antenna in Fig. 4d show very large fields near the ends of the
antenna. This is probably due to the fact that the bare antenna theory is valid only for ph =
179
7I/2. These end effects were found to diminish as radial distance increased and as the
section length approached a quarter wavelength.
For all quantitative calculations of SAR, W/kg/W was calculated using a value of 3098
J/kg*°C for the heat capacity of the HEC phantom (Mark Yeh, memo 1/23/90). Net forward
power at the junction was calculated using the data contained in the memo dated 2/22/90 to
account for line losses.
DISCUSSION / CONCLUSIONS
The qualitative results are consistent with the measured SARs. The maximum SAR
measured at the junction is highest for 2450 MHz and lowest at 433 MHz. (SARs
measured at 433 MHz were uniformly low for this particular antenna, but could be
significantiy improved by increasing the section lengths). Both experiments also show that
the longitudinal distribution is improved at lower frequencies. There is a numerical
discrepancy between the transverse and longitudinal measurements. For example, at the
junction at 2450 MHz, the maximum SAR is about 210 W/kg*W for the transverse
measurement but only about 50 W/kg*W for the longitudinal measurement. The most likely
explanation for this discrepancy is the fact that the longitudinal catheter is attached to the
antenna and essentially surrounded by plastic; therefore, the value for electrical
conductivity, a, is low and the SAR values measured will be proportionally decreased. We
can see much better numerical agreement if we compare the longitudinal measurements
made 5 mm from the catheter with the corresponding transverse results.
e-«r
Since the radial distribution can be expected to decay as r , we can use this to
calculate the radial penetration depth at each frequency (i.e. the distance where the SAR
falls to 1/e of its value measured on the antenna surface). This was done for the theoretical
and measured data at the junction (Fig. 3a) with the results summarized below. The
theoretical bare antenna parameters are summarized in Fig. 5. Theoretical results were also
obtained by Prof. Trembly using insulated antenna theory. The penetration depth, 8, was
calculated using:
5=1. = —1
a
kln(lO) ^
where k is the exponent of the curve fit. The penetration depths based on experimental
measurements were subject to error, as the data did not fit to the exponential curve very
well, especially at 433 MHz. As r increases, SAR decreases so that the measurement noise
is amplified at greater distances. Therefore, the SAR values used were restricted to those
obtained at r < 16 mm for the penetration depth calculations.
Frequency
Bare antenna
Theory
433 MHz
915 MHz
2450 MHz
5.9 mm
5.2 mm
3.9 mm
Penetration depth
Insulated antenna
Theory
11.5 mm
10.8 mm
6.2 mm
Measured
11.6 mm
8.0 mm
4.3 mm
As expected, the lower frequencies radiated over a longer distance. The measured antenna
penetration depths at all three frequencies were generally between the values predicted by
bare and insulated antenna theory. At the lower frequency (433 MHz), the antenna behavior
seemed closer to that of an insulated antenna, while at the highest frequency, the antenna
behavior more closely resembled a bare antenna.
180
The comparison of measurements with theory is summarized in Fig. 7, where the
theoretical exponential curves based on bare and insulated antenna theory are plotted along
with the measured results. It is interesting that both the impedance and SAR measurements
seem to show the insulated antenna theory to be more accurate at the lower frequency (433
MHz).
70
60-
Fig. 1a: Antenna F12-2, jet depth 4 and 12 cm
Theory, 4cm
Theory, 12 cm
Meas, 4 cm
Meas, 12 cm
cn
E
20-
200
500
800
1 100
1400
1700
2000
Frequency (MHz)
40
Fig. 1 b: Antenna F12-2, jet depths 4 and 12 cm
20
-20
Theory, 4cm
Theory, 12cm
Meas, 4 cm
Meas, 12 cm
200
500
800
I 100
1400
Frequency (MHz)
1700
2000
Fig. 2a: Antenna F12-2, |Z|=0, 433 MHz
30
25
20
15
10
5
0
0
3
6
9
12
15
Radial distance (mm)
Fig. 2b: Antenna F12-2, |Z|=0, 915 MHz
70-
DC
<
in
0
3
6
9
12
15
Radial distance (mm)
Fig. 2c: Antenna F12-2, |Z|=0, 2450 MHz
200-
BvtanlenultHoiy
Mewjranieiit
160
S 120
0
3
6
9
Radial distance (mm)
12
15
182
Fig. 3a:
Transverse SAR, |Z|=0
• 433 MHz
• 915 MHz
A 2450 MHz
3
6
9
12
Distance from antenna surface (mm)
Fig. 3b;
75
5 606>
A
A
2450
MHz
g 454
CC 3 0 ^ ^
<
(f)
915
15-
o
433
MHz
15
Transverse SAR, |Z|=10mm
433 (-10mm)
433 (+10mm)
915 (-10mm)
915(+10mm)
2450 (-10mm)
2450 (+10mm)
MHz
8
0
0
3
6
9
12
Distance from antenna surface (mm)
15
183
Fig. 3c:
2450
Transverse SAR, |Z|=20mm
433 (-20mm)
433 (+20mm)
915 (-20mm)
915 (+20mm)
2450 (-20mm)
2450 (+20mm)
MHz
S> 3 0
433
MHz
A
915
MHz
Distance from antenna surface (mm)
Fig. 3d:
Transverse SAR, |Z|=30mm
433 (-30mm)
433 (+30mm)
915 (-30mm)
915 (+30mm)
2450 (-30mm)
2450 (+30mm)
40433
O)
MHz
30915
MHz
2450
MHz
QC
<
en 10-
12
Distance from antenna surface (mm)
15
184
Fig. 4a;
SAR along transurethral thermometry catheter
915 MHz
40-
433 MHz
2450 MHz
CC
<
U)
20-
Antenna
Tip
10-
0
2
4
6
8
10
Distance (cm)
Fig. 4b: SAR 5mm from catheter, avg of 4 catheters
35
30
915 MHz
433 MHz
? 25
O)
2450 MHz
g 20
^ 15
DC
Antenna
Tip
5
0
0
2
4
6
Distance (cm)
8
10
185
Fig. 4c: SAR along catheter surface, bare antenna theory
300
433 MHz
915 MHz
2450 MHz
O)
200-
< 100-
cn
-40
-20
0
20
40
Distance (mm)
Fig. 4d: SAR 5mm from catheter, bare antenna theory
30-
433 MHz
915 MHz
25-
2450 MHz
O)
^ 20-
-40
-20
0
Distance (mm)
20
40
Bare ant parameters
Enter Input parameters:
h. antenna half-lenath (m)
drivinq frequency (MHz)
tissue relative permittivity (1)
tissue conductivity (mho/m)
tissue density (kq/m*3)
input current (A) (arbitrary)
=
=
=
=
=
=
0.028
433
53
1.43
1000
0.1
E
D
C
B
A
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Enter input parameters:
h. antenna half-lenqth (m)
drivinq frequency (MHz)
tissue relative permittivity (1)
tissue conductivity (mho/m)
tissue density (kq/m*3)
input current (A) (arbitrary)
=
=
=
=
=
=
F
Enter input parameters:
0.028 h. antenna half-lenqth (m) =
915 drivinq frequency (MHz) =
tissue relative permittivity (1)
51
tissue conductivity (mho/m)
1.6
tissue density (kq/m*3)
1000
input current (A) (arbitrary)
0.1
=
=
=
=
0.028
2450
47
2.21
1000
0.1
Tissue Properties:
Tissue Properties:
vacuum permittivity (F/m) =
vacuum permittivity (F/m) =
8.854E-12
tissue permittivity (F/m) =
tissue permittivity (F/m) = 4.69262E-10
tissue permeability (H/m) =
tissue permeability (H/m) = 1.25664E-06
Tissue Properties:
8.854E-12
8.854E-12 vacuum permittivity (F/m) =
4.161E-10
4.51554E-10 tissue permittivity (F/m) =
tissue permeability (H/m) = 1.257E-06
1.25664E-06
Miscellaneous Computed Quantities:
59.863005
40.472 alpha (1/m) - Re (qamma) =
357.07564
142.805 beta (1/m) = Im (qamma) =
Miscellaneous Computed Quantities:
alpha (1/m) = Re (qamma) =
beta (1/m) = Im (qamma) =
33.084
73.887
Miscellaneous Computed Quantities:
alpha (1/m) = Re (gamma) =
beta (1/m) = Im (qamma) =
wavelenqth in tissue (m) =
0.085
wavelenqth in tissue (m) =
0.044
wavelenqth in tissue (m) =
0.018
antenna section in wavelenqths =
0.329
antenna section in wavelenqths =
0.636
antenna section in wavelenqths =
1.591
conductivitv/(2n«freq.•permittivity)
=
1.120 conductlvilv/(2n«frea."Permittivity)
=
0.616 conductivitv/(2iffrea.•permittivity)
=
0.345
penetration depth in tissue (m) =
0.030
penetration depth in tissue (m) =
0.025
peneta-ation depth in tissue (m) =
0.017
alpha-h =
0.926
alpha*h =
1.133
alpha^h =
1.676
sinh (alpha-h) =
1.065
sinh (alpha-h) =
1.392
sinh (alpha-h) =
2.579
scale coef. in K&H Eqns. 4.2.24,25= 0.336063003
scale coef. in K&H Eons. 4.2.24,25=
0.38733133
scale coef. in K&H Eqns. 4.2.24.25=
0.425
00
<T>
FIGURES
187
INSULATED ANTENNA THEORY PARAMETERS:
433 MHz
•lis g73
AA =
0.00233
AB=
0.00233
BA=
0.0025
BB=
0.0025
CA=
0.00267
CB=
0.00267
K2= 1 1.832239
0
K3= 1 1.832239
0
0
K2E = 1 1.832239
ER3 =
1,7
HA=
0.026
HB=
0.03
K4= 73.887081 33.083785
KLA = 38.831487 I 1.589686
KLB= 38.831487 1 1.589686
ZCA = 27.127291 "6.0897074
ZCB= 27.127291 "6.0897074
ZHA= 1000000
0
ZHB =
5.69013 "207.81694
FR= 433000000.
APL HERE
VERSION; J0/01 / 8 J
)READ NFCL
)READ CUBIC
)READ G73
ZV=0
RV = 0.008, 0.02, 0.03, 0.05
F1=®=CFIELDR NEFC4 Z
&: 40
0
0.008
13.552
2.3705
0
0.02
5.08206
3.02987
0
0.03
2.70622
"2.63237
0
0.05
0.90898
"1.29678
ZIN 1
7.3241
"17.4869
ZIN 2
9.03529
"8.98447
SEF 0
0
21.3203
1.4686
)WRITE OUTC; F1
)EXIT
22.6617
1.08772
0.510743
0.298135
0.949872
0.125469
1.40871
2.52337
0.0281753
"0.331 1 12
968.107
184.84
25.9162
7.33936
0.827038
188
TNSTTLATED ANTENNA THEORY PARAMETERS:
915 MHz
)read g74
lisglob
SI UNITS;
AA
0.00233 SECTION A CONDUCTOR RADIUS
AB
0.00233 SECTION B CONDUCTOR RADIUS
BA
0.0025 SECTION A DIELECTRIC RADIUS (AIR)
BB
0.0025 SECTION B DIELECTRIC RADIUS (AIR)
CA
0.00267 SECTION A TUBE RADIUS
CB
0.00267 SECTION B TUBE RADIUS
ER3
1.7 RELATIVE DIELECTRIC CONSTANT OF TUBE
HA
0.026 SECTION A LENGTH
HB
0.03 SECTION B LENGTH
K2
25.0035
0 ANGULAR WAVENUMBER OF DIELECTRIC (AIR): B+IA
K3
25.0035
0 ANGULAR WAVENUMBER OF TUBE
K2E
25.0035
0 EFFECTIVE ONE-LAYER DIELECTRIC WAVENUMBER
K4
142.805
40.4723 ANGULAR WAVENUMBER OF MEDIUM
KLA
70.881
23.4631 ANGULAR WAVENUMBER OF CURRENT IN SECTION A
KLB
70.881
23.4631 ANGULAR WAVENUMBER OF CURRENT IN SECTION B
ZCA
22.7957
"8.69274 CHARACTERISTIC IMPEDANCE OF SECTION A: R+JX
ZCB
22.7957
"8.69274 CHARACTERISTIC IMPEDANCE OF SECTION B
ZHA 1000000
0 TERMINATION IMPEDANCE OF SECTION A
ZHB
8.15848
"43.5481 TERMINATION IMPEDANCE OF SECTION B
FR 915000000.FREQUENCY
APL HERE
VERSION: 10/01/81
)READ NFCL
)READ CUBIC
)READ G74
ZV=0
RV = 0.008, 0.02, 0.03, 0.05
F1=®=CFIELD
R NEFC4Z
40
0
0.008
25.1576
"2.05095
1.80715
0
0.02
8.9141 1 "0.345986
0.588316
0
0.03
4.60219
1.03703
0.260682
0
0.05
1.41475
"2.47809 0.0576171
ZIN 1
14.738 "0.607589
ZIN 2
27,8423
"2.61002
SEF 0
0
43.1958
2.97097
)WRITE OUTC; F1
)EXIT
29.9106
0.131642
"2.85806
"1.70922
"0.61958
1.81825
2760.52
636.171
79.8074
21.2481
2.00483
189
INSULATED ANTENNA THEORY PARAMETERS:
2450 MHz
)read g75
lisglob
SI UNITS;
AA
0.00233 SECTION A CONDUCTOR RADIUS
AB
0.00233 SECTION B CONDUCTOR RADIUS
BA
0.0025 SECTION A DIELECTRIC RADIUS (AIR)
BB
0.0025 SECTION B DIELECTRIC RADIUS (AIR)
CA
0.00267 SECTION A TUBE RADIUS
CB
0.00267 SECTION B TUBE RADIUS
ER3
1.7 RELATIVE DIELECTRIC CONSTANT OF TUBE
HA
0.026 SECTION A LENGTH
HB
0.03 SECTION B LENGTH
K2
66.9492
0 ANGULAR WAVENUMBER OF DIELECTRIC (AIR): B+IA
K3
66.9492
0 ANGULAR WAVENUMBER OF TUBE
K2E
66.9492
0 EFFECTIVE ONE-LAYER DIELECTRIC WAVENUMBER
K4
357.076
59.8629 ANGULAR WAVENUMBER OF MEDIUM
KLA
148.571
61.9656 ANGULAR WAVENUMBER OF CURRENT IN SECTION A
KLB
148.571
61.9656 ANGULAR WAVENUMBER OF CURRENT IN SECTION B
ZCA
15.3535
"8.63907 CHARACTERISTIC IMPEDANCE OF SECTION A: R+JX
ZCB
15.3535
"8.63907 CHARACTERISTIC IMPEDANCE OF SECTION B
ZHA 1000000
0 TERMINATION IMPEDANCE OF SECTION A
ZHB
21.6237
"138.013 TERMINATION IMPEDANCE OF SECTION B
FR 2450000000. FREQUENCY
APL HERE
VERSION: 10/01/81
)READ NFCL
)READ CUBIC
)READ G75
ZV=0
RV = 0.008, 0.02, 0.03, 0.05
F 1 =@::CFIELD
RNEFC4 Z
<S>\ 40
0
0
0
0
0.008
0.02
0.03
0.05
25.4846
5.60972
2.01721
0.363381
ZIN 1
14.7732
"9.9249
ZIN 2
14.6503
"8.4915
SEF 0
0
60.4056
2.93359
)WRITE OUTC; F1
)EX1T
"0.424711
0.849064 "2.81395
650.186
"2.41528
0.38131 1
0.723622
31.6143
1.2053
0.16726
"2.59823
4.0971
2.25124 0.0276668
"2.53348
0.13281 1
40.1885
0.575458
5263.95
Fig. 6a: Exponential cun/e fit, 433 MHz
120
QC
<
40-
V)
20102.15 * 10»(-3.57S3e-2x)
0
2
4
6
H«2 - 0.250
8
10
12
14
16
Radial distance (r, mm)
Fig. 6b: Exponential cun/e fit, 915 MHz
300
DC
<
100-
in
332.98 * 10'*(-5.44188 2x)
0
2
4
6
R"2 - 0.634
8
10
12
14
16
Radial distance (r, mm)
Fig. 6c: Exponential cun/e fit, 2450 MHz
800
O)
jg 600E
E
5
400-
cc
<
200y - 1731.9 • 10»(-0.10168x)
0
2
4
6
R«2 . 0.787
8
10
12
Radial distance (r, mm)
14
16
120
Fig. 7(a) Antenna F12-2, Z=0, 433 MHz
•
# Measured data
5 100
o>
80
E
E
60
QC
<
40
•
X.
•
•
*
•
__^ulated antenna theory
tn
*
Bare antenna theory^^""*"-^
•
'
—
*
204
#
0
3
6
9
15
12
Radial distance, r (mm)
Fig. 7(b) Antenna FI2-2, Z=0, 915 MHz
300-
• Measured data
*
CD
J/
E
E
200-
*
•
N. #
Insulated antenna theory
#
S.
•
cr 100
<
(/)
1
Bare antenna theory
—,
*
0
3
6
9
12
Radial distance, r (mm)
Fig. 7(c) Antenna Fl 2-2, Z=0, 2450 MHz
1600-
o>
1200-
E
E
800-
• Measured data
Insulated antenna theory
tr
<
(T)
*
Bare antenna theory
3
6
9
Radial distance, r (mm)
15
192
THAYER SCHOOL OF ENGINEERING
D
A R
T
M
0
U
f
H
C
O
L
L
E
G
E
MEMORANDUM
TO:
Drs. Heaney, Douple, Trembly, Coughlin, Taylor, Stafford, Hoopes,
Thomas Ryan
FROM:
Terry Wong- and Eirik Jonsson
DATE:
December 11,1989
SUBJECT: Summary of Dog prostate hyperthermia (10/9/89)
OBJECTIVES
1)
2)
3)
Measure temperature distributions resulting from the transurethral microwave antenna in
vivo.
Gain experience with the surgical technique required for future toxicity studies
Histological examination of the heated tissue for evaluation of thermal damage and
locaUzation of thermometry.
MATERIALS AND METHOPS
PROSTATE APPLICATOR:
The antenna used was prototype F12-2. This antenna was built around a Bard siliconecoated 12 Fr foley catheter with 5 cc balloon. The drainage channel was taken up by the
antenna feedline and sealed. Therefore, the catheter did not drain the bladder. The antenna was
equipped with a 19 AWG thermometry catheter which was attached along the outside of the
antenna insulating layer with silicone sealant. Some of the relevant antenna parameters:
Parameter
2a
2c
c/a
h
Antenna element O.D.
Overall O.D. (w/o therm, cath.)
Insulation ratio
Section length
Section A
Section B
4.75 mm
5.72 mm
1.20
2.6 cm
4.57 mm
5.33 mm
1.17
3.0 cm
Junction location: 3.5 cm proximal to balloon edge
Junction width: 5-6 mm
EOUIPMENT:
American Microwave Technologies 915 MHz microwave power source
HP 435 Power meters (one each for forward and reflected power)
HP 778D Dual-directional coupler
DOG PROSTATE HYPERTHERMIA 11/9/89
193
Luxtron 3000 fiberoptic thermometry system, calibrated at 42.0°C
In-house IBM data acquisition and temperature mapping system
PROCEDURE:
A 28 kg dog was anesthetized using an appropriate induction dose of pentobarbital and
placed in a supine position. The urethra was filled with K-Y jelly using a 5 cc syringe and an
18 ga angiocath. A 6 Fr filiform catheter was inserted and followed by a 14 Fr follower. A 3
cm midline incision was made in the perineum, over the palpable urethral catheter at the
junction of the inferior pubic rami and the symphysis pubis. The urethra was exposed, and a
longitudinal incision made creating a 1 cm opening. The urethra was secured using four 4-0
silk stay sutures, and the filiform catheter and follower removed. The transurethral antenna was
lubricated generously with K-Y jelly and easily inserted into the bladder. The foley balloon was
inflated with 5 cc water and pulled gently to the bladder neck. 10 W of forward power was
then applied to the antenna (reflected power was 90 mW) and maintained for one half hour,
temperature maps were made along the antenna at 5 minutes and 30 minutes. The maximum
temperature occurred at the juncion, as expected. The power was then adjusted to maintain the
junction at 45°C (this required about 14-16 watts input) for another 30 minute period.
Transurethral temperature distributions were measured at 20 and 30 minutes.
After these initial transurethral temperature maps were completed, microwave pov/er was
discontinued and further surgery was performed to expose the prostate. An incision was made
extending from the mid-abdomen to the symphysis curving along the right side of the penis.
The symphysis was exposed and the prostate bluntly dissected Aom underneath the the upper
edge of the symphysis pubis. The prostate was approximately 2 cm long and 1.5 cm in
diameter. Two 18 ga plastic pointed blind-end thermometry catheters were inserted into the
prostate perpendicular to the urethra in an anteroposterior direction. Temperatures at the
junction were maintained at 45°C for 20 minutes, at which time temperature distributions were
obtained from all three catheters. Power was then increased to 15 W to obtain a larger
therapeutic volume, and temperature distributions measured measured after 20 minutes.
The superior and inferior pubic rami were divided using a bone cutter, permitting the
entire symphysis to be removed. This yielded excellent exposure, and the bladder, prostate,
and urethra (up to the perineum) were removed en block. The thermometry catheters were
clipped but not removed, and the complete block of tissue was immediately placed in a formalin
solution. The dog was euthanized humanely with a standard solution.
The diagram below (not drawn to scale) shows the orientation of the thermometry
catheter (dark line) alongside the transurethral applicator. Distances in mm are shown which
correspond to the distance scale used in the transurethral thermal maps (Figures 1, 2, and 5),
and represent the distance from the proximal balloon edge.
35
4 9
41
72
1
90
i
Distance from
balloon edge (mm)
>
Section A
Foley Balloon
Section B
Tissue boundary
DOG PROSTATE HYPERTHERMIA 11/9/89
194
Measurements made after removal of the bladder and prostate revealed that the prostate
was small, approximately 1.5 cm in diameter and 2 cm in length. The cephalic edge of the
prostate was measured to be 1 cm from the balloon edge; therefore, we conclude that the
prostate surrounded section A of the antenna and extended from 10 mm to 30 mm in the
temperature distribution plots. The junction of the antenna as well as section B were in the
distal urethra.
Bladder
Antenna tip
Prostate
d = distance from antenna catheter
surface to thermometry catheter
Z = distance from antenna junction to
thermometry catheter
dL
Section A
Section B
= ?
do
'R = ?
Zl = 1.43 cm
Antenna Junction
Zpj = 1.93 cm
Urethra
The diagram above illustrates the dimensions measured after the bladder, prostate, and
urethra were removed en block after the completion of the experiments. L and R represent the
left and right thermometry catheters inserted into the prostate. 10% was added to the measured
values to account for fixation shrinkage. There was a 1 cm length of gross hemorrhage noted at
the junction region of the antenna, and the center of this region was used to determine distances
l and r. The distances dL and dR will be measured after the histological cuts are made.
Z
Z
DOG PROSTATE HYPERTHERMIA 11/9/89
195
The table below summarizes the thermometry results which are included on the following
pages:
MAP LABEL
DESCRIPTION
Figure 1
MAP 2:26
MAP 2:53
Transurethral map, after 5 min @ 10 W
Transurethral map, after 30 min @ 10 W
Figure 2
MAP 3:28
MAP 3:38
Transurethral map, max controlled @ 45°C for 20 min^
Transurethral map, max controlled @ 45°C for 30 min^
Figure 3
MAP5:07R
MAP5;36R
R side of prostate, max controlled @ 45°C for 20 min^
R side of prostate, after 20 min @ 15 W
Figure 4
MAP5:07L
MAP5:36L
L side of prostate, max controlled @ 45°C for 20 min^
L side of prostate, after 20 min @ 15 W
Figure 5
MAP 5:07F
MAP 5:36F
Transurethral map, max controlled @ 45°C for 20 min^
Transurethral map, after 20 min @ 15 W
Notes: ^Required 14-16 W forward power
^Required 10-12 W forward power
Figure 1, 2, and 5 are temperature distributions measured alongside the transurethral
applicator. Figures 3 and 4 are temperature distributions made in the prostate through
perpendicular catheters L and R implanted for the second phase of the. experiment.
DOG PROSTATE HYPERTHERMIA 11/9/89
196
FIGURE 1: Transurethral temperatures, 10W input
MAP 2:26
MAP 2:53
Section A
Section B
4
6
Distance (cm)
0
8
FIGURE 2: Transurethral temperatures, max at 45°C
46
—a
o
o
10
MAP 3:2B
MAP 3:3 3
44
2
3 42
5
2
(D
E 40
Se ction A
Sectior B
ma
<
38
0
4
6
Distance (cm)
8
10
DOG PROSTATE HYPERTHERMIA 11/9/89
197
FIGURE 3: Thermometry catheter in prostate (right side)
43
42
o
o
41
1
11 • • ...
1
1
1
/
1
2
l/
E
39
•
MAP 5:1)7 R
r
MAP 5:36 R
i
1
s
1
S> 40
'
0
Q.
0
\
1
/
1
1
1
1
38
Prosta ;e
\i
1
\
1
37
0
h
1
2
Distance (cm)
FIGURE 4: Thermometry catheter in prostate (left side)
42
41
1
1
1
^
^
'' V'
11
1
1
40
1 39
1
1
1
1
1
1
0
Q.
1
1 38
1
1
p
rostate
1
j-x
37
0
1
2
Distance (cm)
1
—
MAP 5:07 L
»—
MAP 5:36 L
E
1
O
0
1
DOG PROSTATE HYPERTHERMIA 11/9/89
198
FIGURE 5: Transurethral temperatures at 20 min.
49
47
MAP 5:07 F
MAP 5:36 F
P 45
£3
4-«
s
o
Q.
E
Section A
0
2
Sectioih B
4
6
Distance (cm)
8
10
DISCVSSKON
Unfortunately, the prostate in this particular dog was very small, and most of the heating
volume was distal to the prostate. The transurethral temperature distributions (Figures 1, 2, and 5)
demonstrate that the antenna heats generally as expected with the maximum clearly at the junction
region (there is some measurement error apparent in the figures probably due to small slippage in the
stepper motor drive).
The temperatures near the tip of section A show a reproducible dip possibly consistent with a
nearby blood vessel; this part of section A was within the prostate. This dip was not seen in
preliminary liquid crystal SAR patterns of a similar foley antenna, but has been seen in other antennas.
Quantitative SAR measurements will be run along this antenna to determine if this is due to an SAR
dip.
Section B of the antenna showed an excellent temperature distribution which dropped off
sharply at the proximal end, indicating an effective choke and confining the hyperthermia to the desired
region. With a maximum (junction) temperature of 45°C, a length of approximately 3 cm is heated
along the urethra to > 43°C, and approximately 4.5 cm is heated to ^ 42°C.
In Figures 3 and 4, temperatures are shown within the prostate along catheters L and R. Note
that this is in the "dip" part of the transurethral temperature distribution and are thus relatively low.
Overall, the distal location of these measurements along section A means that they represent a nearly
worse-case situation. Maximum temperatures of 42°C to 43°C were only attainable if the junction
temperature was around 48°C. Evidently from these distributions, one half of the prostate was more
effectively heated than the other. This is probably attributable to boundary effects: the proximal part of
the thermometry catheter (-2-4 cm) was packed with moist gauze which prevented heat dissipation. It
is also difficult to draw conclusions without knowing dL and dR.
DOG PROSTATE HYPERTHERMIA 11/9/89
199
We are awaiting histological results along the urethra and within the prostate. There
was a region of hemorrhage noted on the gross section along the urethra at the antenna
junction. It should be noted that the thermal dose at this point exceeds the equivalent of 24
hours (1440 minutes) at 43°C.
CONCLUSIONS
1)
The transurethral antenna built around a 12 Fr. Foley catheter can be easily
accommodated in the dog using the perineal urethrostomy technique for access.
2)
The prototype transurethral antenna generally performs as expected in terms of
thermal distributions in vivo. The temperature dip along section A of the antenna is
probably due to blood flow. However, quantitative SAR measurements need to be
done on this antenna.
3)
The dog prostate model in this case was too small for efficient application of the
antenna prototype.
4)
The present antenna design effectively confines hyperthermia to the region of interest
with minimal feedline heating. Reflected power was low (about 1%) at 915 MHz.
5)
The transurethral temperature catheter is easily adapted to TP Ryan's stepper motor
sweep system, which will prove to be invaluable during clinical treatments with this
antenna.
6)
Histology results are pending.
Histology addendum: There was an area of hemorrhage grossly visible in the periurethral
region where the antenna junction had been located. Histological findings included mild
submucosal edema with hemorrhage and loss of the adjacent urothelium. The basement
membrane was largely intact. Overall, the changes were mild in view of the large thermal
dose delivered to this area.
200
T HAYER S CHOOL OF ENGINEERING
D
A R
T
M
O
^
U
T
H
C
O L L E G E
Terence Z. Wong, M.S.
(603) 646-8248
MEMORANDUM
TO:
Drs. Heaney, Jonsson, Douple, Trembly, Coughlin, Taylor, Stafford,
Hoopes, Thomas Ryan
FROM:
Terry Wong
DATE:
March 12,1990
SUBJECT: Dog Experiment (3/9/90)
OBJECTIVES
From the first dog experiment on 11/9/89, it has been concluded that the Transurethral
Microwave Antenna CTMA) performed well at 915 MHz. The urethra was examined
histologically along its length, and it was concluded that only minimal damage occurred even in
the areas of very high thermal dose. The study was limited, however, because that particular
dog had a very small prostate which was inadequate for measurement of temperature
distributions. The primary objective of this experiment then was to measure temperature
distributions in the prostate.
The Transurethral Microwave Antenna (TMA) was designed for optimum performance at
915 MHz, which is suitable for heating lengths of about 4 cm. However, SAR measurements
on the TMA demonstrate that it can be driven efficiently at lower and higher frequencies (i.e.
433 MHz and 2450 MHz). These studies suggest that 433 MHz may be useful for treating
longer lesions along the urethra (6-7 cm), while 2450 MHz may be used to treat smaller lengths
(2-3 cm). Therefore, the second objective of this experiment will be to measure the temperature
distributions along the antenna and in the prostate with the antenna driven at 433 MHz, 915
MHz, and 2450 MHz.
MATERIALS AND METHODS
PROSTATE APPLICATOR:
The antenna used was prototype F12-2 (the same antenna used in the first dog
experiment). This antenna was built around a Bard silicone-coated 12 Fr foley catheter with 5
cc balloon. Please refer to the data from the first dog experiment for dimensions and other
details (memo 12/11/89).
DOG PROSTATE HYPERTflERMIA 3/9/90
201
EOIJIPMENT:
American Microwave Technologies 915 MHz microwave power source
M/A Com 433 MHz generator
Cheung Labs 2450 Wiz generator
HP 435 Power meters (one each for forward and reflected power)
HP 778D and 777D Dual-directional couplers
HP calibrated attenuator (-lOdB)
Luxtron 3000 fiberoptic thermometry system, calibrated in 42.0°C water bath
In-house IBM-based data acquisition and temperature mapping system
PROCEDURE:
A 41 kg dog was anesthetized with pentobarbital, dosed to the desired effect. A perineal
urethrostomy was performed in the same manner as in the previous experiment (see 12/11/89
memo). The bladder and prostate were exposed using an anterior approach. The prostate was
found to be very small and barely palpable; it was too small to allow insertion of any
thermometiy catheters. Therefore, we decided to measure transurethral temperatures along the
antenna only for each of the different driving frequencies.
All temperatures were obtained using the automatic mapping system developed by
Thomas Ryan et al. For each frequency, baseline sweeps were made prior to applying power.
After baseline temperatures were confirmed to be uniform, power was applied for 10-15
minutes. The maximum temperatues were found to be at approximately 30 mm from the end of
the thermometry catheter, so this was taken to be the reference point. At each frequency, power
was adjusted to maintain the temperature between 47°C and 48°C at the reference point. After
10-15 minutes, temperature distributions were measured.
RESULTS
Thermometry measurements were obtained along the urethra using the 20 AWG catheter
which is attached to the antenna with silicone adhesive. The important measurement landmarks
along the thermometry catheter are shown below:
6 7
V
69
113
T
I
Distance from
balloon edge (mm)
1
Section A
i
Section B
Foley Balloon
(not to scale)
Tissue boundary
The power was controlled at each frequency to maintain a control temperature of between
47°C and 48°C at 30 mm from the edge of the Foley balloon. The data summarized in the
technical memorandum dated 2/22/90 was used to calculate the net forward power at the
antenna junction at each frequency:
Frequency
433 MHz
915 MHz
2450 MHz
Indicated Power
Forward
Reflected
21.0 W
13.5 W
12.0 W
0.48 W
0.24 W
0.72 W
Calculated Power at Junction
Forward
Reflected
ZM
13.5 W
11.2 W
5.95 W
0.51 W
0.35 W
1.03 W
13.0 W
10.8 W
4.92 W
DOG PROSTATE HYPERTHERMIA 3/9/90
202
From the above table, the antenna clearly required less power at higher frequencies. This
makes intuitive sense, because a larger volume is heated at the lower frequencies. This is also
consistent with the SAR experiments which have been done on this particular antena.
The temperature distributions at these frequencies are shown in figures 1 and 2. In
general, the heating length increases with decreasing driving frequency as expected. Figure 1
shows the raw ste^y-state temperature data as recorded for each frequency. In figure 2, the
baseline temperature Ty was taken to be 37.9°C, which was confirmed before applying power
at each frequency. The normalized temperature Tnorm was then calclulated from the measured
temperature T:
T
= T-Tb
^ norm rp
rp
Amax"
This data compensates for the small differences in the measured Tmax at each frequency and
simplifies the intercomparison.
PISCUSSION
In general, the behavior of the antenna is consistent with the previous dog experiment of
11/9/89 and with the longitudinal and transverse SAR experiments which have been performed
on this antenna. The SAR and in vivo data suggest that this antenna can be operated at any of
these frequencies, and that the treatment length along the antenna is a function of the driving
frequency.
There are local peaks which occur at the terminal ends of each antenna section (especially
the antenna tip), particularly at 433 MHz and 915 MHz. The reason for this is not clear, and
they were not apparent in the previous experiment. One possibility is that the tip of the antenna
was in the bladder and in contact with urine that had been heated over time during the
experiment. Another possibility is that the terminal end of the antenna was not perfectly sealed,
and the tip of the antenna section was therefore not perfectly insulated.
CONCLUSION
It was very unfortunate that this dog had only a very small prostate. However, we were
able to obtain longitudinal temperatures from the TMA at three frequencies in vivo. As with the
first dog, we will examine the urethra histologically for evidence of mechanical and/or thermal
damage.The next step will be to find a large dog with a big prostate to measure temperature
distributions in the prostate at the three frequencies.
DOG PROSTATE HYPERTHERMIA 3/9/90
203
Figure 1: Temperature along catheter
48
433 MHz
915 MHz
2450 MHz
46
0
V 44
1
o 42
E '
H 40
SMion 6
38
0
2
4
6
Distance (cm)
8
10
Figure 2: Normalized temperature
1.00
433 MHz N
915 MHz N
2450 MHz N
^ 0.80
<3
•§ 0.60
N
"cB
E 0.40
o
z
0.20
Seel ion A
Section B
0.00
0
2
4
6
Distance (cm)
8
10
204
T HAYER S CHOOL OF ENGINEERING
D A R T M O U T H
C
O L L E G E
Terence Z. Wong, M S .
(603) 646-8248
MEMORANDUM
TO:
Drs. Heaney, Jonsson, Douple, Trembly, Coughlin, Taylor, Stafford,
Hoopes, Thomas Ryan
FROM:
Terry Wong
DATE:
April 24,1990
SUBJECT: Dog Experiment (3/26/90)
ORIRCTIVES
From the first dog experiment on 11/9/89, it has been concluded that the
Transurethral Microwave Antenna (TMA) performed well at 915 MHz.and resulted in only
minimal histological damage even in the areas of very high thermal dose. During a second
experiment (3/9/90), temperature distributions along the TMA were measured with the
antenna operating at different frequencies, but the prostate was too small to implant with
thermometry catheters. The primary objective of this third experiment was to select for an
aged uncastrated dog and measure temperature distributions in the prostate at 433,915, and
2450 MHz driving frequencies.
MATERIALS AND METHODS
PROSTATE APPLICATOR:
The antenna used was prototype F12-2 (the same antenna used in the first two dog
experiments). This antenna was built around a Bard silicone-coated 12 Fr foley catheter
with 5 cc balloon. Prior to the experiment, the area near the tip of section A was resealed
with silicone sealant. The thermometry catheter was also repositioned (it tends to shift
slightly after each experiment) and attachment to the Foley antenna reinforced with silicone
sedant. The TMA is diagrammed in figure 1 with relevant landmarks.
DOG PROSTATE HYPERTHERMIA 3/26/90
205
EOIJTPMENTR
American Microwave Technologies 915 MHz microwave power source
M/A Com 433 MHz generator
Cheung Labs 2450 MHz generator
HP 435 Power meters (one each for forward and reflected power)
HP 778D and 777D Dual-directional couplers
HP calibrated attenuator (-lOdB)
Luxtron 3000 fiberoptic thermometry system, calibrated in 42.0°C water bath
IBM-based data acquisition and temperature mapping system
PROCEDURE:
The dog weighed approximately 60 lbs (27.3 kg) and was anesthetized in the usual
manner with pentobarbital. Rectal examination under anesthesia revealed a good-sized
prostate. A perineal urethrostomy was performed as in the previous experiments, and a
long midline incision made in the abdomen to expose the bladder and prostate. The urethra
was narrowed, making it difficult to insert the TMA. The bladder was opened and an 8 Fr
catheter passed through the urethra from the bladder. Suture was tied around the TMA, and
the 8 Fr catheter was used to aid in pulling the TMA through. The TMA Foley balloon was
inflated with water, and good position within the prostate was verified by feeling the
antenna through the prostatic tissue and noting that the balloon was properly seated in the
bladder neck. Two thermometry catheters were implanted in the prostate perpendicular to
the TMA in an anteroposterior direction (posterior = dog's spine). The tips of these
catheters protruded 2-3 mm from the posterior side of the prostate, and this was verified by
feeling the tips by a rectal exam. The catheter orientation and locations are shown in figure
2.
All temperatures were obtained using the automatic mapping system developed by
Thomas Ryan et al. For each frequency, baseline sweeps were made prior to applying
power. After baseline temperatures were confirmed to be uniform, power was applied for
15-20 minutes. Based on a temperature map made during the initial heating, a reference
point was chosen along the TMA thermometry catheter.35 mm from the tip. Input power
was adjusted to maintain this reference point between 47°C and 48°C. After 15-20 minutes,
steady-state temperatures were measured along each thermometry catheter. Temperatures
were allowed to equilibrate 20-30 minutes between hyperthermia sessions.
RESULTS
Thermometry measurements were obtained along the urethra using the 20 AWG
catheter which is attached to the antenna with silicone adhesive. The important
measurement landmarks along the thermometry catheter are shown below:
3 6
33
Section A
Foley Balloon
38
68
145
1
Distance from
balloon edge (mm)
Section B
(not to scale)
Tissue boundary
As with the previous experiment, the power was controlled at each frequency to
maintain a control temperature of between 47°C and 48°C at a reference point at the junction
DOG PROSTATE HYPERTHERMIA 3/26/90
206
35 mm from the edge of the Foley balloon. The data summarized in the technical
memorandum dated 2/22/90 was used to calculate the net forward power at the antenna
junction at each frequency:
Frequency
433 MHz
915 MHz
2450 MHz
Indicated Power
Forward
Refiwtgd
23.0 W
20.0 W
14.0 W
0.67 W
0.58 W
0.73 W
Calculated Power at Junction
Forward
Reflected
14.77 W
16.51 W
6.93 W
0.71 W
0.85 W
1.05 W
14.1 W
15.7 W
5.9 W
As observed in the previous dog experiment, the antenna required significantly less
power at 2450 MHz. Presumably, this is because a shorter length is heated at this
frequency. This was shown by previous SAR experiments.
The steady-state longitu(Anal temperature distributions at the three frequencies are
shown in figure 3. In general, the heating length increases with decreasing driving
frequency as expected, although the effect is not as pronounced as in the previous
experiment, nor is it a dramatic as shown by SAR distributions. This suggests that the
prostate has relatively low blood flow.
Figure 4 shows the steady-state temperature distributions in left lobe of the prostate
perpendicular to the TMA. The catheter was inadvertently pulled back prior to the 2450
MHz temperature sweep, so that the distribution is shifted. We are also therefore not certain
of the distance between this thermometry catheter and the antenna, although it was
estimated to be 5 mm from the antenna surface after the prostate was cut. In figures 4 and
5, distance 0 refers to the posterior aspect of the prostate and temperatures were mapped in
an anterior direction.
Figure 5 shows steady state temperatures at the three frequencies obtained in the right
lobe of the prostate. This catheter was a minimum of 1.1 cm from the surface of the
antenna, as measured upon removal of the prostate. Therapeutic temperatures were
obtained at this location with a maximum urethrA temperature of about 48°C.
DISCUSSION
Again, the behavior of the antenna is consistent with the previous dog experiments
and with the longitudinal and transverse SAR experiments which have been performed on
this antenna. Temperatures measured in the prostate demonstrate that the antenna can
deliver therapeutic temperatures over a total diameter exceeding 2 cm in the prostate. In
general, the blood flow in the prostate appears to be relatively low. However, there was a
"dip" in the temperature distribution within the prostate (figure 5), which may be due to a
vessel or an area of high blood flow.
From the temperature distributions in the prostate, it appears that the antenna gives the
best radial distribution at 915 MHz and radiates least well at 433 MHz. The differences are
not nearly as great based on these steady-state temperatures as seen in previous SAR
experiments.
rONmiSTON
Temperature distributions were obtained along the transurethral antenna as well as in
the prostate for 433 MHz, 915 MHz, and 2450 MHz. The results were consistent with the
previous SAR measurements and animal experiments. Therapeutic temperatures can be
obtained at a distance over 1 cm away from the antenna in the prostate at all three
frequencies. This was accomplished with a maximum urethral temperature of 47°C to 48°C.
Clinically, it may be possible to use higher temperatures, because this would be in the
DOG PROSTATE HYPERTHERMIA 3/26/90
207
region of the hyperplasia (and this is the tissue that would be scraped out if a TURP were
done). Longitudinal temperature distributions show that the antenna radiation is limited to
the region of the antenna elements.
This experiment was well documented photographically (especially the complete
bladder / catheter / prostate / urethra specimen). Histological results are pending.
Interestingly, a region having the appearance of an old infarct was noted after the prostate
was cut. This was located along the right thermometry catheter on the anterior side, which
would correspond to the right third of figure 5.
DOG PROSTATE HYPERTHERMIA 3/26/90
208
FIGURE 2: Orientation of catheters
within prostate. The left catheter
slipped during the experiment, but
was estimated to be approximately
5 mm from the antenna surface.
mmMi
3.3 cm
1.4 cm
1.1 cm
Prostate
Section A
Section B
Figure 3; Temperatures along TMA
a—
Prostate
hi 40
Section A
Section B
40
Distance (mm)
915 MHz
•—
433 MHz
«—
2450 MHz
DOG PROSTATE HYPERTHERMIA 3/26/90
209
Figure 4: Temperatures, left catheter
915 MHz
42-
433 MHz
2450 MHz
u 41 -
0)
40 -A-
i 39-1
<D
Q. 38E
(D
X.
37
Prostate
36
T
0
44.
0
1
20
1
'
1
30
40
Distance (mm)
50
Figure 5: Temperatures, R catheter
915 MHz
433 MHz
42-
2450 MHz
40(U
Q.
E
<D
Prostate
20
30
40
Distance (mm)
60
Hyperthermia for BPH
210
Dartmouth-Hitchcock Medical Center
TRANSURETHRAL HYPERTHERMIA PROTOCOL FOR THE
TREATMENT OF BENIGN PROSTATIC HYPERPLASIA
John A. Heaney, M.D.
Terence Z. Wong, M.S.
Eirikur Jonsson, M.D.
Christopher T. Coughlin, M.D.
James Taylor, M.D.
James H. Stafford, M.D.
B. Stuart Trembly, Ph.D.
Stuart M. Selikowitz, M.D.
September 5,1989
Hyperthermia for BPH
211
TART.F. OF CONTENTS
1. Introduction
1.1 Benign prostatic hyperplasia
1.2 Treatment of BPH using hyperthermia
1.3 Technical aspects of prostatic hyperthermia
1.4 Animal studies
2. Aims
3. Eligibility
3.1 Eligible Patients
3
3
3
6
7
8
8
8
3.2 Ineligible Patients
4. Patient Entry
5. Treatment Schema
6. Toxicities
9
9
9
10
7.
8.
9.
10.
11.
11
12
13
14
15
Monitoring of Patients
Criteria Response
Statistical Considerations
Off Study Criteria
References
Appendix I -
Subjective symptom scoring
Appendix H - Toxicity symptom evaluation
Appendix IE - Treatment and Follow-up Flow Sheets
18
20
21
Hyperthermia for BPH
212
1. TNTRONNRTTON
1.1 Benign prostatic hyperplasia
Benign prostatic hyperplasia (BPH) is a common disease of elderly men; more than
90% of all men develop BPH by the eighth decade of life (3). It is the most common cause
of urinary obstruction in men, and 10%-20% of men will require prostatic surgery at some
time in their lives to relieve obstructive symptoms (3,10). The mean age for developing
symptoms from BPH is 65 for whites and 60 for blacks (3).
The treatment of choice for symptomatic BPH is surgery, for which there are
several major indications:
1) unacceptable decrement in urinary flow,
2) persistent residual urine,
3) acute urinary retention secondary to obstruction,
4) recurrent urinary tract infections secondary to obstruction,
5) hydronephrosis, and
6) acquired bladder diverticula and/or vesicular calculi.
The fact that many patients suffering from BPH are elderly and may not be candidates for
surgery suggests that nonsurgical alternatives for therapy warrant consideration. Non­
surgical treatments for BPH have included medications (alpha-adrenergic antagonists, 5-a
reductase blockers, hormones) and mechanical dilatation (10). Recently, there has been
evidence that hyperthermia may be a useful modality for the management of symptomatic
BPH (11,17,20,21,26).
1.2 Treatment of BPH usinp hyperthermia
Hyperthermia, the heating of tissues to temperatures ranging from 43°C to 60°C, is
being studied at Dartmouth-Hitchcock Medical Center (DHMC) and other institutions as an
adjunct to radiation therapy and chemotherapy for the treatment of cancer (22). In
particular, DHMC and the Thayer School of Engineering have developed an interstitial
microwave antenna array hyperthermia (IMAAH) system which has been used to treat
superficial and deep-seated tumors (4,5,15). Although the precise mechanism by which
hyperthermia causes cell death is not fully understood, it is now believed that heat disrupts
both the cellular membrane and nuclear function (16). Hyperthermia has been used in
treating the prostate for cancer (19,23,26,28,29), BPH (11,17,20,21,26), and chronic
nonbacterial prostatitis (20,21). These studies have demonstrated that hyperthermia may be
a valuable nonsurgical treatment for obstructive symptoms caused by these prostatic
diseases, and causes little morbidity.
Hyperthermia for BPH
213
Lindner et al. (11) treated 6 patients with symptomatic BPH requiring an indwelling
catheter. Previous attempts to wean these patients from their catheters had failed (following
a 5-day course of phenoxybenzamine). The patients were given 5-10 hyperthermia
treatments (1-2 treatments per week) using a microwave (915 MHz) water-cooled
transrectal applicator, and were followed for 6 months. Five of the six patients showed
subjective and objective improvement of their symptoms, and were able to be relieved of
their indwelling catheters. The only treatment failure was a patient with a large, tender
prostate which could not be treated effectively. No complications were observed in this
study.
The largest series of patients undergoing prostate hyperthermia is currently under
study by two investigators in Israel. To date, Servadio and Yerushalmi have used a watercooled transrectal microwave applicator to deliver 500 hyperthermia treatments to 74
patients with benign and malignant prostatic diseases (20). Yerushalmi et al. (27)
summarize their results for the treatment of BPH. In this study, 29 patients with severe
symptoms and for whom surgery was contraindicated were given an average of 14
hyperthermia treatments on a twice weekly schedule. Eleven of the patients had chronic
indwelling catheters. The patients were followed for up to 26 months and evaluated using a
scoring scale for symptoms of frequency, nocturia, urgency, and hesitancy. The authors
noted that symptoms markedly improved after 6-8 treatments, but they felt that a total of
12-15 treatments was optimum. All of the patients who did not have an indwelling catheter
showed symptomatic improvement, and eight of eleven (73%) patients who had indwelling
catheters resumed normal voiding, with a post-void residual of <60 ml. At 18 months
follow-up, none of the previously catheterized patients had developed recurrence of severe
obstructive symptoms or urinary retention. No side effects were noted in this study, and
the transrectal hyperthermia was found to cause no damage to the rectal mucosa.
Most recently, Astrahan and Sapozink et al. have developed a transurethral
microwave hyperthermia applicator (1,2,17) for the treatment of BPH. They have treated
21 men with a total of 177 hyperthermia treatments and a mean follow-up of 10 months. In
terms of objective response (residual urine volume and urine flow rate), 17/21 (81%) of the
patients showed improvement. In terms of subjective parameters (frequency and stream
force), 15/21 (71%) of the patients showed improvement.
Table 1 summarizes the results for hyperthermia as a treatment for BPH. Overall,
the clinical studies to date support the following conclusions;
1) Hyperthermia alone is effective in managing the symptoms of BPH in terms of
both subjective and objective parameters. Overall response rate of current studies is
84%.
Table 1. Hyperthermia for Benian Prostatic Hyperplasia - Summary of clinical data
^
I
Lindner et al.
Yerushalmi / Servadio et al.
Sapozink et at.
# Patients
# Treatments
6
47
29
340
21
177
Applicator
Frequency (MHz)
Rectal, water-cooled
915
Rectal, water-cooled
2450
Transurethral
635, 915
Treatment length
# treatments / patient
60 min.
5-10
45 - 50 min.
7-18
(?)
8.4 (avg.)
Temperature
Thermal dose equivalent
39.4° - 45.2°C
131-510 (total)
42° - 43°C
14-50 (per treatment)
44° - 46°C
(?)
Follow up period
6 mo.
2 - 26 mo., (9.7 avg.)
10 mo. (avg.)
Response (subjective)
Response (objective)
Catheter removed
5/6 (83%)
5/6 (83%)
5/6 (83%)
26/29 (90%)
15/21 (71%)
17/21 (81%)
Oyerall response rate:
Pts. with catheter:
(47/56)
(13/17)
•
8/11 (73%)
84%
76%
N)
4^
Hyperthermia for BPH
215
2) Hyperthermia may be useful in patients requiring an indwelling catheter, as 76%
of these patients were able to have their catheters removed after treatment.
3) Twice weekly treatments with a total of 10-14 hyperthermia sessions has
resulted in the above response rates.
4) Complications and toxicities have been minimal.
1.3 Technical aspects of prostatic hvperthermia
The prostate is accessable through two intracavitary approaches: hyperthermia
applicators may be inserted either through the rectum or through the urethra. The transrectal
approach is preferable for treating prostatic cancers for several reasons: 1) the rectum can
accommodate a larger applicator which in turn can heat a larger volume of tissue, 2) most
prostatic cancers lie in the posterior portion of the prostate and are accessible transrectally,
and 3) the urethra offers an easily accessible central location in the prostate for monitoring
temperature. All transrectal applicator designs feature water cooling to prevent damage to
the rectal mucosa. Technical details of the transrectal microwave hyperthermia applicators
used in clinical studies are provided by Mendecki et al. (13), Petrowicz et al. (14), and
Scheiblich and Petrowicz (18).
Although the transrectal hyperthermia applicators are advantageous for treating large
prostatic lesions such as cancer, the transurethral approach has significant advantages for
treating BPH. First, the water-cooled transrectal applicator delivers a maximum temperature
several millimeters beneath the rectal mucosa and in the posterior prostate, whereas the
transurethral microwave applicator delivers maximum temperature periurethrally and
concentrates the hyperthermia around the symptomatic lesion. Secondly, the transurethral
applicator can be easily localized within the prostate using a balloon catheter and/or imaging
techniques; the transrectal applicator must be properly "aimed" at the prostatic lesion.
Finally, the transurethral approach is less likely to cause complications resulting from
damage to the rectal mucosa. Astrahan et al. (1,2) have recently described a transurethral
microwave applicator consisting of a modified Foley catheter to which three microwave
antenna catheters are attached. This apparatus permits hyperthermia to be delivered to the
prostate using standard interstitial microwave antennas, such as those used at DartmouthHitchcock. One disadvantage of the small interstitial antennas is that they tend to heat
proximal to the junction along their feedlines; this means that parts of the urethra distal to
the prostatic urethra may be heated.
A new transurethral microwave antenna has been developed at DartmouthHitchcock which is also designed around a standard Foley catheter, but has several
advantages over the 3-antenna Astrahan design. First, the microwave antenna and
Hyperthermia for BPH
216
associated fiberoptic thermometry cannulae are built directly onto the catheter.
Measurements of specific absorption rate (SAR) demonstrate that the single larger antenna
has a power deposition pattern which is more uniform than that produced by three smaller
antennas. Secondly, the new antenna features a choke design, which significantly reduces
heating along the distal urethra (25). Finally, the antenna elements are located adjacent to
the Foley balloon, assuring consistently accurate placement of the antenna within the
prostatic urethra.
The DHMC transurethral antenna is a linear dipole based on the antenna theory
developed at the Thayer School of Engineering by Trembly and King (8,24) for an
insulated antenna in a conducting medium. The antenna has been tested in muscleequivalent phantom materials using the same procedures which have been used to test the
previously designed interstitial and surface microwave applicators. Antenna impedance
measurements demonstrate that the antenna reflects little power at the design frequency of
915 MHz, and thus efficiently radiates the power into the surrounding tissue. The SAR
(power deposition) pattern generated by the transurethral antenna has been evaluated using
liquid crystal techniques and demonstrates that the antenna: 1) effectively radiates into the
tissue in the manner predicted by theory, and 2) provides a desirable radiation pattern
which is confined to the prostatic region of the Foley catheter.
Minor refinements are continuing on the DHMC transurethral antenna. First, the
flexibility of the antenna elements is being improved using a custom-designed braid
material. Second, DHMC is collaborating with a fiberoptic thermometry manufacturer
(Luxtron, Inc.) to develop a small fiberoptic temperature sensor to be built directly onto the
antenna. Finally, construction techniques are being refined to allow the entire antenna
system to be incorporated onto smaller Foley catheters. These refinements should improve
the clinical utility of the transurethral antenna without affecting the antenna design or
performance; however, SAR and impedance testing will continue on these antennas as
modifications are made. At this time the design of the DHMC transurethral antenna is
considered to be proprietary while the possibility of a patent is being explored.
1.4 Animal studies
The effect of hyperthermia on the prostate has been previously studied in dogs and
rabbits. Leib et al. (9) used a 915 MHz water-cooled transrectal antenna to treat 20 dogs.
The dogs were divided into 7 treatment groups with various treatment schedules and
temperatures ranging from 40°C to 47°C. Histological examination of the prostate was
performed either immediately or 1 week after hyperthermia. They concluded that
hyperthermia treatments of 42.5°C for 90 minutes resulted in no histological damage after 6
Hyperthermia for BPH
217
treatments and could be repeated safely. Yerushalmi et al. (30) used a 2450 MHz transrectal
applicator without water cooling to treat 64 rabbits in a preclinical study. The rabbit
prostates were given either one or two hyperthermia treatments (42.6®C - 42.8°C) for 30
minutes. Histological examination took place at intervals ranging from 1 day to 3 months
after treatment, and demonstrated this dose of hyperthermia to cause little damage. At the
other extreme, Magin et al. (12) treated 8 dogs with 15 minute treatments of 60°C
hyperthermia (equivalent to treating continuously at 43°C for 3.8 years (7)) and found total
destruction of the prostate.
Astrahan et al. (1) has measured temperature distributions resulting from their
transurethral microwave antenna in a dog. Temperatures were measured longitudinally
along the antennas as well as in a radial direction by surgically exposing the prostate and
implanting multiple temperature probes. For the transurethral antennas operating at 915
MHz, they found that with the maximum temperature fixed at 46°C, the therapeutic
temperature region (43°C to 46°C) extended approximately 3.5 cm in the longitudinal
direction and 0.2 to 0.5 cm in the radial direction.
2. AIMS
2.1 To evaluate the feasibility, toxicity, and efficacy of transurethral hyperthermia
as a treatment for benign prostatic hyperplasia in patients for whom surgery is
contraindicated, who refuse surgery, or who wish a trial of a non-surgical
intervention prior to the surgical treatment.
3. ELIGIBILITY
3.1 Eligible Patients
3.11
3.12
Men with benign prostatic hyperplasia having severe symptoms and for
whom TURF is recommended, but who refuse surgery or are not
candidates for surgery. Patients are considered severely symptomatic if 1)
they score one or more on at least three of the obstructive symptoms listed
in Appendix I, and 2) urinary peak flow is <12 ml/sec.
Men with symptomatic benign prostatic hyperplasia who desire an
alternative treatment to surgery. It is important to note that a TURP may
still be performed in the future should hyperthermia fail to alleviate
symptoms.
Hyperthermia for BPH
218
3.2 Ineligible Patients
3.21
3.22
3.23
Patients with previous prostatic surgery.
Patients with a history of fistula involving the urethra, rectum, or bladder.
Patients on hormonal therapy, steroids, or a-adrenergic medications.
3.24
Immunocompromised patients.
4. PATTFNT ENTRY
4.1 All patients considered for entry onto the study must meet the eligibility
criteria outlined in section 3.
4.2 All patients will undergo a complete history, including careful medication
history, and physical examination. Laboratory studies will include
urinalysis, CBC, electrolytes, BUN, creatinine, and prostate specific
antigen levels.
4.3 Diagnosis of bladder outlet obstruction secondary to prostatic enlargement
will be confirmed by cystoscopic examination or other appropriate
studies. Baseline urodynamic studies will be obtained on all patients,
including maximum urinary flow rate and post-void residual volumes.
Pre-treatment symptomology scores will also be obtained at this time
(Appendix I).
4.4 The prostate will be examined digitally for evidence of carcinoma. If
necessary, ultrasound, CT, and/or needle biopsy will be used to rule out
prostatic carcinoma.
4.5 Prostatic volume will be assessed by transrectal ultrasound.
5. TREATMENT SCHEMA
Patients will receive hyperthermia using a treatment schedule based on that used
previously at other institutions (11,17,27). The transurethral microwave antenna may be
inserted either prior to each hyperthermia treatment, or may be left in place for patients who
require an indwelling catheter. Patients may be treated on an outpatient basis.
5.1 Each hyperthermia treatment will last a total of 60 minutes, and will be
delivered using the same computer-controlled microwave generating
equipment that is used at DHMC for treating superficial and deep-seated
tumors. Temperatures will be constantly monitored using fiberoptic
temperature probes along the transurethral antenna, and the microwave
power will be adjusted either manually or through automatic feedback
control to maintain therapeutic hyperthermic temperatures at the antenna.
Hyperthermia for BPH
219
As in other hyperthermia treatments, temperature data is recorded
automatically and plotted by computer. The target temperature along the
antenna will be a minimum of 43°C, but microwave power will be limited
so that no temperature at any point along the antenna exceeds SO°C. Initial
treatments will be conservative. Patient discomfort will always mandate
that the microwave power be reduced, even if this means that the target
temperature is not achieved. In this case, treatment will continue, but at
the lower temperatures tolerated by the patient
5.2 Hyperthermia treatments will be given twice per week with a minimum of
72 hom-s separating treatments to minimize thermotolerance (27). From
previous clinical experience, symptomatic improvement can be expected
after 6-8 treatments, and the optimum total number of treatments is 12-15
(27). Based on this finding, the first ten patients on this study will receive
a total of 12 hyperthermia treatments. After the statistical data have been
evaluated from these patients, the target number of treatments may be
adjusted accordingly.
5.3 Prior to each hyperthermia treatment, patients will be scored for subjective
irritative and obstructive symptoms (Appendix I), as well as toxicities
(Appendix II).
6. TOXICITIES
In general, localized hyperthermia has been tolerated well by patients here and at
other institutions, and complications are rare. Hyperthermia has been used to treat several
diseases of the prostate, including carcinoma, BPH, and chronic nonbacterial prostatitis. In
most of these studies, hyperthermia was delivered using a water-cooled transrectal
microwave applicator. Hyperthermia of the prostate has been well tolerated in previous
clinical studies with few side effects and only very rare complications.
6.1 Lindner et al. (11) in their series of six BPH patients with indwelling
catheters reported no complications. Their one treatment failure involved a
patient with an enlarged prostate who tolerated the treatments poorly due
to prostatic tenderness.
6.2 In their series of 29 BPH patients who were poor surgical risks and given
transrectal hyperthermia, Yerushalmi et al. (27) reported that the
treatments were well tolerated noted no complications or side effects.
Three patients eventually died from their underlying medical conditions
Hyperthermia for BPH
220
(which had contraindicated surgery), and one patient withdrew from the
study (reason not mentioned).
6.3 Yerushalmi et al. (29) reported on 32 patients treated using a transrectal
applicator for carcinoma of the prostate. 2/4 patients treated with
hyperthermia alone had mild diarrhea. 13/20 patients treated with
hyperthermia and external beam radiation therapy had diarrhea. Other side
effects which occurred when hyperthermia was combined with radiation
included: radiation cystitis (1/20), tenesmus (1/20), proctitis (4/20),
frequency (1/20), dysuria (1/20), and rectal bleeding (1/20). 8 patients
treated with hyperthermia and hormonal therapy had no side effects. They
concluded that hyperthermia did not cause more morbidity than that
expected from radiation therapy alone.
6.4 Servadio et al. (21) reported on a series of 32 patients with various
prostatic diseases given a total of 192 transrectal hyperthermia treatments.
They reported the occurrence of small prostatorectal fistulas in two
patients. One patient had a previous history of prostatic abscess and the
other had chronic proctitis. The authors felt that, in retrospect, these two
patients were unsuitable candidates for transrectal hyperthermia.
6.5 Sapozink et al. (17) have treated 21 patients with a total of 177
transurethral hyperthermia treatments. They report that there were no
complications in this series. Acute toxicity was found to occur frequently,
but was mild and easily managed by supportive measures. The reported
toxicities included bladder spasm (26% of treatments, 71% of patients),
hematuria (23% of treatments, 71% of patients), dysuria (9% of
treatments, 48% of patients), and urethral pain (8% of treatments, 43% of
patients). In one patient, urethral pain limited the power which could be
applied to the microwave antennas.
The specific acute toxicites to be evaluated in this study are listed in Appendix II, along
with the irritative symptoms listed in Appendix I.
7. MONTTORTNG OF PATTENTS
INITIAL WORKUP: Each patient will undergo complete physical exam with laboratory
tests as outlined in section 4.
TREATMENT COURSE: Prior to each hyperthermia treatment, subjective data will be
recorded relative to the patient's irritative and obstructive symptoms (Appendix I), and
toxicities (Appendix II). Laboratory data (i.e. urinalysis / culture) will be obtained if
Hyperthermia for BPH
221
warranted. Objective urological parameters may be obtained (i.e. maximum flow rate,
post-void residual) to confirm changes.
HYPERTHERMIA TREATMENTS: Temperature data are monitored and recorded
automatically by computer during treatments. If patient discomfort occurs, microwave
power will be reduced until the temperatures are tolerated by the patient (see section
5.1). The temperature data and narrative report on each hyperthermia treatment will be
maintained in a separate patient folder which is stored in the hyperthermia treatment
suite, and relevant clinical information will be recorded in the patient record.
POST-TREATMENT FOLLOW-UP: The follow-up schedule will be identical to that used
by Yerushalmi et al. (27):
1) 2 and 4 weeks following the last treatment, then at
2) 2, 3, and 4 months, and then
3) every 3 months.
After 12 months, the appropriate follow-up interval will be determined on an individual
basis by the patient and physician. Post-treatment evaluation will include digital
examination of the prostate, labs if necessary, and scoring of subjective symptoms
(irritative and obstructive). Objective data will also be obtained at each follow-up visit.
8. TRITRRIA RESPONSE
8.1
SUBJECTIVE DATA: Overall patient response will be based on the total
score from the symptom evaluation form (Appendix I). In addition, each
individual symptom will be evaluated to determine which are modified most
significantly by the hyperthermia treatments.
8.2
POST-VOID RESIDUAL: This parameter may be evaluated using either
ultrasound or direct catheterization. The actual volumes determined at each
follow-up visit will be statistically correlated to determine treatment effect
(section 9.2).
8.3
MAXIMUM VOLUMETRIC FLOW RATE: The urinary flow rate measured
at each follow-up visit will also be evaluated statistically to assess treatment
effect (section 9.2). Due to "training" effects, this parameter can be expected
to improve several percent over time.
8.4
DURATION OF RESPONSE: At the present time, the longest patient followup period is approximately 2 years. The time over which treatment response
can be expected to be maintained is not known, and the changes of the
quantified subjective and objective responses will clarify the duration of
response as patients are followed over time.
Hyperthermia for BPH
222
9. STATTSTTrAI. CONSTDKRATIONS
9.1
9.2
ENDPOINTS: The objective of this study is to evaluate hyperthermia as a
possible treatment modality for symptomatic benign prostatic hyperplasia.
Endpoints will be toxicity of treatment as well as subjective and objective
response. The specific parameters which will be used in this evaluation
are listed below (see Appendices I and II).
Toxicity: bladder spasm, hematuria, urethral pain, irritative symptoms,
retrograde ejaculation, and impotence.
Subjective svmptoms: hesitancy, intermittency, dribbling, force and size
of stream, sensation of urinary retention, frequency, nocturia, and
urgency.
Objective tests: post-void residual volume, maximum urinary flow rate,
prostatic size (rectal ultrasound).
PATIENT ACCRUAL: Surgery is currently the primary treatment for
symptomatic BPH (10). Initially, this study will address those patients for
whom surgery is desirable but contraindicated. In addition, some patients
may desire a trial of a non-surgical alternative before committing
themselves to surgeiy. It is anticipated that 10-20 patients will be accrued
annually on this protocol. The initial objective of this study is to determine
whether there is a change in the subjective and objective symptoms
resulting from the hyperthermia treatments (not a comparison between
hyperthermia and surgery). Both subjective and objective changes will be
evaluated using paired t-tests. For any endpoint, if it is assumed that the
average change (from pre-treatment to post-treatment) equals the standard
deviation of the change, then there is an 80% probability of detecting it at
the 0.05 level of significance with 10 subjects. For 90% power, 13
patients are needed. If the standard deviation is double the average, the
sample sizes required for 80% and 90% power are 34 and 44,
respectively. Repeated measurements analyses will also be performed as
9.3
follow-up data accumulates (6).
DATA COLLECTION; Subjective scoring of symptoms will be tabulated
at each follow-up visit without investigator or patient knowledge of
previous responses (see Appendix I). Data will be recorded on a protocol
flow sheet (Appendix HI).
Hyperthermia for BPH
223
9.4
9.5
Toxicity will be evaluated in terms of the treatment side effects listed in
Appendix n and the irritative symptoms listed in Appendix I.
Correlation between treatment parameters (i.e. temperatures, number of
treatments) and clinical response will be made.
to. OFF STiinv RUTTRINRA
10.1 Unacceptable progression of obstructive symptoms.
10.2 Toxicity which, in the opinion of the principle investigator or attending
physician, indicates that it is in the best interest of the patient to
discontinue therapy.
10.3 Patient refusal to continue.
Hyperthermia for BPH
224
H. KRFKRBNCRS
I.
Astrahan MA, Sapozink MD, Cohen D, Luxton G, Kampp TD, Boyd S, and
Petrovich Z. Microwave applicator for transurethral hyperthermia of benign prostatic
hyperplasia. Int. J. Hyperthermia 5(3): 283-296 (1989).
2.
Astrahan MA, Sapozink MD, Jozsef G, Cohen D, Luxton G, Boyd S, Kampp TO,
and Petrovich Z. Microwave applicators for transurethral hyperthermia of benign
prostatic hyperplasia. 9th Annual Meeting, North American Hyperthermia Group,
March 18-23,1989 (ABSTRACT).
3.
Braunwald E, Isselbacher KJ, Petersdorf RG, Wilson JD, Martin JB, and Fauci AS,
eds. Harrison's Principles of Internal Medicine. 11th ed., McGraw-Hill, New York,
p. 1582 (1987).
4.
Coughlin CT, Douple EB, Strohbehn JW, Eaton WL, Trembly BS, and Wong TZ.
Interstitial hyperthermia in combination with brachytherapy. Radiology 148: 285288 (1983).
5.
Coughlin CT, Wong TZ, Strohbehn JW, Colacchio TA, Sutton JE, Belch RZ, and
Douple EB. Intra-operative interstitial microwave-induced hyperthermia and
brachytherapy. Int. J. Radiation Oncology Biol. Phys. 11: 1673-1678 (1985).
6.
Dain B. Biostatistics and epidemiology group, Dartmouth Medical School, Hanover,
NH, personal communication, Aug. 24,1989.
7.
Dunlop PRC, Dickinson RJ, Hand JW, and Field SB. The use of thermal dose in the
clinicd application of localised hyperthermia. In Hvperthermic Oncologv 1984. vol.
1, Overgaard J, ed., Taylor and Francis, London, pp.187-190 (1984).
8.
King RWP, Trembly BS, and Strohbehn JW. The electromagnetic field of an
insulated antenna in a conducting or dielectric medium. IEEE Trans. Microwave
Theory Techniques MTT-31: 574-583 (1983).
9.
Leib Z, Rothem A, Lev A, Servadio C. Histopathological observations in the canine
prostate treated by local microwave hyperthermia. Prostate 8(1): 93-102 (1986).
10. Lepor H. Nonoperative management of benign prostatic hyperplasia. J. Urology
141: 1283-1289 (1989).
11. Lindner A, Golomb J, Siegel Y, and Lev A. Local hyperthermia of the prostate gland
for the treatment of benign prostatic hypertrophy and urinary retention- a preliminary
report. Br. J. Urology 60: 567-571 (1987).
12. Magin RL, Fridd CW, Bonfiglio TA, and Linke CA. Thermal destruction of the
canine prostate by high intensity microwaves. J. Surgical Res. 29: 265-275 (1980).
13. Mendecki J, Friedenthal E, Botstein C, Paglione R, and Sterzer F. Microwave
applicators for localized hyperthermia treatment of cancer of the prostate. Int. J.
Radiation Oncology Biol. Phys. 6(11): 1583-1588 (1980).
Hyperthermia for BPH
225
14. Petrowicz O, Scheiblich J, Crucius A, Erhardt W, Wriedt-Liibbe I, Ultsch B, and
Bliimel G. High-frequency transmitter for the localized heat treatment of the prostate
gland. Natl. Cancer Inst. Monogr. 61: 473-476 (1982).
15. Roberts DW, Coughlin CT, Wong TZ, Fratkin JD, Douple EB, and Strohbehn JW.
Interstitial hyperthermia and iridium brachytherapy in the treatment of malignant
glioma: a phase I clinical trial. J. Neurosurgery 64:581-587 (1986).
16. Roti Roti JL and Laszlo A. The effects of hyperthermia on cellular macromolecules.
In Hyperthermia and Oncology, yol. 1, Urano M and Douple E, eds. VSP BV,
Utrecht, pp. 13-56 (1988).
17. Sapozink M, Boyd S, Astrahan M, Jozsef G, Nourparyar S, and Petroyich Z.
Transurethral hyperthermia for benign prostatic hyperplasia - preliminary results. 9th
Annual Meeting, North American Hyperthermia Group, March 18-23, 1989
(ABSTRACT).
18. Scheiblich J and Petrowicz O. Radiofrequency-induced hyperthermia in the prostate.
J. Microw. Power 17(3): 203-209 (1982).
19. Seryadio C, and Leib Z. Hyperthermia in the treatment of prostate cancer. Prostate
5(2): 205-211 (1984).
20. Seryadio C, Leib Z, and Lev A. Diseases of prostate treated by local microwave
hyperthermia. Urology 30(2): 97-99 (1987).
21. Servadio C, Leib Z, and Lev A. Further observations on the use of local
hyperthermia for the treatment of diseases of the prostate in man. Eur. Urol. 12: 3840 (1986).
22. Stewart JR. Prospects for hyperthermia in cancer therapy. In Hyperthermia and
Oncology, vol. 1, Urano M and Douple E, eds. VSP BV, Utrecht, pp. 1-12 (1988).
23. Szmigielski S, Zielinski H, Stawarz B, Gil J, Sobczynski J, Sokolska G,
Jeljaszewicz J, and Pulverer G. Local microwave hyperthermia in treatment of
advanced prostatic adenocarcinoma. Urol. Res. 16(1): 1-7 (1988).
24. Trembly BS. The electric field of an insulated linear antenna embedded in an
electrically-dense medium. PhD Thesis, Dartmouth College, Hanover, NH. (1982).
25. Wong TZ and Trembly BS. Performance of an intracavitary microwave antenna with
choke. 9th Annual Meeting, North American Hyperthermia Group, March 18-23,
1989 (ABSTRACT).
26. Yerushalmi A. Hyperthermia treatment of prostate tumors. In Hyperthermia in Cancer
Treatment, vol. Ill, Anghileri LJ and Robert J, eds. CRC Press, Boca Raton, pp. 89101 (1986).
27. Yerushalmi A, Fishelovitz Y, Singer D, Reiner I, Arielly J, Abramovici Y,
Catsenelson R, Levy E, and Shani A. Localized deep microwave hyperthermia in the
treatment of poor operative risk patients with benign prostatic hyperplasia J. Urology
133: 873-876 (1985).
Hyperthermia for BPH
226
28. Yerushalmi A, Servadio C, Leib Z, Fishelovitz Y, Rokowsky E, and Stein JA, Local
hyperthermia for treatment of carcinoma of the prostate: a preliminary report. Prostate
3(6): 623-30 (1982).
29. Yerushalmi A, Shani A., Fishelovitz Y, Arielly J, Singer D, Levy E, Katsnelson R,
Rakowsky E, and Stein JA. Local microwave hyperthermia in the treatment of
carcinoma of the prostate. Oncology 43: 299-305 (1986),
30. Yerushalmi A, Shpirer Z, Hod I, Gottesfeld F, and Bass DD. Normal tissue response
to localized deep microwave hyperthermia in the rabbit's prostate: a preclinical study.
Int. J. Radiation Oncology Biol. Phys. 9(1): 77-82 (1983).
Hyperthermia for BPH
227
APPENDIX T -
SlIBTECTTVF. SYMPTOM SmUTNO
ObstrHttiyg Symptoms
SYMPTOM
SEVERITY
SCORE
DESCRIPTION
Hesitancy
0
1
2
Occasional (occurs in ^0% of attempts to void)
Moderate (occurs in 20-50% of attempts to void)
Frequent (occurs in >50% of attempts, but not dways,
and may last up to 1 minute)
Always present and lasts for 1 minute or longer
Intermittency
0
1
2
Occasional (occurs in ^0% of attempts to void)
Moderate (occurs in 20-50% of attempts to void)
Frequent (occurs in >50% of attempts, but not dways,
and may last up to 1 minute)
Always present and lasts for 1 minute or longer
3
Terminal
Dribbling
0
1
2
3
Occasional (occurs in ^0% of attempts to void)
Moderate (occurs in 20-50% of attempts to void)
Frequent (occurs in >50% of attempts, but not Always)
Always present and lasts for 1 minute or more, or
wets clothing
Size / Force of
Urinary Stream
0
1
2
3
Absence of symptom
Impaired trajectoiy
Size and force are restricted most of the time
Pt. urinates with great effort and stream is interrupted
Sensation of
Incomplete
Bladder
Emptying
0
1
2
3
Absence of symptom
Occasional sensation after voiding (£50% of the time)
Frequent sensation after voiding (>50% of the time)
Constant and urgent sensation and no relief upon voiding
Hyperthermia for BPH
228
APPENmx T. STlRTKrTTVF. SYMPTOM SrORTNO (CONT'D^
Irritative Svmntoms
SEVERITY
SYMPTOM
SCORE
DESCR]
Nocturia
0
1
2
3
Absence of symptom
Ft. awakens Ix each night to urinate
Ft. awakens 2-3x each night to urinate
Ft. awakens >4x each night to urinate
Daytime
Frequency
0
1
2
3
Pt. urinates l-4x daily
Pt. urinates 5-7x daily
Pt. urinates 8-12x daily
Pt. urinates >13x daily
Urgency
0
1
2
Absence of symptoms
Occasionally difficult for pt. to postpone urination
Frequently difficult (>50% of time) to postpone urination,
may rarely lose urine
Always difficult to postpone urination and pt. sometimes
loses urine
3
Dysuria
0
1
Absence of symptom
Occasional burning sensation during urination
Frequent (>50% of time) burning sensation during
unnation
Frequent and painful burning sensation during urination
Hyperthermia for BPH
229
APPKNOTX
TT - TOXTriTY SYMPTOM RVAIJIATTON
SYMPTOM
SEVERTTY
SCORE
DESCRIPTION
Bladder
Spasm
0
1
2
3
Absence of symptom
Occasional, minor
Frequent, interfering with daily activity
Frequent and severe, a major problem for the patient
Hematuria
0
1
2
3
None
Microscopic only
Grossly visible
Severe, clot retention
Urethral Pain
0
1
2
Absence of symptom
Mild pain during or between treatments
Moderate pain during or between treatments, requiring
occasional reduction in power
3
Severe pain during or between treatments, requiring
hyperthermia treatments to be curtailed or cancelled
Impotence
0
1
No
Yes
Retrograde
Ejaculation
0
1
No
Yes
during treatment
(Rev 9/28/89)
230
DARTMOUTH-HITCHCOCK MEDICAL CENTER
Consent for rnvestiyational Theranv
Transurethral Hyperthermia for the Treatment of
Benign Prostatic Hyperplasia
INTRODUCTION
You die, being asked to participate in a research treatment program. Your participation is
voluntary, and this treatment, though directed toward cure or improvement, may result in
neither. There are alternatives to this treatment as indicated in the alternatives section. You
are not obligated to participate in this research program, or any other program, in order to
seek medical care for your disease, and you niay withdraw at any time without
compromising your care.
Before agreeing to participate in this study, please read the following information very
carefully and ask you physicians and nurses any questions so that you will understand
what your participation in this study will involve.
SUMMARY OF RFSFARCH PROTFCT
Your physician has informed you that you have urinary bladder outlet obstruction
secondary to a benign enlargement of your prostate. This condition is usually curable by
the operation of prostatectomy, which can be performed by open surgery or transurethral
resection. Drug therapy is dso an alternative.
Hyperthermia is a treatment in which tissues are heated to temperatures of 42°C to 55°C
(108°-131°F). You are being asked to participate in an investigational treatment program
testing a new hyperthermia applicator which is designed to heat the part of the prostate
which surrounds the urethra. It is hoped that the hyperthermia will help shrink that part of
the prostate which is contributing to the obstruction. Similar devices have been studied at
other institutions; in these preliminary studies, 82% of men showed improvement in their
symptoms and 81% showed measurable improvement in urinary flow. However, followup has generally not exceeded one year, and therefore, it is not certain whether this
improvement is permanent. The purpose of our study is to determine the short- and longterm effectiveness and the side effects of this treatment
TRRATMENT PROGRAM
Before you are asked to participate in this program, you already will have undergone a
routine urologic evaluation to confirm the presence of bladder outlet obstruction by
prostatic enlargement. This evaluation will have included measurement of the urinary flow
rate, residual bladder urine measured after voiding, visual examination of the urethra and
bladder by a cystoscope (operating telescope), and routine studies of the blood and kidney
function.
Treatment will consist of six to twelve one-hour hyperthermia sessions, with a period of
several days between treatments. The exact number of treatments will be determined by
your doctor. The treatments will take place in the hyperthermia suite at the Department of
Hyperthermia for BPH - CONSENT FORM
231
Radiation Therapy. Each treatment session will require approximately two hours to allow
for set-up and treatment Prior to each treatment, a local anesthetic jelly will be placed into
the urethra. The microwave hyperthermia applicator will then be inserted through the
urethra into the bladder in the same manner Aat a standard Foley catheter is inserted. Xrays may be taken to verify the location of the microwave antenna elements within the
prostatic urethra. Microwave power will then be applied to the applicator for the one-hour
treatment. Temperatures will be constantly monitored during the treatment using a
computer-controlled system. The hyperthermia applicator will be removed at the end of the
treatment session.
After all of the treatments have been completed, you will return periodically for assessment
of you symptoms, a physical examination, blood tests, urine analysis, a urinary flow rate,
and post-void residual volume. These studies will be repeated at 2 weeks, 4 weeks, 2
months, 3 months, 4 months, and then every 3 months for one year. After one year, your
doctor will determine when it is appropriate for you to return.
RISKS
Known side effects from any catheterization of the urethra and bladder include pain,
bleeding, and/or burning sensation of the urethra within six hours of the procedure.
Subsequently, infection, and/or stricture may occur at any time after treatment. Bladder
spasm may occur during or between treatments, and may cause you to lose urine
occasionally. Urination may occasionally be painful between treatments. These side effects
usually disappear after the treatment course has been completed, and/or can be treated with
medication. Unanticipated side effects that have not yet been reported may occur. These
complications may result in minor inconvenience or may be so severe as to result in loss of
life. Retrograde ejaculation (semen falling back into the bladder at orgasm), impotence, and
urinary incontinence are possible side effects. These side effects must be viewed in the
context of the alternative Aerapy of transurethral prostatic resection (TURF), which results
in almost certain retrograde ejaculation, impotence in approximately 10% of men, and, with
significant frequency, the complications of stricture, bladder neck contracture, and
incontinence.
B ENEFITS
It is hoped that this procedure will benefit you by improving the drainage of your urinary
bladder, thus reducing your symptoms and the chance of subsequent urinary infection or
kidney damage. The risks of general anesthesia and increased morbidity of prostatectomy
may be avoided. However, should the hyperthermia treatments fail to produce the
anticipated results, prostatectomy may remain a viable option.
It is further hoped that the information obtained during the course of this study will be of
benefit to others.
A LTERNATIVES
For symptomatic bladder obstruction secondary to BPH, the alternatives include 1)
"watchful waiting," whereby some patients might improve spontaneously; 2) prostatectomy
by either open surgery or TURP; 3) experimental drug therapy with medication that relaxes
the prostatic muscle, but may have side effects; and 4) balloon dilatation, another
experimental therapy in which the urethra is expanded mechanically. These alternatives,
Hyperthermia for BPH - CONSENT FORM
232
including no treatment, have their own risks and benefits, which should be discussed with
your doctor.
FTNANCTAL C ONSTDFRATIONS
This research study is not sponsored by a corporation or an agency of the Federal
Government. There will be no extra cost to you for participation in this study. You will not
be charged for the extra tests or treatments described in this study, but routine tests will be
charged to either you or your insurance carrier. The hyperthermia treatments will be
provided at no cost to you.
rONFTnFNTTALTTY
The results of your treatment will be compared with the results of other patients who have
been treated with the same procedures. Statistical data will be collected, analyzed, and
exchanged by the appropriate medical investigational groups, but all identifying information
will be kept confidential.
FIIRTHFR INFORMATION
This research is being conducted by the Sections of Urology and Radiation Therapy of the
Dartmouth-Hitchcock Medical Center. You have had an opportunity to have your questions
answered about this research. Any questions you have in the future may be brought to the
attention of John A. Heaney, M.D., Section of Urology, D^mouth-Hitchcock Medical
Center, Hanover, NH; (603) 646-5080. If Dr. Heaney is not available, other members of
the Sections of Urology or Radiation Therapy will be available to answer you questions at
any time.
i
DISCLAIMER S TATEMENT
The Dartmouth-Hitchcock Medical Center (DHMC) and its component institutions wish to
make you aware of its position on compensation for an injury as a result of a research
procedure since possible side effects can never be totally excluded. The policy is stated
below.
Participating in medical research may result in an injury. If you have such an injuty,
DHMC will help you obtain medical treatment, but will not provide you with financial
compensation or reimbursement for this care. By signing this consent form, you are not
giving up any legal rights. If the research is conducted in a negligent manner and causes
you some harm, you may be able to recover the costs of care and damages from DHMC. If
you have any questions about the legal responsibility of DHMC, please call the Mary
Hitchcock Memorial Hospital Office of Risk Management between the hours of 8:00 A.M.
and 5:00 P.M. on Monday through Friday at (603) 646-7864. At all other times call (603)
646-5000 and ask for the Hospital Administrator On-call.
233
Hyperthermia for BPH - CONSENT FORM
CONSENT
I have read and understand the above information about hyperthermia of the prostatic
urethra and agree to participate in this study. I have been given a copy of this informed
consent for my own records.
Patient's Signature
Date
Witness' Signature
Date
Документ
Категория
Без категории
Просмотров
0
Размер файла
4 666 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа