close

Вход

Забыли?

вход по аккаунту

?

Transperineal interstitial microwave therapy for recurrent localized carcinoma of the prostate following brachytherapy seed implantation: Phantom studies

код для вставкиСкачать
INFORMATION TO U SERS
This manuscript has been reproduced from the microfilm m a ste r UMI films
the text directly from the original or copy submitted. Thus, som e thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of th is reproduction is dependent upon th e quality of the
copy subm itted. 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 com er and continuing
from left to right in equal sections with small overlaps.
ProQ uest Information and Learning
300 North Z eeb Road, Ann Arbor, Ml 48106-1346 USA
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Transperineal Interstitial M icrowave Therapy for Recurrent Localized
Carcinoma of the Prostate Following Brachytherapy Seed
Implantation: Phantom Studies
By
Claire McCann
A thesis submitted in conformity with the requirements
for the degree of Master of Health Science
Graduate Department of the Institute of Biomaterials and Biomedical Engineering
Clinical Engineering Program
The University of Toronto
© Copyright Claire McCann 2002
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1*1
National Library
of Canada
Bfcliotteque rationale
du Canada
Acquisitions and
Bibliographic Services
Acquisitions et
sendees bibiiographiques
3S5WaNnglon Stra*
Ottawa ON K1A0N4
385, ruaWMington
Ottawa ON K1A0N4
Canada
Your I
Ourm
The author has granted a non­
exclusive licence allowing the
National Library o f Canada to
reproduce, loan, distribute or sell
copies of this thesis in microform,
paper or electronic formats.
L’auteur a accordd une licence non
exclusive permettant a la
Bibliotheque nadonale du Canada de
reproduire, prdter, distribuer ou
vendre des copies de cette these sous
la forme de microfiche/film, de
reproduction sur papier ou sur format
dlectronique.
The author retains ownership o f die
copyright in this thesis. Neither the
thesis nor substantial extracts from it
may be printed or otherwise
reproduced without the author’s
permission.
L’auteur conserve la propridte du
droit d’auteur qui protege cette these.
Ni la these m des extraits substantiels
de celle-ci ne doivent etre imprimes
ou autrement reproduits sans son
autorisadon.
0-612-73799-3
CanadS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Transperineal Interstitial Microwave Therapy (TIMT) for Recurrent
Localized Carcinoma o f the Prostate Following Brachytherapy Seed
Implantation: Phantom Studies
Claire McCann B.A.Sc., M.H.Sc.
Institute of Biomaterials and Biomedical Engineering
Clinical Engineering Program
The University of Toronto, 2002
We are investigating the feasibility of TIMT as a salvage treatment for prostate cancer
following failed Brachytherapy. Given the electrically and thermally conductive nature
of brachytherapy seeds, concern existed as to their potential to scatter the microwave
energy and increase the extent of the temperature distribution respectively, leading to
areas of excessive or inadequate heating, which could result in damage to surrounding
critical structures or inadequate dose to areas of the tumour.
To characterize these
effects, a helical antenna was inserted into phantom material with and without seeds,
where specific absorption rate (SAR) and temperature distributions were measured using
Infrared (IR) thermography and thermometry. Comparisons of the no-seed and seed
cases revealed statistically insignificant differences in the SAR and temperature
distribution measurements. These results suggest that TIMT as a salvage treatment for
failed Brachytherapy is possible, as the seeds do not compromise the localization of
heating to the target volume.
ii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgments
I would like to thank my supervisors, Dr. Michael D. Sherar and Alf Dolan whose
intellectual support and guidance were integral to the completion of this project. I would
also like to thank Dr. J. Carl Kumaradas who provided direction and unlimited support in
finite element modeling.
I would like to thank the other researchers and students in the
hyperthermia/thermal therapy group at Princess Margaret Hospital including Lee Chin,
Vanessa Choy, Sean. R.H. Davidson, Dr. M. Gertner, Dr. M Kolios, Mihaela Pop, Arthur
Worthington and Xia Wu for their support, both technical and intellectual, as well as
encouragement in all aspects of development of this project. I am greatly appreciative of
all of your help.
A special thanks to my family, especially my parents for their unlimited
emotional support throughout these many years of school. A very special thanks to Bill
for all of his encouragement, emotional support and his laboratory expertise. You were
such an integral part of the completion of this project.
Finally, I would like to thank the Ontario Cancer Institute for providing the
facilities and the environment that make research such a wonderful experience. I would
also like to thank NSERC for their funding of this project.
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
Abstract.........................................................................................................................ii
Acknowledgments........................................................................................................ iii
Table of Contents......................................................................................................... iv
List of Figures............................................................................................................... v
List of Tables................................................................................................................vi
List of Abbreviations...................................................................................................vii
Chapter 1 Introduction
1
1.1
Motivation.........................................................................................................1
1.2
Prostate Cancer Incidence and Description....................................................1
1.3
Brachytherapy.................................................................................................. 3
1.3.1
Brachytherapy as a Primary Treatment Option for Localized
Prostate Cancer...................................................................................3
1.3.2
Brachytherapy Seed Arrangements...................................................5
1.3.3
Incidence of Prostate Cancer Recurrence Following First-Line
Brachytherapy....................................................................................7
1.3.4 Treatment Options for Prostate Cancer Recurrence Following
Brachytherapy....................................................................................8
1.4
1.5
Thermal Therapy.............................................................................................9
1.4.1
Clinical Implementation of Microwave Thermal T herapy............9
1.4.2
Biological Rationale.........................................................................10
1.4.3
Thermal Therapy Modalities........................................................... 11
Microwave Thermal Therapy........................................................................15
1.5.1
Electrical Tissue Response to Microwave Radiation.....................15
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1.5.2
Microwave Applicators.................................................................. 15
1.5.3
Specific Absorption Rate (SAR)....................................................17
1.5.4
Infrared Thermography.................................................................. 19
1.5.5
Magnetic Resonance (MR) Thermometry..................................... 20
1.5.6
Discrete Thermometry................................................................... 21
1.6
Thermal Conduction..................................................................................... 22
1.7
Theoretical Modeling.................................................................................... 25
1.8
1.7.1
Overview: Methods of Modeling................................................... 25
1.7.2
Numerical Modeling of SAR..........................................................26
1.7.3
Numerical Modeling of Temperature Distribution....................... 27
Summary and Objectives...............................................................................29
Chapter 2 Methodology
31
2.1
Introduction.................................................................................................... 31
2.2
Overview: Material & Methods, Calibration and OperationProcedures..32
2.3
2.2.1
Thermal Imaging System................................................................ 33
2.2.2
Helical Antenna...............................................................................39
2.2.3
Microwave Generator..................................................................... 40
2.2.4
Biomedical Luxtron Temperature Probes..................................... 41
2.2.5
Brachytherapy Seeds and Experimental Seed Configurations
2.2.6
Phantom Materials.......................................................................... 45
2.2.7
Theoretical Modeling: Finite Element Simulations...................... 48
42
SAR and Extended Heating Experiments and Simulations: A Detailed
Description....................................................................................................50
2.3.1
Infrared Imaging Experiments........................................................ 50
2.3.2
Luxtron Temperature Measurements.............................................52
2.3.3
Electromagnetic Finite Element Simulations................................ 52
2.3.4
Thermal Finite Element Simulations.....................
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
Chapter 3 Results
3.1
57
Introduction: SAR Experiments and Plane Wave Simulations.................. 57
3.1.1
Two-Dimensional SAR Measurements..........................................58
3.1.2
Three-Dimensional SAR Measurements....................................... 60
3.2
Finite Element Plane Wave Simulations...................................................... 70
3.3
Extended Heating Experiments..................................................................... 73
3.4
3.3.1
Extended Heating Temperature Distributions: Antenna Plane
3.3.2
Three-Dimensional Extended Heating Measurements................. 79
3.3.3
Extended Heating Luxtron Temperature Measurements..............82
Finite Element Extended Heating Simulations............................................ 86
Chapter 4 Discussion
4.1
74
90
Introduction: SAR Experiments and Plane Wave Simulations..................90
4.1.1
SAR Experiments............................................................................ 91
4.1.2
Plane Wave Simulations.................................................................. 92
4.1.3
Correlation: Experiments and Simulations.....................................94
4.2
Future Work: Modeling the Helical Antenna...............................................95
4.3
Extended Heating..........................................................................................100
4.3.1
Experimental Temperature Measurements................................... 100
4.3.2
Multiple Seed Lines Case...............................................................102
4.3.3
Thermal Simulations.......................................................................103
4.3.4
Clinical Implications.......................................................................104
References
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
List of Figures
1.1
Artistic Rendering of Prostate Gland......................................................................... 2
1.2
Radiograph: Coronal Perspective of Brachytherapy Seed Lines in the Prostate.. .4
1.3
Brachytherapy Seed Line Configuration....................................................................6
1.4
Radiofrequency Ablation of a Tumour Volume....................................................... 12
1.5
Ultrasound Generated Thermal Lesion..................................................................... 14
2.1
Thermograph: Copper Pin Array for Pixel Size Determination............................. 34
2.2
Schematic: Theoretical Determination of Pixel Size............................................... 35
2.3
IR Thermovision Camera and Styrofoam Block for Phantom............................... 37
2.4
Digital Picture: Helical Antenna............................................................................... 39
2.5
Solid Model: Helical Antenna.................................................................................. 40
2.6
Model 3100 Fiberoptic Thermometers..................................................................... 41
2.7
Schematic: Brachytherapy Seed............................................................................... 43
2.8
Brachytherapy Seeds....................................................................
2.9
Brachytherapy Seed Configurations....................................................................... .44
43
2.10 Schematic: Axial View of Phantom, Antenna and Seeds....................................... 45
2.11
2.12
Mold configuration for phantom layer manufacturing........................................... 48
Flow Chart: Finite Element Approach to Electromagnetic and Thermal
Simulations............................................................................................................... 49
2.13 Flow Chart: Finite Element Thermal Simulation Steps.......................................... 56
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3.1
Schematic: Contour Characteristics..........................................................................58
3.2
2-D SAR Contour Pattern Antenna Plane: No-Seed Vs. Seed Case-A..................59
3.3
2-D SAR Contour Characterization Bar Graphs: Antenna Plane........................... 60
3.4
3-D SAR Contour Pattern Antenna Plane: No-Seed Vs. all Seed cases................ 61
3.5
3-D SAR Contour Characterization Bar Graphs: Antenna Plane...........................62
3.6
3-D SAR Contour Pattern 2.5mm above Antenna Plane........................................ 64
3.7
3-D SAR Contour Characterization Bar Graphs: 2.5mm above Antenna Plane...65
3.8
3-D SAR Contour Pattern 5mm above Antenna Plane............................................ 67
3.9
3-D SAR Contour Characterization Bar Graphs: 5mm above Antenna Plane...... 68
3.10
3-D Reconstruction of Experimentally Determined SAR for the No-Seed...........69
3.11
Single Seed Plane Wave Simulation-Worst Case.................................................... 71
3.12
Two Seed Lines-Plane Wave Simulation................................................................ .72
3.13
Schematic: Orientation and Location of Y and X-axis Temperature Profiles.......74
3.14
Extended Heating IR Temperature Distribution Plot: Antenna Plane................... 75
3.15
Normalized X and Y-axis Temperature Profiles: Experiments..............................76
3.16
X and Y-axis Profile Width Characterization Bar Graph........................................78
3.17
Normalized 3-D Reconstruction of Experimentally Determined Extended Heating
Temperature Distribution: No-Seed........................................................................ 80
3.18
Normalized 3-D Reconstruction of Experimentally Determined Extended Heating
Temperature Distribution: Seed Case-A
3.19
.....................................................81
Thermometry Experiments: Average Temperature Rise at Antenna HotSpot:
Control Vs. Seed Case Scenarios.............................................................................83
3.20 Thermometry Experiments: Average Temperature Rise 1cm Away from and
Parallel to Antenna Hot Spot: Control Vs. Seed Case Scenarios..........................84
3.21
Thermometry Experiments: Average Temperature Rise 0.5cm Away from and
Parallel to Antenna Hot Spot: Control Vs. Seed Case-A Scenario........................ 85
3.22 2-D Coronal Images: Theoretical Extended Heating Simulations..........................87
3.23 2-D Axial Images: Theoretical Extended Heating Simulations..............................88
3.24 Normalized X and Y-Axis Temperature Profiles: Simulations.............................. 89
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4.1
Plane Wave Simulations: Wave Propagation Perpendicular and Parallel to Seed
Axis............................................................................................................................ 93
4.2
Rhinoceros Model: Full Helical Antenna................................................................. 96
4.3
Top of Helical Antenna: Rhinoceros Solid Model and Mesh................................. 97
4.4
Helical Antenna Junction: Rhinoceros Model and ICEM Mesh............................ 98
4.5
End of Helix: Rhinoceros Solid Model and ICEM Mesh.......................................99
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Tables
1.1
Thermal and Material Properties of Titanium and PAG Phantom.........................24
2.1
Electrical and Thermal Properties of Phantom-Equivalent Muscle Tissue at 2S°C
and Muscle Tissue at 37°C at a frequency of 915MHz......................................... .45
3.1
No-Seed Vs. Seed Case-A Temperature Rise Values: Simulations.......................87
3.2
Y-Profile Width Measurements: Simulations...........................................................89
3.3
X-Profile Width Measurements: Simulations...........................................................89
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Abbreviations
1. IR: Infrared
2. PAG: Polyacrylamide gel
3. SAR: Specific Absorption Rate
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 1
INTRODUCTION
1.1
Motivation
To determine the viability of Transperineal Interstitial Microwave Therapy (TIMT) for
the treatment of recurrent prostate cancer, following failed Brachytherapy.
1.2
Prostate Cancer Incidence and Description
Prostate cancer is the second most common cause of cancer related death in men (second
to lung cancer) in Canada and the United States but is the most frequently diagnosed type
of cancer with 185 cases diagnosed per 100,000 men in 1997. (NCIC: Canadian Cancer
Statistics, 1998). While these statistics are alarming, the situation is expected to worsen.
Using a bivariate multiplicative model of prostate cancer incidence and mortality,
Morrison and MacNeill predict that the number of prostate cancer incidences will rise
from 11,355 in 1990, to 26,900 by the year 2010 and to 35,200 by the year 2016. As a
result, the number of deaths is forecasted to climb from 3424 observed in 1991, to an
estimated 6300 by the year 2010. (Can J Public Health, 1995)
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
The large numbers of men with prostate cancer today and the increasing number of cases
predicted in the near future suggests in turn increased numbers of primary treatment
failure.
Adenocarcinoma of the prostate is the clinical term for cancer that begins as a
tumour in the prostate gland. As prostate cancer grows, it may spread to the interior of
the gland, to tissues near the prostate, to seminal vesicles and to distant parts of the body
(bones, liver, lungs). This paper will focus on prostate cancer that is confined to the
gland, that is, localized prostate cancer. The prostate gland is located in the pelvis,
posterior to the bladder, superior to the urethral sphincter and the penis and anterior to the
rectum (see figure 1.1). The prostate is comprised of glandular tissue and muscle fibers
that surround the urethra. The gland is covered by a capsular membrane that produces
prostate-specific antigen (PSA), a glycoprotein involved in the liquefaction of the seminal
gel. As the prostate gland increases in size as a result of prostate abnormalities including
prostate cancer, PSA levels usually increase. Therefore, PSA measurements are used in
conjunction with the digital rectal examination, as indicators of prostate cancer. PSA
levels are also used as a marker for determining treatment failure. According to the
American Society for Therapeutic Radiation Oncology (ASTRO), prostate cancer
recurrence is defined as three consecutive rises in PSA levels following treatment.
Figure 1.1 Artistic Rendering of Prostate Gland and surrounding tissue structures. (Oncology
Nursing, 1991)
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
13
Brachytherapy
1.3.1
Brachytherapy as a Primary Treatment Option for Localized Prostate Cancer
Brachytherapy is one of a number of primary treatment options available to those
patients with localized prostate cancer. Other options include no treatment, hormone
therapy, prostatectomy, and external beam radiation therapy (EBRT). The selection of
treatment is based on a number of factors including the health of the patient as well as
their personal preferences. Brachytherapy has increased as the primary treatment option
of choice because of its ability to strategically deliver highly localized low energy
radiation to the intended diseased areas of the prostate. (Brawer, 2001) In 1996, of all
patients with nonpalpable prostate cancer, 5.8% of this population was treated with
permanent Brachytherapy. This number is expected to increase to 19% once the statistics
for the year 2000 are compiled (Mettlin et al., 1999).
In Brachytherapy, radiation sources are in direct contact with the tumour, so as to
produce precisely targeted irradiation at short distances. Radiation treatment uses ionizing
radiation to destroy the cancerous cells via direct and indirect mechanisms.
Ionizing
events which occur on the DNA can result in strand breaks, while ionizing events
occurring near the DNA can result in the production of free radical species, which attack
the DNA following reactions with molecules in solution at near diffusion controlled rates.
Whether direct or indirect mechanisms result, there is no guarantee that all cancerous cells
will be destroyed, as the DNA may not be in the vicinity of a radiation induced event.
Permanent interstitial Brachytherapy involves the insertion of strategically located
radioactive seeds into the prostate gland. The success of the treatment depends in great
part on the location and density of the seeds within the prostate. Transrectal ultrasound is
used to determine the size and shape of the prostate such that the seeds can be accurately
located consistent with the geometry of the target volume. Computer programs are used to
calculate the number, activity and distribution of the seeds within the gland needed to
cover the target volume. The radioactive seeds are implanted transperineally with needles
3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
using transrectal ultrasound guidance. The radioactive element contained in these seeds
can be either Iodine-125 (I25I) or Palladium-103 ( l03Pd), both types being encapsulated by
an outer titanium shell. These isotopes have a different half-life, which affects the initial
dose rate of the implant. 1-125 has a half-life of 60 days and emits radiation at a rate of 810 cGy per hour. Pd-103 has a half-life of 17 days and emits radiation at a rate of 20-24
cGy per hour, both at the time of the actual implant. (Radge, 2001) After several half
lives, the radiation decays to an inert state, leaving only the outer metal shell.
Brachytherapy seeds emit low energy radiation (24 KeV), therefore many seeds are needed
to cover the entire tumour.
Radto«c<w todtne «»id« in th« p«Mwt on mMm tofcn lo vfUy t t f pm Won o* Mw
l i i ^ a g ild
H M la lifiM C iff
Figure 1.2. Radiograph: Coronal Perspective of Brachytherapy Seed Lines in the Prostate. The
radiograph depicts multiple seed lines spaced throughout the prostate gland. The seed lines are uniformly
spaced, as are the seeds within the lines. The template that guides the insertion o f the seed-loaded needles is
also depicted on the radiograph as a uniformly spaced grid.
(http://www.kcc.tju.edu/RadOnc/brachy/imp.htm)
4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
1.3.2
Brachytherapy Seed Arrangements
Transperineal ultrasound guided prostate Brachytherapy may require anywhere
from 40-120 radioactive seeds implanted into the prostate depending on the size and shape
of the gland and the extent of the disease.
In order to ensure the most effective
Brachytherapy treatment, the seed loading approach used is of critical importance, as it
dictates the final seed arrangement and therefore the radiation dose distribution within the
prostate. In an effort to achieve the most effective arrangement of the implanted sources,
templates are used to guide the placement of the seed-loaded needles into the tumour
volume. This allows for control over the entire prostate target volume and specification of
source placement at any point within the gland. (Beyer, 2001) Generally speaking, seeds
are located on coronal planes throughout the gland in the form of seed lines with each line
spaced about 1cm apart. Within each seed line, there are usually 5 seeds per line and the
distance between the end o f one seed and the start of the next is about 5-6mm (see figure
1.3 (a)). In the transverse plane, the seed lines are also spaced about 1cm apart (see figure
1.3 (b)). Such evenly spaced and symmetrical seed line arrangements are achieved with
what is generally termed a uniform seed loading approach. The anatomy of the prostate
and its proximity to surrounding ‘healthy’ tissue structures including the bladder, rectum
and urethra, require that some adjustments to this approach be made, as it can produce an
intolerably high radiation dose to unwanted tissue areas, particularly along the urethra.
Therefore, a modified seed loading approach (Butler, 1999) is used, as it deposits 2/3 of
the seeds on lattice points of a 1cm cubic grid with two planes of the lattice defined by the
base plane of the prostate and the posterior (lowest) implant row. This approach produces
a uniform dose distribution, while still maintaining dose levels less than 150% of the
minimum peripheral dose (MPD) in the regions surrounding the urethra.
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
The modified uniform loading approach results in seed arrangements that are fairly
uniform throughout the prostate gland. Therefore, for the purposes of this study, the seed
arrangement resulting from this approach was selected for testing because of its use in a
clinical setting as well as its ease of reproducibility. Specifically, seed lines consisting of
S seeds, each seed spaced 5mm apart, were located on coronal planes, with each seed
plane separated by 1 cm.
PTV
Base o f
Prostate
1 cm
1 cm
5 mm
^
Brachytherapy
Seed
Apex o f
Prostate
Figure 1.3 (a) Coronal View of Brachytherapy Seed Line Configuration. Consistent with the modified
uniform loading approach, the seeds were generally configured in seed lines spaced 1cm apart throughout the
Planning Treatment Volume (PTV).
6
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
Top o f Prostate
(Bladder End)
1 cm
1 cm
Cross-Sectional View of
Brachytherapy Seed Line
Bottom o f Prostate
(Rectum End)
Figure 1 J (b) Transperineal View of Brachytherapy Seed Line Configuration. Consistent with the
modified uniform loading approach, the seeds were configured in seed lines spaced 1cm apart throughout the
Planning Treatment Volume (PTV). A cross-sectional view of the seed lines is depicted. While a
homogeneous distribution of seed lines is presented in this schematic, the number of rows and the number of
seed lines in each row was dictated primarily by the size of the PTV and the shape o f the prostate gland. For
example, there would be 1-3 lines of seeds at the bladder and rectum ends of the PTV.
1.3.3
Incidence of Prostate Cancer Recurrence Following First-Line Brachytherapy
While cure rates for prostate cancer are steadily improving, with 5-year actual
survival rising from 67% in 1976 to 93% in 1994 (Parkin, 1999), still there is a significant
portion of the population inflicted by recurrence of the disease.
For Brachytherapy
treatments in particular, Radge et al, reported ten-year disease free survival at 66% for
patients with clinically organ-confined prostate carcinoma and treated with brachytherapy
only (Radge et al., 2000). According to the American Society for Therapeutic Radiology
and Oncology (ASTRO), a disease recurrence (or biochemical failure) is defined as three
consecutive rises in PSA levels, with each measurement taken at least three months apart.
Once a biochemical failure has been established, it is necessary to determine if the active
cancer is local or systemic, as this information dictates the next form of treatment.
7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
1.3.4
Treatment Options for Prostate Cancer Recurrence Following Brachytherapy
For a local recurrence of prostate cancer following Brachytherapy, there are some
treatment options available-each with their own merits and shortcomings. While hormonal
therapy is the most common treatment used after any radiotherapy failure, it is not
curative, as it is employed only to control tumour growth. Generally, the objective of
hormonal therapy then is to keep cancer in remission in patients who are likely to die from
something else.
Unfortunately, prostate tumours inevitably progress to an androgen-
resistant state, thereby nullifying the intended effect of the hormonal treatment. Therefore,
salvage therapies are now being designed with curative intent.
In the case of
Brachytherapy failure, local salvage therapy may be a viable alternative so long as the
disease is still confined to the prostate gland. To my knowledge, there is no published data
on well-established salvage therapy options following failed Brachytherapy. Thus, there is
an evident need to develop and/or investigate a treatment option for the recurrence of
localized prostate cancer following Brachytherapy.
8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
1.4
Thermal Therapy
1.4.1
Clinical Implementation of Microwave Thermal Therapy
Thermal therapy works to destroy solid tumours with the use of heat. The success
of current Phase I/H clinical trials using interstitial thermal therapy to treat recurrent
prostate cancer following EBRT at the Princess Margaret Hospital in Toronto, Canada, has
encouraged the investigation of this treatment as a salvage therapy option for failed
Brachytherapy patients.
In clinical trials of prostate thermal therapy, microwave radiation is used to
generate heat, which destroys the tumour volume. An array of S helical antennae, which
operate at 9ISMHz, are inserted transperineally into the prostate under transrectal
ultrasound guidance. The use of an antenna array serves to ensure adequate thermal
coverage of the tumour volume.
The antennas are driven incoherently to prevent
electromagnetic interference between antennae. This ensures that the power deposition
pattern characteristic of a single antenna is retained, which thus simplifies the
determination of the entire antenna array, heating pattern. With accurate knowledge of
this pattern then, localization of heating to the target area and preservation of the critical
tissue structures can be achieved. (Sherar et al., 2001)
This type of thermal therapy results in temperatures (55°-90°C) which are
destructive to the tumour, but which are also hazardous to normal tissue structures
including the bladder, rectum and urethra. In order to prevent damage to these tissues,
methods are employed to reduce the amount of thermal toxicity these tissues are subjected
to. Active cooling of the urethra is achieved with a modified Foley catheter (Sherar et al,
2001).
It is inserted into the urethra in the form of a closed loop system, which
recirculates water that is actively maintained at temperatures ranging from 5° to 10°C.
Efforts to protect the rectum are achieved with ‘hydrodissection’ (Lancaster et al, 1999,
Sherar et al, 2001).
This involves inserting a needle transperineally into the region
between the prostate and the rectum. A sterile solution of saline is injected into this
9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
intracavitary space thereby lifting the prostate off of the rectum. This space thus permits
the application of much higher temperatures at the posterior aspect of the prostate, while
preserving the rectum beneath. During the course of the thermal treatment of recurrent
prostate cancer, thermometry both in the target and normal tissues provides feedback as to
the needed power levels of the antennas, which can be adjusted in real time.
The results of these Transperineai Interstitial Microwave Therapy (TIMT) clinical
trials for the treatment of recurrent prostate cancer, have demonstrated that for 25 patients
at 6 months post treatment, 52% had PSA levels < 0.5ng/ml, while 40% of patients had
PSA levels between 0.51ng/ml and 4ng/ml. In addition, 64% of patients had negative
biopsies (Sherar, M.D. et al, 2001). The success of these trials have in part motivated the
investigation of its use for Brachytherapy failures, as the current salvage treatment options
available to this population of patients are limited.
1.4.2
Biological Rationale
At temperatures between approximately 55°C and below 90°C, maintained for a
specific amount of time, the resulting thermal dose is sufficient to cause coagulation,
resulting in necrosis of the tumour volume. Coagulation involves the denaturation of the
structural proteins that form the extracellular matrix and cytoskeleton of all cells.
Following thermal denaturation of proteins, cell changes including collagen denaturation
(measured by birefringence changes) and hyalinization, muscle cell birefringence changes
and cell shrinkage. (Thomsen, 1999) Following thermally induced coagulation, the
epithelial cells such as in the prostate gland tend to elongate from their usual columnar
shape to form thin spindled cells. Subsequently, this epithelial spindling forms a distinct
thermal damage boundary that can be used to map the extent of thermal damage
(Thomsen, 1999). In order to ensure that this damage results in complete necrosis o f the
desired tumour volume, a specific dose of thermal energy is required.
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I. INTRODUCTION
The thermal dose necessary to induce coagulation is dependent on a number of
factors including the anatomy of the target tissue, the thermal properties of the target, the
tissue temperature, and the duration at which those temperatures are maintained.
(Thomsen, 1999). The interdependence of the latter two parameters, in conjunction with
the activation energy (E») and the frequency factor (Afreq), together define the Arrhenius
damage integral, which quantifies the amount of thermal damage achieved (Henriques,
1947) The Arrhenius damage integral is defined as follows:
£2(r,t) = \ Afreq exp[-Ea/(RT(t))]dt
0<t
(1.1)
where
£2(r,t) = -ln(c(t)/c(0))
(1.2)
Where £2(r,t) is a dimensionless damage coefficient defined by the concentration of
native undamaged tissue c(t) after a time t of heating, relative to the concentration of
native undamaged tissue before heating begins c(0). The activation energy Ea describes
the amount of thermal energy required for coagulation to occur. The frequency factor
Afreqdescribes the rate of thermally induced molecular agitation and R is the universal
gas constant.
Therefore, while time and temperature are the factors, which will
predominantly ensure the thermal dose necessary to induce coagulation, selection of the
appropriate energy source and corresponding energy delivery system are also of critical
importance, as they dictate the power deposition pattern that subsequently shapes the
resulting temperature distribution.
1.4.3
Thermal Therapy Modalities
Thermal therapy applicators are designed specific to the intended clinical need,
that is, to suit particular anatomical and physiological requirements.
A number of
thermal therapy modalities for the treatment of localized carcinomas are currently being
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I. INTRODUCTION
studied.
These include Microwave, Radio Frequency, Ultrasound and Laser Thermal
Therapy, each of which uses extreme temperatures to induce some type of biological
effect. This may vary from thermally induced coagulation as in TIMT or ablation as in RF
and ultrasound thermal therapy.
RF Thermal Therapy
There are many applicators used in RF thermal therapy, but the electrode currently
being studied in this laboratory is the umbrella electrode, which operates at a frequency of
460KHz. It is used interstitially for the treatment of solid tumours. Depending on the
diameter of the desired area targeted for heating, the electrode will have a specific number
of tines suitable to ensure complete heating of the tumour volume (McGahan and Dodd,
2000) (see figure 1.4). Consequently, generally only one umbrella electrode is required to
heat the entire tumour, unlike other thermal therapy modalities, which require an array of
applicators. The umbrella electrode is inserted percutaneously into the tumour volume
under CT, MRI, or ultrasound guidance (Grasso et al, 2000).
Figure 1.4. Radiofrequency ablation of a tum our volume with the umbrella electrode. This 10-tine
umbrella-shaped applicator, which operates at a frequency o f 460kHz, is pictorially shown to completely
encompass the tumour volume, subsequently inducing coagulation, (http://www.radiotherapeutics.com/)
12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
Laser Thermal Therapy
Laser thermal therapy (LTT) delivers light via a fiber optic system to the tumour
volume. The specific wavelength of the light is chosen based on the depth of penetration
required, which thus depends on the location and size of the tumour in the body. For
example, light with longer wavelengths have fewer scattering events and can thus
penetrate to deeper areas of the body (Roggan, Muller et al., 1995). While the interstitial
delivery of LTT makes it suitable for highly localized energy deposition, it has the
potential to induce temperatures which result in charring o f the tissue. This can occur as a
result of the small aperture through which the light is delivered, as well as the mode in
which the laser is operated. The continuous wave (CW) mode for example, has been
shown to cause tissue ablation, that is, the complete removal of biological material
(LeCarpentier, 1993) Thus the potential for this applicator to induce charring limits its use
to those targets where complete ablation would not create any additional health-related
problems for the patient.
Ultrasound Thermal Therapy
Ultrasound is another experimental thermal therapy modality used in the
treatment of solid tumours. In this case, sound waves emitted at frequencies ranging
from 500kHz-5MHz are used to heat the tumour (G. Ter Haar, 1995). Thermal therapy
via ultrasound may be either applied interstitially or externally.
Ultrasound therapy
applied externally is non-invasive and can be quite accurately focused to target deepseated tumours. Interstitial ultrasound thermal therapy is minimally invasive and targets
localized treatment regions.
Both modes can be subject to problems due to the
interactions of the ultrasound waves with bone and air, where almost complete reflection
occurs. This can cause bone pain and inhomogeneous heating in the target. (G. Ter Haar,
2001).
13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
For therapeutic results, tissue ablation via ultrasound thermal therapy employs a phased
array system, where the ultrasound beam is scanned electronically across the treatment
region or where energy can be deposited with static or scanned multiple-focus patterns.
However, fabrication of such large-aperture phased-array transducers is difficult and thus
limits the viability of ultrasound thermal therapy as a treatment option at this time. (Hong
Wan, 1999)
Figure 1.5. Thermal lesion created by square-shaped ultrasound beam forming aperture for
ultrasound generated thermal therapy. (Kolios, 2001)
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
1.5
Microwave Thermal Therapy
1.5.1
Electrical Tissue Response to Microwave Radiation
Microwave thermal therapy uses microwave energy, typically at a frequency of
9 15MHz, to heat the tumour. In this frequency range, the electromagnetic properties of
tissue are determined primarily by the behaviour of water molecules in the tissue.
Generally, the higher the water content in the tissue, the higher the dissipation of the
energy. In muscle tissue in particular, there is a relatively large water content, as opposed
to fat tissue which contains little water.
As microwave radiation propagates through
tissue, the resulting electric field causes a separation of charges until a sufficient distance
between the charges is reached, at which point an internal electric field is created. While
charges are moving in response to the external electric field and before equilibrium is
reached by the balancing of the internal field, energy is dissipated in proportion to the
amount of charge and distance moved. This dissipation process resulting from charge
movement describes the electrical conduction through the medium. This can be stated
mathematically as:
J c = oE
1.3
where J c is a vector quantity that denotes the movement of charges or current, a (S/m) is
the electrical conductivity of the tissue through which the wave propagates and E (V/m) is
a vector quantity describing the resulting electric field. It is this movement of particles
that causes heating of the medium, which subsequently induces the desired coagulative
effect.
1.5.2
Microwave Applicators
Microwave radiation can be delivered to the target volume externally or internally.
External radiators and short interstitial radiating antennas may be used in either single
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
mode or array configurations. A number of interstitial applicators have been tested for use
in microwave thermal therapy, as in the dipole, multipole and helical antennas. The
helical coil antenna consists of a basic dipole design with a bare winding around the
insulation of the inner conductor, whereas the dipole antenna consists of a flexible
miniature coaxial cable with an extension of the inner conductor at the antenna junction.
The construction of the multipole is very near to that of the dipole except that there are
multiple junctions. Limitations of the dipole and multipole antennas specifically for use in
thermal therapy, include the variability of the heating profile with different insertion
depths, restricted range of possible heating lengths for a given microwave frequency and
inadequately heated ‘dead length’ at the antenna tip (Satoh, 1988 and Ryan, 1991). The
helical antenna on the other hand is not subject to these same limitations as the antenna hot
spot is located very near the tip of the antenna, rather than at the junction. As a result,
over-insertion of the applicator past the treatment point is unnecessary, thereby making the
helical antenna ideal for such therapeutic applications (Satoh, 1988).
The helical antenna emits circularly polarized radiation, that is, polarization
whereby the electric field vector remains constant in length but rotates back and forth
tracing out a circular (or elliptical) path. For biomedical applications, as in interstitial
microwave therapy, the helical antenna operates in the normal mode, which is defined
such that the maximum far field radiation is in the plane perpendicular to the axis of the
helix. Operation of the helical antenna in the normal mode results when the dimensions of
the helix are very small compared to the source wavelength (Lee, 1984). The result is an
energy source whereby the maximum power deposited is biased towards the tip of the
antenna (i.e.: near the end of the helical coils).
The most significant advantage of the helical antenna is that the circularly
polarized nature of the radiation ensures localization of heating to the region surrounding
the coil element and there is no potential for tissue charring. As a result, there are no
heating peaks along the feedline nor is there overheating at the entrance site of the
antenna, as was demonstrated by Satoh et al with the half-wavelength dipole antenna-a
linearly polarized radiator. This suggests then that for treatment o f prostate cancer, where
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
the size of the prostate gland can vary markedly from patient to patient, as can its
proximity to healthy tissue structures, the helical antenna is the most attractive applicator
because it ensures knowledge of the location of the heated region, regardless of these
patient specific variables.
However, for patients where Brachytherapy has failed, the
presence of the spent metal seeds in the vicinity of the helical applicator may change the
shape of the heating pattern and in tum compromise the localization of heating to the
target volume.
1.5.3
Specific Absorption Rate (SAR)
Specific Absorption Rate (SAR) pattern, SAR(x,y,z), also called heating pattern
or power deposition pattern, is a useful parameter in thermal therapy because it identifies
both the magnitude and distribution of energy deposited in the tissue. Specifically, it is
defined as the rate of energy E absorbed or dissipated in an incremental mass contained
in an incremental volume V of a given density rho (p). Mathematically, SAR can be
expressed as:
SAR(x,y,z) = d/dt[dE/pdV] (W/kg) (1.4)
By multiplying equation 1.4 by the density p, SAR can also be defined as Absorption
Rate Density (ARD) and is expressed in terms of power per unit volume (W/m3).
SAR can be determined theoretically with the use of modeling, or empirically
through measurements. SAR can be measured directly with the use of an E-field probe,
which measures the Held strength of the applicator. (Brehonnet, 2000) However, electric
Held measurements are challenging because small, non-perturbing, probes are difficult to
build,
hi addition, use of an E-field probe for direct SAR measurements may
compromise the accuracy of the results, as it can perturb the applicator field pattern
(Iskander and Tumeh, 1989).
The mathematical relationship that describes the
relationship between spatially distributed power deposition and electric Held strength is
as follows:
17
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
SAR (x,y,z) = (CT/p) IE (x,y,z) 12 (W/kg)
(1.5)
where E(x,y,z) is the electric field vector (V/m), and a is the electrical conductivity of the
medium (S/m). (Ryan et al., 1990)
The SAR pattern can also be determined indirectly by measuring temperature rise
in the tissue or an electromagnetic phantom to which the power was deposited. If the
material is not experiencing a change in phase, the energy source term can be expressed
as:
SAR = pcc)T/dt
(1.6) (W/m3)
where pcdT/dt is the time rate of change of the thermal energy of the medium per unit
volume. (Incropera and DeWitt, 1996) Therefore, for a power source in an infinite
homogeneous medium of heat capacity c and density p, the initial temperature rise due to
the conversion of deposited energy to heat, can be measured. According to Roemer and
Fletcher (1985), it is possible then to determine specific absorption rate (SAR) by
measuring the temperature rise from equilibrium following a step input of power.
Assuming that only heat generation occurs, that is, there is no heat flow, then the
conduction term in the heat conduction equation is zero and thus the specific absorption
rate W is linearly related to the initial temperature rise in the medium. In order to create
a situation then where there is no heat flow, temperature rise measurements must be made
following a very short period of heating such that conduction of the energy is minimized.
pcdT/dt =SA R + /tV2T
(1.7)
where k is the thermal conductivity of the material (assuming constant k), T is the
temperature and /tV ^ is the conduction term. Over very short periods of heating,
0, and the above equation reduces to:
pcdT/dt 11=0* = SAR (1.8)
18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
=
CHAPTER 1. INTRODUCTION
Integrating equation 1.8 subsequently yields:
SAR=pcAT/At
(1.9)
where AT = Tf(x,y,z) - Tj(x,y,z) is the temperature rise at a point (x,y,z) in the medium
over some heating interval At. Thus ‘the SAR at a point in the medium is linearly related
to the temperature rise at that point that results from a brief application of power’
(Gladman, 1996). The measurement of temperature rise for the indirect determination of
SAR can be achieved with the use of thermography or thermometry.
1.5.4
Infrared Thermography
An infrared camera can be used to measure temperature rise on a heated surface,
which can be used to determine an SAR pattern. Infrared imagers are used to observe
and/or measure heat emissions radiated from a surface via non-contact mechanisms. All
objects at temperatures above 0°K emit and absorb radiation at frequencies in the
electromagnetic spectrum between visible light and radio waves. The IR portion of the
spectrum spans wavelengths from 0.7pm to 1000pm. Different materials have different
emissivities, that is, the ability to absorb and reflect radiation, and will therefore emit IR
energy at different intensities for a given temperature. The benefits of thermography for
the purposes of SAR measurements include its ability to provide spatial temperature
information, as SAR is a spatially distributed parameter. For the purposes of 2-D SAR
measurements then, this ensures that only the energy on the applicator plane contributes
to the resultant thermal image. While many groups have used thermography for indirect
SAR determination (Khizhnyak, 1994, Gross, 1990), there are some limitations
associated with the technique used to acquire thermographic images. One such limitation
is the presence of a conduction artifact, which results from the time required for the
applicator to deposit power and subsequently generate heat.
Another limitation, the
‘cooling’ artifact, is defined by Gladman (1996) as being the artifact associated with both
19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I. INTRODUCTION
the conduction and convection of heat once the applicator has been turned off. While
both of these artifacts are important factors in dictating thermographical measurement
accuracy, they can be minimized if specific experimental measurement techniques are
employed. The accuracy of the SAR measurements is improved if applicator heating is
short enough to limit the effects of thermal diffusion. According to Gladman (1996), a
heating time of 20 seconds does not compromise the accuracy of the SAR measurements,
however shorter heating times will even further reduce the effects of thermal diffusion, as
thermal diffusion length is related to heating time (t) according to L* = [V(4Jfc/pc)*t]. In
addition, fast image acquisition and minimization of airflow across the phantom surface
post-heating will significantly reduce the artifacts in the SAR measurements due to the
conductive and convective cooling mechanisms respectively.
l.S.S
Magnetic Resonance (MR) Thermometry
MR thermometry can also enable the determination of spatial temperature
information.
Therefore, it can also be used for indirect determination of SAR.
(Raaymakers, 2001)
Unlike the IR camera, it is not subject to the same limitations
associated with the problems of conductive and convective cooling, as the measurements
are obtained interstitially, in real time. MR thermometry provides good spatial resolution
(Graham et al, 1999) but poor temperature resolution in comparison to that provided by
invasive thermometry.
Generally, a temperature measurement accuracy of 1°C is
achieved with MR thermometry (Bohris, 1999). Also, MR compatibility is a significant
consideration, which can limit the use of this measurement device for certain
applications. In addition, the significant expense associated with this equipment and the
difficulty accessing it make this tool a much less viable option for acquiring spatial
temperature and subsequent SAR measurements.
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
1.5.6
Discrete Thermometry
Temperature rise at a single point in space can be measured with a temperature
sensor.
Thermometry is used in many clinical and experimental hyperthermia and
thermal therapy applications as a means to quantify the temperature rise induced in tissue
during the course of heating (Stauffer et al, 1987 and Sherar et al, 2001).
Unlike
thermography and MR thermometry, point thermometry cannot provide as dense a
temperature data set, but is not subject to the cooling artifacts inherent in thermographical
temperature measurements. Many thermometry systems can measure temperature during
the course of heating, which is useful for identifying the time-dependent thermal response
of a material. The choice of a thermometry system depends on the nature of the energy
source. In the case of electromagnetic microwave applicators, as in the helical antenna,
fluoroptical-based thermometry systems are most suitable for temperature rise
measurements, as opposed to a metal tipped temperature probe (Ryan, 1990), in order to
prevent the potential disturbance of the electric field and the temperature readings.
A Luxtron Fluoroptic® Thermometer Model 3100 can be used to make discrete
temperature measurements.
A temperature sensor mounted on the end of the probe
consists of a small quantity of manganese-activated magnesium fluorogermanate, a
temperature sensitive phosphor. A xenon flash lamp excites the phosphor with a pulse of
blue-violet light, thus causing it to fluoresce. As the intensity of the fluorescent radiation
decays, the decay time is measured and then correlated with the phosphor temperature by
comparing the measured decay time with a digital look-up table. (Luxtron 3100 Series
Operating Manual, 1993).
21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
1.6
Therm al Conduction
Thermal therapy is characterized by the application of high temperatures to a
target tissue for extended periods, to ensure thermal damage sufficient to destroy the
tumour is achieved. The resulting temperature distribution (in a non-perfused region) is
determined primarily by the SAR pattern characteristic of the applicator and its
electromagnetic environment, in this case the surrounding Brachytherapy seeds, and the
conductive mechanisms of heat transfer. Thermal conduction can be described as the
transfer of heat across a medium from a region of higher temperature to a region of lower
temperature, due to the presence of a temperature gradient in a stationary medium. This
flow of thermal energy is sustained by physical processes at the molecular level.
Specifically, higher temperatures are associated with higher molecular energies (kinetic
energy), and when neighboring molecules collide, a transfer of energy from the more
energetic to the less energetic particles occurs. This net transfer o f energy by random
molecular motion is referred to as diffusion of energy and represents the mechanism of
conductive heat transfer. (Incropera and DeWitt, 1996) In the presence of an energy
source, conduction of this energy will occur due to the induced temperature differential
thereby causing the temperature in the medium to rise. This relationship is described by
the Heat Conduction Equation.
pc0T/at = W +
(1.10)
where W is the energy source term (W/m3) or characteristic energy input of the applicator
(SAR), and kVT is the heat flow in 3-D space. The proportionality constant k is the
thermal conductivity parameter, which describes the ability of a material to conduct
thermal energy a distance of lm for every degree in temperature rise. In the case of
solids in particular, the conductive transport of thermal energy is due to two effects: the
migration o f free electrons and lattice vibrational waves. These effects are additive such
22
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
that the thermal conductivity k is the sum of the electronic component if* and the lattice
component k\. Thus k = kc + k\. For pure metals, as in titanium, k is determined primarily
by ke, as it is generally, inversely proportional to the electrical resistivity /%, which in pure
metals is usually very small. Metal alloys have somewhat larger resistivity values and
thus the contribution of k\ to k, is no longer negligible. For nonmetallic solids on the
other hand, as in biological or phantom-equivalent tissue, k is determined primarily by k\ ,
which depends on the frequency of the interactions between the atoms of the lattice.
Evidently then, the thermal conductivity of a material is an important parameter in
determining the distribution of the temperature rise in a medium.
Heat capacity (pc) describes the ability of a material to store thermal energy,
which is in turn indicative of the extent of the induced temperature rise. Heat capacity
describes the quantity of energy necessary to increase the temperature of a medium by 1
degree Kelvin.
Therefore, for two materials with largely different heat capacities,
generally speaking, the medium with the larger heat capacity will be better able to store
heat, subsequently requiring more energy with which to heat it to temperatures
comparable to those induced in the material with a lower specific heat value.
Further insight into the thermal response of a material (defined by temperature
rise and distribution) is gained upon consideration of these thermophysical parameters (k,
p, c) together as a single entity, known as thermal diffusivity a . It is the ratio of thermal
conductivity (k) to the heat capacity (pc) and is expressed in units of m2/s. It is the
controlling transport property of transient conduction. Specifically, thermal diffusivity is
defined as follows:
a = k / (pc)
(1.11)
Thermal diffusivity measures the ability of a material to conduct thermal energy relative
to its ability to store thermal energy (Incropera and DeWitt, 1996). Materials of large a
will respond quickly to changes in their thermal environment, while materials of small a
will respond more sluggishly, taking longer to reach a new equilibrium condition.
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I. INTRODUCTION
The thermophysical properties of titanium and a muscle-equivalent tissue phantom are as
follows:
Titanium
PAG Phantom
Real Muscle Tissue
Thermal
Conductivity o
(YV/mK)
22
0.535
0.395
Specific Heat c
(J/kgK)
Density p (kg/m3)
523
3810
3470
4510
1070
1070
Table i.I. Thermal and Material Properties of Titanium and PAG Phantom. (Suroweic et al., 1992,
CRC Handbook of Chemistry and Physics, 2001)
Titanium has a thermal diffusivity approximately 72 times greater than the thermal
diffusivity of phantom material. This suggests that tissue containing Brachytherapy seeds
may require less time to reach thermal equilibrium. In addition, the increased thermal
conductivity of the titanium seeds suggests the potential for increased spreading of the
temperature distribution. Also, the heat capacity quantity (pc) of titanium is almost half
that of phantom material.
This suggests that titanium has a limited thermal storage
capacity as compared to phantom and therefore requires less energy with which to heat
up to the same temperature as phantom.
Materials with different thermophysical
properties will respond differently to the same thermal environment, meaning the
magnitude and extent of the resulting temperature rise distribution, may vary. During
thermal therapy then, tissue containing titanium seeds may have different temperature
distributions than tissue with no seeds.
24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
1.7
Theoretical Modeling
1.7.1
Overview: Methods of Modeling
Thermography and thermometry can be used to measure temperature rise for the
purposes of indirect SAR determination and following periods of extended heating.
Determination of the SAR and the resulting temperature rise induced over long-term
heating is essential to characterizing the efficacy of microwave thermal therapy. These
parameters can be determined using empirical or theoretical means.
determination may be either analytical or numerical.
Theoretical
While analytical solutions are
inherently more accurate, they only exist for simple geometries. Therefore, numerical
methods are commonly employed to solve complex problems, for which an analytical
solution fails to exist and for which development of an experimental model is very
difficult. In addition, because numerical modeling can be used to predict in advance the
ideal outcome without the complications inherent in the experimental model, it can also
be used to help identify and understand the nature of those complications that limit the
detectability and/or accuracy of the experimental results.
Numerical modeling involves breaking up the domain of interest into elements
and then applying the differential equations representative of the problem to be solved, to
each small element. Two commonly used numerical methods are the finite difference
method and the finite element method.
The primary difference between the finite
difference and finite element method is that the finite element method approximates a
solution throughout an element (a 2-D or 3-D region) instead of at a node, a single point
in space, as in the finite difference method.
As a result, the finite element method
requires increased computation times and memory requirements but is capable of solving
problems for more complex geometries. This is because the elements, which are defined
by a number of sides as opposed to a particular shape, are thus capable of conforming to
almost any geometry and can thus more accurately represent it, unlike the rigid nodal
network (grid) of the finite difference method.
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
1.7.2
Numerical Modeling of SAR
SAR can be modeled using either the finite element (Kumaradas, 2001, Jia, 1994)
or the finite difference time domain (FDTD) methods (Rossetto, 2000 and Cherry and
Iskander, 1993). In the case of the finite element method, there are two techniques by
which to implement the finite element method, node-based and edge based.
The
selection of the appropriate technique is dictated by the nature of the parameter to be
solved. The spatially distributed SAR pattern characterizes the power deposition pattern
of the applicator. Thus SAR is directly related to the electric field, which is defined by
Maxwell’s equations. The nature of these equations and the boundary conditions they
impose make the edge based finite element method more suitable for the determination of
electric field, and subsequently SAR. In the node-based finite element method of solving
the electric field distribution, the electric field is forced to be continuous at the surfaces
between the elements, which in fact contradicts the continuity condition imposed by
Maxwell’s equations. The edge-based approach to the finite element method on the other
hand is most suitable for solving the inhomogeneous Helmholtz wave equation, a
derivation from Maxwell’s equations, as it maintains continuity of the tangential
component of the electric field at each edge of a tetrahedral element. (Kumaradas, 2001)
In order to solve for the electric field using the edge based finite element method, the
inhomogeneous Helmholtz differential wave equation is used. (Kumaradas, 2001) In the
case of an electric field (E) generated by a source in the microwave frequency range and
in the absence of excitation currents, the equation becomes:
VxVxE-^EsO
(1.12)
where ifc2 = arpie- jwfJocr. The constants //a n d £ represent the magnetic permeability and
the electric permittivity respectively. The electrical conductivity is <xand eo = 27if where
f is the excitation frequency. (Lorrain and Corson, 1962)
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
The electric field E represents the total electric field, which is equal to the incident
electric field E uk plus the scattered electric field Es.
E = Einc + Es
(1.13)
The edge-based finite element model for determination of the SAR distribution
was developed by Dr. J. Carl Kumaradas. The model is equipped to handle a user
specified incident field, a radially polarized incident field as in the field described by a
coaxial cable and a plane wave field specification. The model is also capable of dealing
with metal objects, which are modeled as perfect conductors.
1.7.3
Numerical Modeling of Temperature Distribution
Temperature distribution can also be modeled using either the finite element or
the finite difference methods. For the purposes of thermal therapy treatment planning,
the finite element method is well suited for the complex geometries associated with the
tissues, blood vessels and other physiological structures. (Martin, 1992, Rine, 1987) In
the case of simpler geometries however, the finite difference technique is also suitable
(Rosetto, 2000).
In addition, although implementation of the well-established finite
difference method for the purposes of thermal modeling can be done with greater ease,
the finite element method is better equipped to handle inhomogeneities in the model.
This arises from the fact that an element in the finite element method is representative of
one material only.
In the finite difference method, an interior node for example is
centered amidst 4 other nodes, each of which may or may not belong to the same
material. As each of these nodes are used in the determination of the temperature at the
interior node, the simple mathematical formulation that described the finite difference
method is much more complex and difficult to solve. Depending on the nature of the
nodal network and the extent of the material inhomogeneity, solution generation may not
be possible.
27
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I. INTRODUCTION
Whether the finite element or the finite difference method is used for thermal
therapy modeling, determination of the temperature distribution or heat flow can be
modeled using the Pennes bioheat transfer equation (BHTE) (Pennes, 1948 and Rosetto,
2000) or the Effective Thermal Conductivity Equation (ETCE) for example. The former
model is described by the heat conduction equation, which incorporates a convective
term to account for the heat sink resulting from the perfusion of blood.
pc3T/9t = W + k V ^
-CbWb(T-Ta)
(1.14)
where W is the energy source term (W/cm3), and /fcVT is the conductive heat flow in 3-D
space. The proportionality constant k (W /c m 'K 1) is the thermal conductivity and p and
c are the density and specific heat of the tissue respectively.
The term ct,wt(T-Ta)
describes the heat sink as a result of blood in the microvasculature with w*, (g/cm 'V )
representing the blood perfusion term. In the case of modeling a homogeneous medium
as in phantom-equivalent tissue, a heat conduction equation (see Eq. 1.10) (which lacks
the heat sink term) can be used to model the simple transfer of heat.
28
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I. INTRODUCTION
1.8
Summary and Objectives
Brachytherapy as a first line therapy for the treatment of localized prostate cancer
is now established as a standard treatment. However, statistics on cancer recurrence
following Brachytherapy indicate many patients will fail. As a result, there is a need for
secondary treatment options for these patients. The success of current clinical trials using
TIMT for the treatment o f recurrent prostate cancer following failed EBRT, has therefore
motivated the investigation of its use as a potential salvage treatment for failed
Brachytherapy patients.
The efficacy of this salvage treatment will be determined
primarily by the thermal dose administered to the tumour and the localization of this
heating to the target volume. Thus anything that could compromise these treatment
efficacy determinants, in turn compromises the viability of this salvage treatment option.
Due to the contrasting electrical and thermal properties of the seeds in comparison
to the surrounding tissue, they may scatter the incident electric field characteristic of the
microwave helical antenna, and increase the magnitude and extent of the temperature
distribution resulting from microwave heating. The presence of the Brachytherapy seeds
in the tissue may compromise the localization of heating to the target volume and the
thermal dose necessary to induce coagulation in the target. In order to determine the
feasibility of this proposed salvage treatment, experiments and theoretical models were
developed and results generated for both the control (no-seed) and seed cases. The goal
was to measure and quantify the deviation of the seed case from the no-seed case and
subsequently determine if this deviation might compromise the quality of the TIMT
treatment.
IR thermography was used to measure relative spatial temperature information for
the purposes of both indirectly measuring SAR patterns, and temperature distributions in
a muscle-equivalent tissue phantom with and without Brachytherapy seeds.
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
We developed an experimental technique for measuring and reconstructing the 3-D SAR
patterns as well as the 3-D extended heating temperature distributions, in phantom tissue,
using Brachytherapy seed arrangements consistent with standard clinical seed loading
protocols. The widths and lengths of SAR contours (obtained from multiplanar SAR
contour patterns) were measured, to quantify the extent of electric scattering induced by
the seeds. Similarly, the widths of temperature profiles (obtained from the temperature
distributions) were measured, such that the extent and thus localization of the temperature
distribution in the presence of seeds could be determined. A Fluoroptical-based Luxtron
thermometry system was used as an additional temperature measurement tool, for the
purposes of comparing temperatures rises induced over extended heating periods, in
phantoms with and without seeds.
While experimental models are generally employed to validate theoretical models,
the latter that can subsequently provide a more efficient means of evaluating complex
problems, this project utilized theoretical models in a slightly different capacity.
Numerical models were used to help understand and confirm the experimental results.
Plane wave simulations based on an edge-based finite element method were run such that
the scattering effects of the seeds could be theoretically determined and compared with
experimental SAR results. In addition, thermal finite element models based on the heat
conduction equation also served to theoretically quantify the temperature distribution
resulting in phantom tissue with seeds.
30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 2
METHODOLOGY
2.1
Introduction
The primary objective of this project was to determine the electromagnetic and/or thermal
impact of brachytherapy seeds on the SAR patterns and extended heating temperature
distributions characteristic of 9 15MHz helical antennae used for prostate thermal therapy.
The problem was studied using theoretical models and experiments in a muscleequivalent polyacrylamide phantom material. The electromagnetic field scattering effect
of the brachytherapy seeds was determined experimentally using 3-D reconstructions of
coronal SAR measurements in phantoms measured using an IR camera. The problem
was simulated theoretically using a finite element model. Over extended heating periods,
IR measurements and simulated temperature distributions were determined in order to
quantify the thermal impact of the brachytherapy seeds in a high temperature
environment. In addition to the IR measurements, a discrete thermometry system was
also employed for validation purposes.
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
Together, the experiments and simulations were designed to produce a complimentary set
of results from which the electromagnetic and thermal impact of the brachytherapy seeds
could be determined. This chapter is organized into two sections. The first, section 2.2,
contains an overview of the materials, methods and calibration procedures. The second
section, 2.3, includes a detailed explanation of the experimental and theoretical steps
employed to determine the scattering and thermal effect of the brachytherapy seeds in an
electromagnetic Held for extended periods.
2.2
Overview: Materials and Methods, Calibration and Operation Procedures
An interstitial helical antenna, powered by a commercial microwave generator, was used
to heat polyacrylamide phantom layers compiled into a solid volume, with and without
brachytherapy seeds. Subsequently, the acquisition of thermometry data was achieved
with an IR camera as well as fluoroptic temperature probes. Simulations employing the
finite element method for theoretical electric field and temperature distribution
determination, made use of a number of commercial software packages, as well as
software designed in-house by Dr. J. Carl Kumaradas.
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
2.2.1
Thermal Imaging System
IR Camera
The Thermal imaging system consisted of the AGEMA Thermovision® 400
infrared camera (figure 2.3) and the accompanying ThermaCam Researcher 2001
software. Infrared thermography was used to measure phantom surface temperatures
following interstitial microwave heating with a helical antenna, powered by the BSD 500
Microwave Generator (BSD Medical Corp, Salt Lake City UT). This real time infrared
imaging system consisted of a thermoelectrically cooled SPRITE detector. It operates in
real time and makes use of an internal microprocessor for the automated control of
operative functions including internal drift compensation, field parameter and emissivity
calculations.
This particular radiometer has a sensitivity of 0.1°C at 30°C and an
accuracy of ±2°C with 140 IR-lines per frame and a 4000Hz IR-line frequency. The field
of view (FOV) is 25° x 25° with a minimum focus distance of 0.4m. (AGEMA Infrared
Systems, 1991)
Pixel Size Determination and Calibration o flR Camera
Prior to thermal image acquisition, coronal pixel size and absolute temperature
validation were determined. A copper pin array set in a Styrofoam base was used to
calculate and validate the pixel size of the system both in the vertical and horizontal
directions of the thermal image data set. First, the copper pin array was located in the
Field of View (FOV) of the camera. The base of the pin array was immersed in a small
water tank, connected to an external water bath. After the temperature of the water bath
was set to 80°C or 90°C (a comparable temperature range to that used in later
experiments), the copper pins were heated, subsequently emitting IR radiation detectable
by the thermal camera (see figure 2.1 (a)). The temperature range and sensitivity of the
camera were set to levels characteristic of those required for experimental purposes so as
to prevent saturation of the thermal image. In addition, the copper pin located directly
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
in the center of the pin array was connected to an unsheathed K-type thermocouple,
which itself was connected to an electronic temperature readout system (which included a
compensation circuit and digital display). As the pins were heated by the water bath
thermal images were acquired at S°C increments. From one of the thermal pin images,
the pixel size was determined by measuring the number of pixels between two copper
pins in the vertical and horizontal directions.
Knowing that the physical distance
separating two copper pins was 20mm, the corresponding pixel size was determined. In
order to ensure an accurate measurement of the copper pin separation distance in the
thermal image, multiple measurements were made of the pin position at selected
locations. Figure 2.1 (b) identifies the pin locations at which these pixel measurements
were made. With these measurements, an average was calculated and the resulting pixel
size determined. In order to validate the measured pixel size o f the system, a simple
calculation employing the knowledge of the FOV and array size was also performed (see
figure 2.2).
xl.^1
•
%
%
•
t %
4 •
•
$
x4
•
•
X8
•
1
t
t
•
«
xS
»
•
•
*
*
(b)
Figure 2.1 Empirical Pixel Size (PS) Determination. Thermal Image o f the copper pin array system is
shown in (a). Pixels were counted between adjacent copper pins, separated by a distance of 20mm.
Multiple pixel measurements were made at selected locations on the pin thermal image as shown in (b).
The average was calculated and the pixel size was then determined by dividing the average value by 20mm.
The pixel size was empirically determined as follows:
For i = 1:8, n = 1:4 and Ax is the # of pixels measured between two copper pins:
Axn = Xi+t - Xi and Axavg = XAx„/n
Therefore, Psx = 140 pixels/Axavg
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
Figure 2.2. Schematic: Theoretical Determination of Pixel Size. Given the distance of the camera lens
to the phantom surface, into addition to the array size as dictated by the IR camera, the pixel size on the
coronal phantom plane was determined. This theoretical calculation was then compared with the
experimentally determined result using the copper pin array.
With a 25° x 25° FOV and a minimum focus distance of 0.4m, the minimum pixel size is:
a = [ 180°-(90°+12.5°)] = 77.5°
tana = h/(x/2)
tan77.5° = 0.4/(x/2) ->x = 0.17735573m = y
Therefore, for a 140 x 140 array:
# pixels/mm = 140 / 177.35573 = ~ 0.79 pixels/mm
For the purposes of this project, given the physical limitations of the experimental set-up,
the focal distance was approximately 0.6m and the magnification of the image was
increased by 2 times. This resulted in a calculated pixel size of:
For 2x magnification:
FOV = 12.5° x 12.5 °
a = [180°-(90°+6.25°)] = 83.75°
tana = h/(x/2)
tan83.75° = 0.6/(x/2) ->x = 0.131421374m = y
Therefore, for a 140 x 140 array:
# pixels/mm = 140 / 131.421374 = - 1.065 pixels/mm
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
Measurement of pixel size in the horizontal and vertical directions yielded results of
1.075pixels/mm and 1.15 pixels/mm respectively, which is in good agreement with the
calculated pixel size of 1.065pixels/mm.
Absolute Temperature Verification
The absolute temperature measurements as indicated by the IR camera, were
compared to the temperature measurements indicated by the K-type thermocouple and
accompanying digital read out system. For each 5° increment, the temperature indicated
by the thermocouple attached to the center copper pin was compared to the temperature
determined by the IR camera at the corresponding pin location on the thermographic
image. It was shown that both temperature readings were within good agreement. The
agreement between the copper pin temperatures indicated by the thermocouples and the
IR camera was due to the minimization of conductive and convective thermal losses.
This was achieved by obtaining thermographic images during the course o f heating from
a surface already exposed to the camera lens.
In addition, the copper pins were
embedded in a Styrofoam insulator, which also helped to minimize artifacts due to
conduction. However, the accuracy of the absolute temperature data can be compromised
where a time delay in image acquisition exists or exposure of the heated surface to the
camera lens is required.
Overview: IR Imaging Procedure
With the IR camera mounted vertically such that the lens was positioned 0.6m
above and parallel to the bench top, the phantom slices, compiled together in a phantom
block and contained within a Styrofoam support mold, were located directly in the FOV
of the camera lens (see figure 2.3).
36
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
The phantom was positioned in a marked area of the mold to ensure registration between
each phantom layer. For the purposes of three-dimensional SAR and extended heating
temperature distribution reconstructions, it was necessary to image each phantom slice
independently. As a result, accurate and reproducible phantom positioning as well as
antenna and seed positioning were critical. A mark was made on the phantom layer
coincident with the antenna position to ensure the consistent location of the antenna for
each experiment. Prior to heating, an initial background thermal image of the phantom
was acquired. This thermal image was later subtracted from the image produced by
antenna heating to correct for variations in baseline temperature in the phantom.
Following acquisition of the background thermal image, the helical antenna was
located on the phantom surface and the upper phantom layers (also compiled into a
block) were placed on top of the antenna plane. Another Styrofoam block was used to
compress the phantom layers and the antenna to ensure complete adhesion between each
slice, to prevent artifacts in the SAR or temperature measurements due to air pockets.
After heating, the upper portion of the phantom section and the antenna were rapidly
removed. The phantom slice of interest was then imaged and the data written to a disk in
the IR camera.
IR Camera
^
Phantom ^
Styrofoam Mold
Figure 2 3 . IR Thermovision Camera and Styrofoam Insulating/Support Block for Phantom. The
camera was mounted 0.6m above the phantom surface to be imaged. The phantom was precisely located
within the Styrofoam block.
37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2 . METHODOLOGY
Software and Processing Code
The ThermaCAM Researcher 2001
software, which provided the data
manipulation and processing environment, was used to extract temperature data from the
background and antenna heating thermal images. The imaging software system allowed
the data to be stored as a 2-D data set of the absolute temperatures at each pixel location.
The 2-D temperature data set was saved as a MATLAB file type and then
imported into the MATLAB environment for subsequent SAR and temperature volume
calculation and determination. Code was written to calculate the SAR value from the
measured temperature rise data and subsequently plotted as 2-D contour patterns. The
raw data was transformed into millimeter distances using the pixel size determined on the
coronal plane and a sampling technique employing a linear interpolation function. This
code, originally developed by Gladman (1996) was adapted for the purposes of this
project, specifically the discrete temperature data input and the 3-D reconstructions. For
both 3-D SAR and extended heating 3-D temperature distribution reconstructions, similar
MATLAB routines were employed. The technique used to generate an SAR pattern from
a coronal plane was also used as a first step in the generation of 3-D SAR and extended
heating temperature distributions. 2-D temperature data obtained from multiple phantom
slices was compiled into a 3-D data set using the CAT function provided by MATLAB.
The spatial resolution in the z direction was determined by the slice thickness ( as
opposed to the FOV and data matrix size dictated by the camera). Due to physical and
practical limitations, the minimum phantom slice thickness that could be manufactured
was 2.5mm. Thus a non-interpolated temperature reading could only be obtained every
2.5mm in the axial direction. With a linear interpolation function and the corresponding
z-axis resolution, the raw temperature data along this plane was also transformed into
millimeter distances.
38
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
Subsequently, MATLAB routines employing the CONTOURF function were used to plot
the SAR volumes and extended heating temperature distributions.
2.2.2
Helical Antenna
The interstitial microwave applicator employed in the experimental treatment of
recurrent prostate cancer was a helical coil antenna (Sherar et al, 2001). The antenna is
operated in the normal mode whereby the wavelength of the electromagnetic radiation
emitted from the antenna is greater than the antenna diameter, but is comparable to its
length. The helical antenna consists of a rigid coaxial feed line 125cm in length with a
3cm extension of a copper inner conductor (0.2825mm diameter). A 3cm long helical
copper coil emitter with an inner diameter of 0.925mm and a turn density of 1.4
tums/mm is wound around the inner conductor and soldered to its tip, thus terminating it
with a short circuit. The end of the helix distal to the antenna tip is terminated by a
coaxial feed point. Both the inner conductor and the helix are made of copper, with an
electrical conductivity of 6.48xe07 S/m (CRC Handbook of Chemistry and Physics,
2001).
A polytetrafluoroethylene (PTFE) dielectric insulator surrounds the inner
conductor and the helix and is encased in an outer flourinated ethylene propylene (FEP)
sheath. Both plastics have a relative permittivity of 2.1 (CRC Handbook of Chemistry
and Physics, 2001). The antenna was designed to operate in the microwave frequency
range, specifically 915MHz and it has a maximum input power level of 25W. (see figures
2.4 and 2.5).
Figure 2.4. Digital Picture: Helical Antenna. The 915MHz interstitial helical antenna is shown beneath
a ruler. The helical portion o f the antenna extends for about 3cm. The junction separating the coils and the
coaxial feed can also be seen.
39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
Figure 2.5. Solid Model: Helical Antenna. Rhinoceros, a 3-D solid modeling program was used to draw
the helical antenna. The figure identifies the helical coils as seen in blue, the inner conductor indicated in
red, the coaxial feed line in green and the dielectric insulator.
2.2.3
Microwave Generator
The interstitial helical antenna was powered with a commercial Model 500
Hyperthermia system (BSD Medical Corp, Salt Lake City UT). The generator operates at
9 15MHz.
It was equipped with an 8-channel interstitial multiplexer, each channel
capable of a maximum power output of 50W. In addition, both the forward and reflected
power of each channel was monitored.
40
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
2.2.4
Biomedical Luxtron Temperature Probes
The measurement of point temperatures at various locations on the phantom
surface was achieved using the SMM (Luxtron, Santa Clara, CA) four-sensor probe
coupled to model 3100 Luxtron fluoroptic temperature measurement instrument.
An
optical fiber serves as the delivery tool for the temperature sensor located at its tip. A
LUXTRON molded plastic four-fiber connector couples the probe to the FOC-3100
extension cable, a 4 0 0 hard clad silica fiber. The four fibers are bundled together.
The probes have an operating range of 0° to 125°C. The hard clad silica fibers are
200pm in diameter with a Tefzel jacket, and have a response time of 0.25seconds in
stirred liquid. The probes were calibrated to a reference temperature corresponding to the
temperature of the phantom (after having allowed to equilibrate to room temperature),
which was measured using a digital thermometer. Once calibrated, the instrument had an
accuracy of ±0.10°C at the calibration temperature and a precision (repeatability) of
±0.10°C RMS at 8 samples per measurement. (LUXTRON, 1993)
The SMM Probe
1.05m
*f (mm)
90
23
20
13
SENSOR
A
B
C
0
Approx. 0.S mm 3
T hw S M M i t
1 fiffw
« * n in r p ro h m
« f f m ir i q a r i t o p ie c e s o f 2 0 0 lim c o re
Figure 2.6 Model 3100 Fiberoptic Thermometers (Courtesy of LUXTRON Operator’s Guide, 1993).
This biomedical sensor was used to measure temperature in phantoms with and without brachytherapy
seeds, following periods o f extended heating.
41
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
Unsheathed Luxtron probes (i.e.: without catheters) were used for the discrete point
thermometry measurements in the extended heating experiments.
Precise probe
positioning was critical to ensuring accurate and reproducible measurements due to the
steep temperature gradients at or near the antenna hot spot.
Another important
consideration in the placement of the probes was their location relative to the antenna and
the brachytherapy seeds. The probes were located on the side of the seeds distal to the
helical antenna. This was to ensure that the probes did not block or interfere with the
heating of the brachytherapy seeds.
2.2.5
Brachytherapy Seeds and Experimental Seed Configurations
Brachytherapy seeds consist of a radioactive iodine (l25I) substrate encased in a
laser welded titanium shell or capsule. Each brachytherapy seed is 4.5mm long and
0.8mm in diameter. For the purposes of this project, ‘dummy’ Brachytherapy seeds were
used. These dummy seeds were exact replicas of the seeds used in actual treatments,
except that they did not contain the radioactive source. Figures 2.7 and 2.8 show the
brachytherapy seeds and their size relative to a nickel, respectively.
Brachytherapy for the treatment of prostate cancer, involves the insertion of 40120 radioactive seeds into the prostate gland. In this project, we attempted to simulate, as
accurately as possible, typical Brachytherapy protocols. Different configurations of the
helical antenna relative to the seed arrangement were tested. In configuration A (CaseA), figure 2.9 (ii), two seed lines oriented parallel to the axis of the antenna were each
located 0.5cm from the antenna with the first seed of the seed line positioned directly
across from the antenna tip. For configuration B (Case-B), 2.9 (iii), two seed lines
oriented parallel to the antenna axis were positioned such that 1 line was adjacent to the
antenna (in close contact) and the other seed line was located Icm away.
42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2 . METHODOLOGY
4.5mm
4
k
0.8mm
Outside
Diameter
Silver Marker
Titanium Capsule
T
Substrate with R a d io a ctiv e Iod in e
Figure 2.7. Schematic: Brachytherapy Seed. This schematic depicts the components of a therapeutic
Brachytherapy seed with the corresponding dimensions, (http://www.implantsciences.com)
ttycomcd Atnennam Uncoseed
Figure 2.8. Brachytherapy Seeds. Six brachytherapy seeds are shown next to an American nickel for a
relative comparison of size. The seeds are 4.5mm long and 0.8mm wide. (hnp-.//www.phoenix5.org/seeds.htmi)
Configuration C (Case-C), 2.9 (iv), consisted of 6 seed lines, each line spaced 1cm apart
with the antenna located midway between the two central lines. For each case, each seed
line consisted of S brachytherapy seeds with seeds spaced 4-Smm apart. In addition, for
case-A and B, the seed line configurations on the antenna plane were also repeated on
parallel planes 1cm above and below the antenna plane (Fig. 2.10). Note figure 2.9(i)
shows the antenna in the no-seed case scenario. The probe locations shown in Fig. 2.9
are discussed in section 2.3.2.
43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
(a)
(b)
Helical
Antenna
■
^
Brachytherapy Seed
^
O" —£
o—
- Tlcm
Jlcm
ElAX
X
X Probe 1
X
Probe I
O Probe 2
O
Probe 2
(c)
( d ) ---------
No-Seed
x 1cm
b
X
X
Probe 1
o
Probe 2
• Probe 3
•
• Probe 3
i -
Tlcm
▼
X
X
Probe I __
O Probe 2
*
--------
Figure 2.9. Brachytherapy Seed Configurations. Brachytherapy seeds were configured according to 3
different protocols. In Case-A (Fig. 2.9 b), two seed lines are oriented parallel to the axis of the antenna
and located 0.5cm from the antenna. Case-B (Fig. 2.9 c), two seed lines are oriented parallel to the antenna
axis with 1 line adjacent to the antenna (in direct contact) and the other seed line located Icm away. CaseC (Fig. 2.9 d) consists of 6 seed lines, each line spaced Icm apart with the antenna located midway between
the two central lines.
44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
1
w
ii
•* i2 V
2cm
t
Figure 2.10. Schematic. Cross-Sectional (Axial) View of antenna
(♦ ) and Brachytherapy seeds (O) located on the antenna plane as
well as 1cm above and 1cm below the antenna plane.
•Antenna Plane
2.2.6
Phantom Materials
Phantom Solution
A tissue-equivalent phantom with the electrical and thermal properties of muscle
tissue was used to model the prostate gland in these experiments. Muscle data were used,
as there was no data available on the dielectric properties of human prostate tissue. A
polyacrylamide solution was prepared according to a recipe and method developed by
Suroweic et al, 1992. For a phantom volume of 1050ml, the composition of the stock
solution included a 40% acrylamide solution (60% of phantom volume by weight), de­
ionized water (33.5% by weight), NaCl (1.05% by weight), N,N’-Methylene-BisAcrylamide, 10% ammonium persulfate solution and the catalyst TEMED (Tetra-MethylEthylene-Diamine).
The stock solution was poured into a mold, where gel
polymerization then occurred following an exothermic reaction.
A summary of the
electrical and thermal properties of the polyacrylamide gel (PAG) phantom tissue
compared to human muscle tissue are shown in table 2.1.
Property
Electrical Conductivity <r(S/m)
Permittivity e (F/m)
Wavelength X (cm)
Density p (kg/m3)
Specific Heat c (J/kg°C)
Thermal Conductivity k (W/m°C)
Muscle
1.37
55
4.3
1070
3470
0.395
PAG
1.39
55.5
4.3
1070
3810
0.535
Table 2.1. Electrical and Thermal Properties o f Phantom-Equivalent Muscle Tissue at 2S°C and
Muscle Tissue at 37°C at a frequency o f 915MHz. (Suroweic et al, 1992)
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
Phantom Sheets
The use of a multilayer phantom made possible the generation of experimental 3D
SAR and temperature distributions composed from several 2-D coronal images obtained
at different layers. While spatial resolution on the coronal plane was dictated by the IR
camera, the axial resolution of the 3-D reconstructions were determined by the thickness
of the phantom layers. The use of thinner layers would improve the spatial resolution in
the axial direction, but as discussed by Suroweic et al, (1992) layers less than 2.5mm
could lead to overestimates of SAR due to the transient effect of thermal diffusivity at
these thicknesses. Due to the relatively slow rate of thermal conduction, this effect would
be most pronounced on phantom layers close to the antenna.
Gladman (1996)
demonstrated very little difference in the SAR patterns measured from phantom layers
with thicknesses of 2.5mm, 5mm and 10mm. Consequently, 2.5mm thick PAG phantom
layers were used. In addition, these layers were still thick enough to endure physical
handling such that they were not easily damaged.
Multiple phantom slices with thicknesses of 2.5mm, 1.0cm and 1.5cm, were used
to reconstruct a phantom volume of 1050ml1. In order to manufacture the PAG slices, a
specialized mold was designed (see figure 2.11). For each phantom slice thickness, a
number of molds of varying depths, each with the configuration of figure 2.11 and having
an inner rectangle with dimensions of 9cm by 11cm, were cut. 8 molds with depths of
2.5mm were prepared, while 3 molds for each of the thicker layers were also constructed.
For the 2.5mm and 1.0cm thick slices, the mold was constructed from plastic, while the
mold used to create the 1.5cm thick slices was made from panels of wood. Mylar was
1The phantom volume, was separated into two halves (each half comprised o f many thin phantom layers),
the upper (top) portion and the lower (bottom) section, which are separated by the antenna plane.
46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
then glued (with epoxy) to either side of the mold and secured with glass plates, fastened
by clips. This ensured flat slices of relatively uniform thickness and prevented leaking
before the liquid polymerized. The purpose of including thicker slices was to create
symmetry in the phantom sections above and below the antenna. In order to achieve
symmetry of the top and bottom phantom, seed configurations were replicated in both
upper and lower halves of the phantom.
Hence, the thicker slices were used to
reconstruct the bottom phantom section, while still accommodating the required location
of seeds at a depth Icm below the antenna plane. The 2.5mm slices were used to
reconstruct the upper half of the phantom such that symmetrical brachytherapy seed
configurations could be achieved, while also providing reasonable spatial resolution in
the axial direction-necessary for SAR and temperature distribution reconstruction.
Additional ‘thick’ slices were placed atop the thinner slices to ensure a phantom volume
sufficient to contain the entire heating volume. A Styrofoam block was used to compress
the complete phantom volume and to provide an insulating boundary condition such that
the rate of temperature rise with respect to the radial distance from the power source was
zero (Neumann Boundary condition).
Following the creation of the molds, the polyacrylamide solution was prepared.
Given the number of layers required and the speed of the solidifying reaction, a stock
polyacrylamide solution without the TEMED catalyst was prepared.
Following the
preparation of the various batch solutions, an appropriate amount of TEMED was added
to each batch and the completed solutions were then poured into their respective molds.
After phantom solidification, the layers were removed from the mold, wrapped in plastic
film and stored in a cold room until needed.
47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
Glass
Plates
Clip
Mold
Figure 2.11. Mold configuration for phantom layer manufacturing. The picture depicts the mold
design used to manufacture phantom slices of varying depths, including 2.5mm, 1cm and 1.5cm. While
this figure indicates a wooden mold with a 9cm x 11cm rectangle cut out, with a depth of 1.5cm, thin
plastic sections were cut to create the thinner molds with the same rectangular dimensions. A mylar sheet
is glued to either side of the mold, while glass plates, also on either side of the mold are clipped together
with the mold and the mylar. This step prevented leaking o f the polyacrylamide solution before
polymerization and helped to ensure flat layers.
2.2.7
Theoretical Modeling: Finite Element Simulations
Finite
element
simulations
were
used
to
theoretically
determine
the
electromagnetic scattering and thermal effects of the brachytherapy seeds to compare
with the experimental results. Codes for electromagnetic and thermal finite element
simulations were developed in house by Dr. J. Carl Kumaradas (Kumaradas, 2002).
Some of the modeling and processing tools utilized in the modeling are commercially
available. They included Rhinoceros3D2-a Nurbs (non-uniform rational B-spline) solid
modeling program, ICEM CFD v4.13-a mesh generation package and Kaskade 3.14-a
software package used specifically to solve the transient heat conduction problem in
2 http://www.rhino3d.com
3 http://icemcfd.com
4 http://www.zib.de/SciSoft/kaskade
48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
3 dimensions.
The electromagnetic solver created in-house iteratively solved the
differential wave equation based on Maxwell’s equations.
In addition, a number of
scripts and routines created in C++ were used to interface each step o f the finite element
approach to electromagnetic and thermal solution generation. The solutions were then
imported into MATLAB or GMV for analysis. The flow chart of figure 2.12 depicts a
simplified representation of the fundamental and sequential steps necessary for both
electromagnetic and thermal finite element simulations.
3-D
Solid Geometric
Modeling
Mesh
Generation
c=>
Edge-based
E-M Solver
Kumaradas
■=>!
ICEMCFD
Rhinoceros3D
3-D Transient
Heat Conduction
Solver
Graphical
Visualization
of Solution
MATLAB
Kaskade 3.1
Figure 2.12. Flow Chart: Fundamental Steps of the Finite Element Approach to Electromagnetic and
Thermal Simulations
49
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
2.3
SAR and Extended Heating Experiments and Simulations: A Detailed
Description
2.3.1
Infrared Imaging Experiments
Both the SAR and extended heating experiments relied on the IR camera as the
primary measurement tool.
As compared to a standard thermometry system, which
employs temperature probes and therefore provides discrete temperature data, this
thermography system offers reasonable spatial resolution (approximately 1mm) on the
coronal plane. A standard imaging protocol for multiplanar temperature measurements
was established for both SAR and extended heating experiments. These experiments
differed however in the power input to the antenna and the duration of heating. For the
SAR experiments, 10 Watts was supplied to the helical antenna for S seconds. This
heating time was selected so that SAR measurement errors due to thermal diffusion
would be minimized.
This was the shortest heating time attainable given the time
required for power deposition to occur.
For the extended heating experiments, the
antenna was powered with 8 Watts for S minutes. This heating duration was chosen to
create a temperature distribution comparable to those observed clinically and to keep
temperatures low enough such that the integrity of the phantom was not compromised5.
The acquisition of multiplanar SAR and extended heating thermal images
involved a number of steps.
Phantom layers were first compiled to reconstruct the
1050ml phantom volume. As described in section 2.2.5, only the phantom layers (2.5mm
thickness) above and including the antenna plane were imaged. A Styrofoam mold was
located within the FOV of the camera. The entire phantom volume was then positioned
in the mold and secured there with additional pieces of Styrofoam and plastic. Stacked
beneath the phantom were 8-12 glass plates, each with a thickness of 2.5mm. The
5 Polyacrylamide is non-toxic when in gel (solid) form. In the liquid state however, it is very poisonous
and carcinogenic. Given that it was unknown then as to the response o f PAG at near vaporizing
temperatures for extended periods of time, care was taken to ensure it did not revert into its potentially
harmful liquid state.
50
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
number of plates was chosen such that when combined with the phantom in the mold, the
distance of the middle phantom plane (antenna plane) to the lens of the camera would be
equal to the distance used for calibration with the copper pin array.
Once accurate registration of all components (camera, phantom, mold, seeds6)
was achieved, a background image of the phantom surface, in this case the antenna plane
was obtained. In order to ensure the pixel size previously determined with the copper pin
array was maintained, the distance from the antenna plane to the camera lens was
adjusted, using the glass plates, to equal the pre-established height h (the distance from
the top of the copper pin array to the lens of the IR camera). The upper phantom layers
were placed atop of the antenna and manually compressed together using another
Styrofoam block. Following each heating (following power shut-off), the phantom layers
above the plane to be imaged, in this case the antenna plane, were quickly removed. The
antenna was then removed and the desired phantom plane imaged.
After sufficient
cooling of the phantom layer, the process was repeated for the next layer, 2.5mm above
the phantom plane. In order to ensure that the distance of this phantom surface (2.5mm
above the antenna plane) to the camera lens was equal to h, one of the 2.5mm thick glass
plates was removed from beneath the bottom phantom layers. The entire process was
repeated until a phantom plane was reached at which no heating was measurable.
Generally, for SAR measurements, a depth of i.25-i.5cm away from the antenna plane
was imaged, while depths of about 3-3.5cm were imaged for extended heating
experiments. These experiments thus provided the ability to generate SAR and extended
heating temperature distributions via a reconstruction method detailed in section 2.2.1.
Accurate registration of all components (Phantom layers, antenna, camera, seeds) was
critical in order to reconstruct one, complete 3-D SAR or temperature distribution.
However, registration of all components between each experiment was very difficult to
achieve due to the need to relocate various components. Therefore, the absolute position
of the antenna and seeds may vary from one experiment to the next, but the relative
positions are maintained.
6 The Brachytherapy seeds were located throughout the phantom volume consistent with the pre-established
configurations, Case-A, B and C.
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
2.3.2
Luxtron Temperature Measurements
Thermometry probes, model 3100 Luxtron fluoroptic temperature probes, were
used to measure temperature at specific locations during the course of an extended
heating experiment. The results obtained with this measurement device were compared
to those measured thermographically. Luxtron probes were placed at locations relative to
the antenna and the brachytherapy seeds (dependent on the seed configuration) measured
using a standard ruler with an accuracy of approximately 0.5mm. The probe locations for
each scenario are shown in figure 2.9.
The Luxtron data acquisition software was configured to acquire temperature data
(in degrees Celsius) every second. The calibration temperature established according to
the protocol described in section 2.24 was used, to which all probes were then referenced.
Subsequently, the BSD was set to power the antenna with 8W for 5 minutes. Following
heating, the temperature data written to a .pm file was then imported into EXCEL and
plotted as a function of time. As with the IR measurements, 4 experiments were repeated
for each seed configuration in order to generate statistical confidence in the
measurements.
2.3.3
Electromagnetic Finite Element Simulations
The edge-based finite element model for electric field determination was
employed to calculate the SAR pattern resulting from a plane wave incident on a single
brachytherapy seed and lines of brachytherapy seeds. The primary objective was to
determine the electromagnetic scattering nature of the seeds at a frequency of 915MHz.
First, a solid model, representative of the desired bounding region and scattering objects
was created in Rhinoceros.
A brachytherapy seed (or two seed lines configured
according to the Case-A arrangement) modeled as a solid cylinder (with clinical
dimensions), was located at the center of a 2cm cubic volume representative of the
phantom material.
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
The boundary of the phantom volume was modeled as an absorbing boundary in order to
reduce the scattering effects of the boundary. This first order boundary approximation is
representative of the Sommerfeld radiation condition, which specifies that the scattered
field is a spherical outgoing wave (away from the scatterer) absorbing the scattered field
incident on it. All metals were assumed to be perfect conductors.
Subsequent to the generation of the solid model, a Parameters file7 was edited to
include the corresponding electrical properties and conditions, characteristic of the
materials and surfaces in the solid model.
Due to the nature of the model, only
specification of the electrical conductivity and permittivity of the phantom material8 was
required. Other input parameters included the frequency, propagation and polarization
directions.
In the single seed and two seed lines simulations, two propagation and
polarization direction combinations were tested; (i) propagation perpendicular to the seed
axis and polarized parallel to the axis, and (ii) propagation parallel to the seed axis and
polarized perpendicular to the axis. In both propagation scenarios, the magnitude of the
incident field was set arbitrarily to IV/m at the center of the bounding region. The size of
the bounding region selected by the user then dictated the magnitude of the incident field
at the boundary. The parameters Hie was also edited to include the mesh element sizes.
Specifically, a reference element length o f 0.25mm was selected, to which all elements in
the model were referenced. The element lengths of the domain and seed, relative to the
reference length, were selected to be 4 and 1 respectively. These were the smallest
achievable values suitable for a practical computation time. After implementing all of the
necessary model parameters of the Parameters Hie, it was then saved and compiled in
C++. ICEM CFD, a commercial mesh generator, was used to mesh the solid model with
the element sizes dictated by the Parameters file.
7 The Parameters file is a type of control file whereby all necessary simulation parameters, conditions and
parameters are inputted, from which subsequent processes in the model invoke as needed.
The PAG phantom material had a relative permittivity o f 55.5 and an electrical conductivity o f 1.39S/m.
(Suroweic era/, 1992)
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
The edge-based electromagnetic solver was then run.
In this step, the Helmholtz
differential wave equation was applied to each tetrahedron for which the unknown
electric field within each element was approximated by a sum of vector-valued basis
functions with unknown coefficients. The system of equations or coefficient matrix was
solved with an iterative method known as the quasi-minimal residual method (QMR).
The solution was then imported into MATLAB for analysis. MATLAB scripts were
created with which user-specified slices were plotted, so that the results could be viewed
on any axial, coronal or sagittal plane. (Kumaradas, 2002)
2.3.4
Thermal Finite Element Simulations
A node-based finite element model was employed to calculate the 3-D
temperature distribution following a S-minute period of heating. The solver used for this
simulation was Kaskade3.1, a commercial package that solves the differential transient
heat conduction equation with an experimentally determined SAR pattern as the input
power pattern. This thermal model was applied to the control case (no-seed) and the seed
configuration A scenario. Similar to the electromagnetic simulations, a solid model was
first created in Rhinoceros. In the no-seed case, a phantom volume was drawn, while for
the seed case, the same volume was created with brachytherapy seeds configured
accordingly. The phantom volume was chosen to be sufficiently large to ensure that the
Dirichlet thermal boundary condition imposed at the phantom edges was a reasonable
approximation.
The dimensions of the phantom volume were I6cm along the axis
parallel to the antenna axis, and I4cm in each of the other directions9. In the Parameters
file, the temperature of the Dirichlet boundary was set to 0, such that the final solution
would represent temperature rise.
9 The dimension parallel to the axis of the antenna was chosen to be larger than the other two dimensions,
as, the SAR pattern characteristic of the antenna was largest in this dimension. Hence, given the effects of
conduction over time, the extent of this dimension would be further increased. Thus, this dimension was
selected as the dimension, which dictated the maximum possible extent of the resulting temperature
distribution following S minutes of heating.
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
As the power input to the system was the experimentally determined SAR distribution10,
it was necessary to accurately position these seeds in the geometric model relative to the
input power source. There were two reasons for this: (i) the SAR was measured with the
seeds in this configuration, hence its resulting pattern was dictated by this particular seed
arrangement; (ii) the expected thermal impact of the seeds on the resulting temperature
distribution was dependent on the proximity of the seeds to the source. Therefore, to
accurately simulate the experiments, it was essential to maintain the same relative
positions and orientations of seeds and antenna. Other important considerations included
specification of the thermal properties of the titanium seeds and the phantom-equivalent
bounding domain11. These properties as well as the element sizes were specified in the
parameters file. The element lengths of the seeds and the phantom-equivalent domain
were specified to be 1 and 4 respectively, relative to a model reference element size of
0.5mm.
These element sizes were the smallest achievable values consistent with a
practical computation time. As in the electromagnetic finite element simulations, the
output solution was specified as a MATLAB file format for subsequent data processing,
manipulation and 3-D display using MATLAB scripts previously developed.
Once the model was meshed using the ICEM mesh generator software in
accordance with the element sizes specified in the parameters file, a command file
required of the Kaskade thermal solver was created. This command Hie defined problem
specifics not included in parameters file, including the type of thermal problem, the
dimensions in space, the heating time and the global precision; that is, the conversion
criterion. Specifically, a transient heat conduction problem was specified with a heating
time of 5 minutes. Next, the experimentally determined 3-D SAR data set was prepared
for input into the model under the pre-defined problem parameters and conditions. This
step involved using MATLAB to create additional data points in the SAR data set
10 The 3-D SAR pattern measured in the no-seed case and the experimental SAR data of the seed Case-A
scenario were used in the corresponding, no-seed and seed thermal models respectively.
11 The PAG phantom had a thermal conductivity of 0.535 W/mK, a specific heat o f 3810 J/kgK and a
density o f 1070 kg/m3. Titanium had a thermal conductivity of 22 W/mK, a specific heat of 523 J/kgK and
a density o f 4510kg/m3. (Suroweic, 1992, and Handbook of Chemistry and Physics, 2001)
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2. METHODOLOGY
through bilinear interpolation, as the measured data set was shown to be too sparse for the
solver. Once inputted, the thermal solver was executed and a solution was generated.
The results were then imported into MATLAB for analysis. The following flow chart
describes the steps required for the execution of the thermal model.
PARAMETERS FILE
3-D
Solid Geometric
Modeling
Mesh Generation
ICEMCFD
RhinocerosSD
3-D Transient Heat
Conduction Solver
Kaslcade Command File
Kaskade 3.1
<=□
Experimentally
Determined
SAR
Distribution
Analysis
MATLAB
Figure 2.13. Flow Chart: Finite Element Thermal Simulation Steps
56
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 3
RESULTS
3.1
Introduction: SAR Experiments and Plane Wave Simulations
The efficacy of TIMT depends on the localization of thermal dose administered to the
target volume. The helical antenna is an applicator well suited for this thermal therapy
treatment because it ensures localization of heating to the region surrounding the coil
element. The use of TIMT as a salvage treatment for failed Brachytherapy generated
concern as to the ability of the electrically conductive brachytherapy seeds to scatter the
incident microwave energy, thereby compromising the size and shape of the heating
pattern characteristic of the helical antenna. In order to quantify the scattering effects of
the seeds, 2-D and 3-D SAR contour patterns were measured in phantoms with and
without brachytherapy seeds using IR thermography. These SAR results, which were
measured under a variety of brachytherapy seed configurations, indicated that the
titanium seeds had a negligible effect on the antenna’s SAR pattern. Finite element
simulations involving a plane wave incident on a brachytherapy seed were subsequently
run so as to provide a better understanding of the electromagnetic interactions of the
metal seeds in a microwave field and the reasons for which the expected seed effects
were undetectable experimentally.
57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3.1.1
Two-Dimensional SAR Measurements
Following a 10W, 5-second power input to the helical antenna embedded in a
muscle-equivalent tissue phantom, SAR contour patterns were measured on the phantom
surface using an Infrared camera.
Figures 3.1 (a) and (b) show example 2-D SAR
contour patterns measured on the antenna plane of phantoms with and without seeds. In
this particular set of experiments, the seeds were configured according to the Case-A
protocol described in Chapter 2, section 2.2.5. The contours indicate the magnitude of
the power deposited, relative to the maximum power deposited, as well as the distribution
of power. A visual assessment of the shape and size of the contour patterns of figures 3.2
and 3.3 showed only minor differences between the seed and no-seed case.
Quantification of the widths and lengths of contours corresponding to 70%, 50% and
30% of the maximum power deposited also indicated statistically insignificant
differences between the no-seed (control) and seed (Case-A) scenarios. These contour
characteristics are identified in the schematic of figure 3.1.
D: 70%
D: 30%
L: 70%
D: 50%
L:50%
L: 30%
Figure 3.1: Schematic: Contour Characteristics. This figure identifies the 70%, 50% and 30% SAR
contour lengths and widths.
58
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
Two Seed Lines (Case-A) SAR Contour Pattern:
Antenna Plane 10W, 5tec
No-Seed SAR Contour Pattern: Antenna Plane 10W.5MC
45
40
LEGEND
35
30
3.0%
50%
10%
MM
70*
2S
40*
20
40*
IS
70%
90%
90%
Max. SAR • 10.3428 W/cm*
max. SAR • 10.1216 W/cm’
10
5
10
20
30
40
Diatanca (mm)
50
60
70
10
(a)
20
30
40
Dixunce (mm)
50
60
(b)
Figure 3.2: 915MHz Helical Antenna 2-D SAR Contour Pattern Antenna Plane: The SAR contour
pattern was measured on the antenna plane of a phantom without brachytherapy seeds-Fig. 3.2 (a), and with
brachytherapy seeds-Fig. 3.2 (b). A 10W, Ssec power input was supplied to the helical antenna. Each
contour represents the magnitude and extent of the power deposited, relative to the maximum SAR. The
90% contour is indicated in red and the 20% contour is indicated in dark purple.
The results of figure 3.3 (a) indicate only statistically insignificant differences
between the measured widths of the 70%, 50% and 30% contours. For both cases, as
distance from the source increased, a comparable fall off in SAR was demonstrated. This
suggests that the incident wave emitted from the helical antenna was propagated and
attenuated in the phantom in the same manner as the no-seed case, in spite o f the
inclusion of brachytherapy seeds at 0.5cm distances from the antenna axis. The results
obtained for the 70%, 50% and 30% contour length measurements o f the control and seed
scenarios of figure 3.3 (b) also demonstrated statistically insignificant differences
between the no-seed and seed cases, where multiple measurements were made for
statistical accuracy.
59
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
CHAPTER 3. RESULTS
2-0 SAR Contour Width Characterization: NoSaad Va. Saod, Antenna Plano
2-D SAR Contour Langth Characterization:
No-Seed Vs. Send, Antenna Plane
■No-Seed
I Seed
(Ca»e-A)
70%
50%
■Seed
(Case-A)
70%
30%
50%
30%
% Contour
% Contour
(a)
(b)
Figure 3 J : 2-D SAR Contour Characterization Bar Graphs. Multiple 2-D SAR contour patterns
obtained from the antenna plane were used to measure the contour widths o f the 70%, 50% and 30%
contours respectively, as shown in Fig. 3.3 (a) and the contour lengths, as shown in Fig. 3.3 (b).
3.1.2
Three-Dimensional SAR Measurements
An Infrared camera was used to measure SAR contour patterns in the plane of the
antenna (center plane), as well as multiple planes above it. The coronal images were
acquired at 2.5mm intervals, which corresponded to the minimum phantom layer slice
thickness. Power was supplied at 10W for 5 seconds to the helical antenna in a phantom
with no-seeds (i.e.: control), as well as phantoms with seeds configured according to the
following case scenarios: Case-A, Case-B and Case-C. Figures 3.4, 3.6 and 3.8 show the
2-D SAR contour patterns measured on the antenna plane, 2.5mm above the antenna
plane and 5mm above the antenna plane respectively, for each of the four experimental
scenarios.
60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
A qualitative comparison revealed the SAR contour patterns of the control and
seed case scenarios A-C (depicted in figures 3.6-3.9) to be indistinguishable. The relative
magnitudes, sizes and shapes of the SAR patterns obtained from the antenna plane of the
control and seed case scenarios were all similar. The magnitude of the maximum power
deposited varied by less than 0.5W/cm3 between each case.
Quantitative analysis
confirmed this, as multiple measurements of the 70%, 50% and 30% contour widths of
the no-seed and seed cases, demonstrated statistically insignificant differences (see Fig.
3.5 (a)).
Two Seed Linee (Caae-A) SAR Contour Pattern:
Antenna Plane 10W. Seec
No-Sood SAR Contour Pattam: Antenna Plane 10W, Ssec
LEGEND
90%
90%
Mu. SAR • 10.3426W/cm1
nun. SAR - 10.1216 W/cm’
10
20
30
40
□Mane* (mm)
SO
so
10
70
20
30
40
□Wince (mm)
SO
50
(a)
(b)
Two Seed Linee (Caea-B) SAR Contour Pattern:
Antenna Plane 10W, Seec
Mulipte SmcJ Lin m SAB Contour Pattern:
Antenna Ptene 10W, Soac
5Q%
70
50%
30%
¥»
90%
90%
MU.SAR-06S61 W/cm1
w/cm
70%
10
20
30
40
SO
□Wince (mm)
so
70
10
(C)
JO
40
OManca(mm)
SO
•0
(d)
Figure 3.4:915MHz Helical Antenna SAR Contour Pattern. The SAR contour pattern was measured on
the antenna plane o f phantoms with and without brachytherapy seeds. The no-seed SAR pattern is shown
in Fig. 3.4 (a). Seed case-A is shown in Fig. 3.4 (b). Seed case-B is shown in Fig. 3.4 (c) and Seed case-C
is shown in Fig. 3.4 (d). Each contour represents the magnitude and extent of the power deposited, relative
to the maximum SAR. The 90% contour is indicated in red and the 20% contour is indicated in dark
purple.
61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3-0 SAR Contour Characterization: Contour Widths
Antsnna Plans
■No-Seed
■Seed (Case-A)
□Seed (Case-B)
□Seed (Case-C)
70%
50%
30%
% Contour
(a)
3-D SAR Contour Characterization: Contour Lengths Antenna
Plane
■No-Seed
■Seed (Case-A)
□Seed (Case-B)
□Seed (Case-C)
70%
50%
30%
% Contour
(b)
Figure 3.5. SAR Contour Characterization B ar Graphs: Antenna Plane. Multiple 2-D SAR contour
patterns obtained from the antenna plane were used to measure the contour widths o f the 70%, 50% and
30% contours respectively, as shown in Fig. 3.5 (a) and the contour lengths, as shown in Fig. 3.5 (b). The
no-seed and seed cases A-C are represented, as indicated in the legend of the bar graph. Error bars denote
the standard deviation from the mean.
62
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
Similarly, figure 3.5 (b), which shows the lengths of the 50% and 30% contours
also showed only insignificant differences. The 70% contour length measurements on
the other hand did demonstrate some differences. For the seeds configured according to
case-B and C, the 70% contour lengths were statistically smaller than those measured for
the no-seed and seed (Case-A) cases. The latter two cases however, had much larger
standard deviations. The ‘larger’ 70% contour lengths are indicative of 70% of the
maximum power deposited occurring in the upper and lower half of the figure eight SAR
pattern. The ‘smaller’ 70% contour lengths with sizes ranging from about 7.4mm-9mm
resulted because 70% of the maximum power deposited was limited to the upper half of
the figure eight SAR pattern. The large standard deviations associated with the no-seed
and seed (Case-A) cases are indicative that for some measurements, 70% of the
maximum power was deposited in both the upper and lower half of the figure eight, while
for other measurements-70% of the maximum power was only observed in the upper half.
It is difficult to associate a seed-induced trend in this contour parameter then, given the
large variability associated with the no-seed case and seed Case-A.
As in the results obtained from the antenna plane, the SAR contour patterns of
figure 3.6 (a)-(d) were difficult to differentiate visually. Even 2.5mm above the antenna
plane, there was little difference in the maximum power deposited as well as the size and
shape of the heating patterns for all four cases. Inspection of the 70%, 50% and 30%
contour widths at this depth-wise distance from the antenna again revealed statistically
insignificant differences (see Fig. 3.7). The statistics associated with the lengths of the
70%, 50% and 30% contours also disclosed nominal differences 2.5mm above the
antenna plane between each scenario. A relatively large amount of variability was again
evident in the 70% contour length measurements of the no-seed and seed (Case-A) cases.
Similar to the results measured on the antenna plane, seed Case-C demonstrated a
decrease in the area of 70% contour 2.5mm above the antenna plane as well.
63
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
Two Seed Line* (Case-A) SAR Contour Pattern:
2.5mm Above Antenna Plane 10W, 5aec
No-Soed SAR Contour Pattern:
2.5mm Above Antenna Plane 10W, Saec
4540
35-
LEGEND
_ _
(
50%
4 -^ 2
-Al/~--
80*
"e 30 •
E
s '25
70*
i ~
j
w
3.0%
C' t o o ° ^
n '
E 30
60*
| 20
IS
10
*
9055
50*
—
40*
30%
30*
Mil. SAR • 7.5136 W/cm*
70%
V 90%
70% ' —'J
90%
SAR - S.537S W/cm
20*
5
_.
10
20
Q_
30
40
0Manc*(mm)
50
70
20
10
30
40
Distance (mm)
SO
60
(a)
(b)
Two Seed Lines (Case-B) SAR Contour Pattern:
2.5mm Above Antenna Plane 10W, 5sac
Mulipie Seed Line* SAR Contour Pattern:
2.5mm Above Antenna Plane 10W, Saec
70
40-
30%
3S -
50%
30%
30-
IT30
r\ f
2520-
90%
Max. SAR-6.1743 W/cm'
15-
70%
Max. SAR - 7 7002
10-
S10
20
30
40
Distance (mm)
SO
60
70
10
(C)
20
30
40
Distance (mm)
SO
60
(d)
Figure 3.6: 915MHz Helical Antenna SAR Contour Pattern-2.5 mm above Antenna Plane. The SAR
contour pattern was measured 2.5mm above the antenna plane of phantoms with and without brachytherapy
seeds. Power was supplied at 10W for 5s to the antenna. The no-seed SAR pattern is shown in Fig. 3.6 (a).
The SAR pattern of Seed case-A is shown in Fig. 3.6 (b). The SAR pattern o f Seed case-B is shown in Fig.
3.6 (c) and the SAR pattern of Seed case-C is shown in Fig. 3.6 (d). Each contour represents the magnitude
and extent of the power deposited, relative to the maximum SAR. The 90% contour is indicated in red and
the 20% contour is indicated in dark purple.
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3-D SAR Contour Characterization: Contour Widths
2.5mm Above Antenna plana
■No-Seed
■Seed (Case-A)
□Seed (Case-B)
□Seed (Case-C)
E
E,
•C
S
3
70%
50%
% Contour
30%
(a)
3-0 SAR Contour Characterization: Contour Lengths
2.5mm Above Antenna Plane
■No-Seed
■Seed (Case-A)
□Seed (Case-B)
□Seed (Case-C)
70%
50%
% Contour
30%
(b)
Figure 3.7. SAR Contour Characterization Bar Graphs: 2.5mm above Antenna Plane. Multiple 2-D
SAR contour patterns obtained 2.5mm above the antenna plane were used to measure the contour widths of
the 70%, 50% and 30% contours respectively, as shown in Fig. 3.7 (a) and the contour lengths, as shown in
Fig. 3.7 (b). The no-seed and seed cases A-C are represented, as indicated in the legend o f the bar graph.
Error bars denote the standard deviation from the mean.
65
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
As shown in figure 3.8, the SAR contour patterns measured for the control and
seed case scenarios 5mm above the antenna plane generally retained the characteristic
figure eight pattern, but were irregular in shape, as a result of the decreased signal to
noise ratio at this distance from the antenna. There was no evidence of significant
differences in the SAR contour patterns between any o f the cases. In addition, the width
and lengths of the contours measured 5mm above the antenna plane were similar for all
four cases (see figure 3.9 (a)-(d)). However, the 30% contour width of seed case-C was
smaller than that of the control case.
While this difference was not statistically
significant, this result was observed on each plane.
Three-dimensional SARs were reconstructed for the no-seed and seed cases A-C.
Examination of the SAR distributions on coronal, saggittal and axial planes showed
negligible differences in the relative sizes and shapes, between each case. As the 3-D
SAR distribution of each case was similar, only the no-seed SAR distribution is shown in
figures 3.10 (a) and (b). In both figures, the 3-D cut was taken through the location
corresponding to the maximum SAR.
66
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
No-Saed SAR Contour Pattern:
5mm Above Antenna Plane 10W. 5eec
45
J .—i ‘
LEGEND
----- 90%
35
------ 80%
E30 a
70%
§ 2* ------ 60%
------ 50%
520
IS ------ 40%
10 — 30%
20%
Two Seed Uhee (Caae-A) SAR Contour Pattern:
5mm Above Antenna Plane 10W,5eec
.
^50%
40 D
45
3(j%
4035
5
5
' J
-
—
|3 0 -
i
'70%
90%
20
30
40
DliUnca (mm)
90%
2
15-
Max. SAR - 3.1101 W/cm*
IQ­
S'
.
10
Q
S*25•
i» -
50
70
0 —O
Max. SAR-1.7850 W/cm*
20
10
30
40
□Mane* (mm)
(a)
50
60
70
(b)
Two Seed Lines (Case-8) SAR Contour Pattern:
5mm Above Antenna Plane 10W, 5sec
Multiple Seed Linee SAR Contour Pattern:
5mm Above Antenna Plane 10W, 5sec
30%
6 30
:r.
90%
Max. SAR • 1.5366 W/cm1
70%
10
20
30
40
Distance (mm)
50
60
70
10
(C)
20
30
40
Distance (mm)
so
60
(d)
Figure 3.8: 915MHz Helical Antenna SAR Contour Pattern-5mm above Antenna Plane. The SAR
contour pattern was measured 5mm above the antenna plane o f phantoms with and without brachytherapy
seeds. Power was supplied at 10W for 5s to the antenna. The no-seed SAR pattern is shown in Fig. 3.8 (a).
The SAR pattern o f Seed case-A is shown in Fig. 3.8 (b). The SAR pattern o f Seed case-B is shown in Fig.
3.8 (c) and the SAR pattern of Seed case-C is shown in Fig. 3.8 (d). Each contour represents the magnitude
and extent of the power deposited, relative to the maximum SAR. The 90% contour is indicated in red and
the 20% contour is indicated in dark purple.
67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3-0 SAR Contour Characterization: Contour Widtha
5mm Above Antenna Plana
■No-Seed
■Seed (Case-A)
□Seed (Case-B)
□Seed (Case-C)
70%
50%
30%
% Contour
(a)
3-D SAR Contour Characterization: Contour Lengths
5mm Above Antenna Plane
■No-Seed
■Seed (Case-A)
□Seed (Case-B)
□Seed (Case-C)
70%
50%
30%
% Contour
(b)
Figure 3.9. SAR Contour Characterization Bar Graphs: 5mm above Antenna Plane. Multiple 2-D
SAR contour patterns obtained 5 mm above the antenna plane were used to measure the contour widths of
the 70%, 50% and 30% contours respectively, as shown in Fig. 3.9 (a) and the contour lengths, as shown in
Fig. 3.9 (b). The no-seed and seed cases A-C are represented, as indicated in the legend o f the bar graph.
Error bars denote the standard deviation from the mean.
68
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3-D Reconstruction of Experimentally-Determined SAR:
Controj C ase, 10W /5sec
0 0
(a)
3-D Reconstruction of Experimentally Determined SAR:
Control C ase, 10W, / 5sec
mm
(b)
Figure 3.10. 3-D Reconstruction o f Experimentally Determined SAR for the No-Seed (Control) Case.
Figure 3.10 (a) depicts slices taken through the x, y and z planes and figure 3.10 (b) depicts slices taken
through the x and z planes, at the location o f maximum power deposition. The colour bar on the right hand
side of the figure identifies the percent magnitude o f the power relative to the maximum power deposited.
The x, y and z axes are expressed in units of millimeters.
69
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3.2
Finite Element Plane Wave Simulations
Plane wave models, using an edge-based finite element program, were simulated in the
presence of single seeds and seed lines to provide a general understanding of the
interactions of the microwave energy with the metal seeds in three dimensions, and the
reasons for which the suspected interactions were not detectable experimentally. The
titanium seed or seed lines were strategically located in the center of a finite domain
electrically equivalent to phantom muscle tissue.
At the center of this domain, the
amplitude of the plane wave was set to lV/m, which is equal to 0.69SW/m3 (for average
SAR = o/21E12) in this particular medium. Multiple scenarios were simulated where the
direction of polarization and propagation were varied relative to the long axis of the seed.
The most significant perturbations (‘worst’ case) in the deposition of electromagnetic
power were demonstrated when the plane wave was polarized in the direction of the seed
axis and was propagated perpendicular to the axis of the seed. Simulations involving
other polarization and propagation directions relative to the seed orientation revealed
small perturbations to the resulting deposition of electromagnetic power. Consequently,
the figures of only the ‘worst-case’ simulations are plotted. The results of figures 3.11
(a)-(c) show the deposition of electromagnetic power (W/m3) for the single seed case, and
figures 3.12 (a) and (b) depict the two seed-lines simulations results.
In figures 3.11 (a), (b) and (c), a single brachytherapy seed was oriented with its
longitudinal axis parallel to the direction of polarization of the incident plane wave, and
perpendicular to the wave’s propagation direction. Figure 3.11 (a) outlines the edges of
the seed in black. At the edges of the seed, ’hot spots’ were evident. The maximum
magnitude of these hot spots was equal to 9 W/m3, which is approximately 13 times the
magnitude of the power deposited in that same location. On this 2-D slice, the area of the
‘hot spot’ was l-2mm2 for 100% to 20% of the maximum power. Encircling the middle
of the seed, there was evidence of a ‘cold spot’; that is, a region where the power
deposited was less than it would have been in that same region without a seed. Figure
3.11 (c) isolates this cold spot as the slice is taken through the very center o f the seed.
70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
CASE I: Single Seed
W/m3
Hot Spots:
9W/m
Background:
- 0.7W/mJ
-0 .0 1
- 0 .0 0 0 - 0 0 0 A - 0 0 0 4 - 0 0 0 2
0
0 .0 0 2
0 004
0 .0 0 0
0 .0 0 0
Background:
- 0.7W/m3
0 .0 1
(a)
(b)
Slice Location relative to seed axis.
(C)
Figure 3.11. Single Seed Plane Wave Simulation-Worst Case. In this scenario, the plane wave was
polarized in the direction of the seed axis and propagated in the direction perpendicular to the seed axis.
Each figure is a two-dimensional slice extracted at a known location from the three-dimensional power
deposition volume. A slice through the center o f the seed -parallel to the seed axis and the direction of
wave propagation is shown in fig. 3.11 (a). A slice just past the end of the seed-perpendicular to the seed
axis and parallel to the direction of wave propagation is shown in (b). A slice through the middle of the
seed-perpendicular to the seed axis and parallel to the direction of wave propagation is shown in (c). Both
(a) and (b) show the hot spots at the end o f the seeds, while figures (a) and (c) show the cold spots around
the seed middle. The colour bar on the right hand side of the figure indicates the magnitude of power as a
function of colour.. The results are plotted on an area 0.02m x 0.02m.
71
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
CASE II: Two Seed Lines
Figures 3.12 (a) and (b) show the plane wave simulations for a two seed line
configuration.
In order to simulate a seed density typically used in clinical
Brachytherapy applications, the seed lines were spaced 1cm apart. As in the ‘worst-case’
single seed results, the plane wave was polarized in the direction of the seed axis and
with the propagation direction perpendicular to this axis. There was evidence of ‘hot
spots’ at the seed edges and ‘cold spots’ encircling the center of the seeds. The intensity
of the hot spots were more pronounced at the outer ends of the seed lines as compared to
the hot spots at the seed edges contained within the seed line. The latter result may be
attributed to destructive interference occurring between the ends of adjacent seeds.
•
0.01
m
00
Figure 3.12. Two Seed Lines-Plane Wave Simulation. In this scenario, the plane wave was polarized in
the direction of the seed axis and propagated in the direction perpendicular to the seed axis. Each figure is
a two-dimensional slice extracted at a known location from the three-dimensional power deposition
volume. A slice through the center of the seeds-parallel to the seed axis and perpendicular to the wave
propagation direction is shown in fig. 3.12 (a). A slice through the center of the sMds-parallel to the seed
axis and the direction o f wave propagation direction is shown in fig. 3.12 (b). Both (a) and (b) show the
hot spots at the end of the seeds and the cold spots around the seeds’ middle. The colour bar on the right
hand side of the figure indicates the magnitude of power as a function o f colour. The results are plotted on
an area 0.02m x 0.02m.
72
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
33
Extended Heating Experiments
In addition to the electromagnetic scattering effects, we hypothesized that the
brachytherapy seeds may have an effect on the temperature distribution due to their high
thermal conductivity and low heat capacity, in comparison to the phantom material.
Therefore, experiments involving extended heating were performed to determine the
impact of the largely different thermophysical properties of the seeds on the resulting
temperature distributions.
Following a S minute 8W power input to the antenna,
temperature distributions were measured in phantoms with and without brachytherapy
seeds.
Multiple two and three-dimensional temperature distribution images were
measured using the infrared camera, from which temperature profiles were determined.
IR thermography was employed for the purpose of comparing relative temperature
distributions of the no-seed and three seed cases. As there was no evidence of seedinduced changes to the relative temperature distributions using IR thermography, for
comparison, Luxtron temperature probes were employed to measure temperature rises at
specific locations, under the same experimental conditions.
The thermometry
measurements also indicated no differences between the no-seed and seed cases.
Therefore, preliminary finite element thermal modeling was investigated as a means to
quantify the expected thermal impact of the seeds on the resulting temperatures in tissueequivalent phantom material, not detectable using IR thermography or point
thermometry. In addition, the simulations helped to generate an understanding as to the
reasons for which there was no experimental evidence of seed effects.
73
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3.3.1
Extended Heating Temperature Distributions: Antenna Plane
The Infrared camera was used to measure spatial temperature information
following extended heating experiments. Power was supplied at 8W for 5 minutes to the
helical antenna in a phantom with no-seeds (i.e.: control), as well as phantoms with seeds
configured according to the following case scenarios: Case-A, Case-B and Case-C, from
which coronal images of temperature distributions on the antenna plane were measured.
Figures 3.14 (a)-(d) show examples of the temperature distributions measured
thermographically for all four cases. From a qualitative perspective, there were few
discernible differences between each case. Specifically, the width and length of the area
corresponding to a temperature rise greater than 1S°C, was consistently 1.0-1.2cm and
3.5cm respectively, in all four cases. Thus there was no quantitative evidence of the
increased thermal conductivity of the seeds increasing the width of the temperature
distribution. In addition, the width and length of the area of maximum temperature rise
was also measured. For these measurements there was a significant amount of variability
between each case. The lengths of the areas of maximum temperature rise ranged in size
from 0.5cm to 1.5cm, while the widths ranged from 0.3cm to 1.0cm. This variability was
investigated using X and Y-axis temperature profile measurements (see figure 3.13). The
widths of the profiles were measured at locations corresponding to 90%, 70%, 50% and
30% of the maximum temperature rise in order to quantify and compare the relative
extent of the temperature distributions of the no-seed and seed cases. Examples of the
temperature profiles measured for each case are shown in figure 3.15 (a) and (b), and the
profile width statistics are indicated in figure 3.16 (a) and (b).
Figure 3.13.
Schematic: Orientation and
Location of Y and X-axis Temperature Profiles.
The temperature distribution plots were used to
determine the temperature profiles. Both the x and
y-axis temperature profiles were obtained through
the antenna hot spot. The normalized widths
corresponding to 90%, 70%, 50% and 30% of the
maximum temperature rise were measured from
these nmfiles.
74
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
Figure 3.14. Extended Heating IR Temperature Distribution Plot: Antenna Plane. 8W was supplied
to antenna for S minutes in phantoms with and without brachytherapy seeds. Following heating, the
antenna was removed and the phantom surface corresponding to the antenna plane was exposed for
imaging. The resulting temperature distribution of the no-seed case is shown in Fig. 3.14 (a) and the
maximum temperature rise measured for this case was SS°C. The resulting temperature distribution o f seed
case-A is shown in Fig. 3.14 (b) and the maximum temperature rise measured for this case was 55.5°C.
The resulting temperature distribution of seed case-B is shown in Fig. 3.14 (c) and the maximum
temperature rise measured for this case was 53°C. The resulting temperature distribution of seed case-C is
shown in Fig. 3.14 (d) and the maximum temperature rise measured for this case was S6° C. The area of
maximum temperature rise in indicated in deep red and the background, representative of zero temperature
rise, is indicated in dark blue.
75
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
Nanaoiiad X-Axis Immpmrtmn Pro*#*: KjumkM HM hg WT, t a i l
Sesc(Casft-Ai
NoSeod
100
100 -
(a)
ao>
80!
40;
20i
Monaoiiad
TMHrmn
20
Mm
40
00
Sw d ICise-Bj
S m O (C 4 M -C )
100 r
100
80!
601i
00
40;
40
20!
20
60
60
20
0.
80
20
40
00
DbtMM (aiai)
N o m d a d V-Axi* TMomrmre
(b)
KJOMdMl H u d a g W , 8 M b
S*?ntl (Casa-*)
No-Stcd
1001-
oojI
60;
40|
20 !
NorauNad
Tm m n w *
20
40
60
20
00
40
60
00
60
00
S«ad(C4M-C)
2«»(J iC.iv)-81
100
00
60
40
20
0.
20
40
ObiMM (faai)
Figure 3.15. Normalized X and Y-axis Temperature Profiles. These X-axis profiles, shown in (a), were
measured through the axis o f the antenna. The Y-axis profiles, shown in (b), were measured through the
antenna hot spot perpendicular to the antenna axis. The horizontal axis is the distance in millimeters. The
vertical axis is the normalized temperature rise.
76
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
The X-axis temperature profile was obtained directly through the antenna axis
while the Y-axis temperature profile was obtained through the antenna hot spot,
orthogonal to the axis of the antenna. The temperature rise for each case was normalized
to the maximum. There was no significant difference in the widths of the temperature
profiles of each of the case scenarios. For each case, the X-axis temperature profile had
two humps, and the Y-axis temperature profile had one sharp peak, as wouid be expected
given the locations of the hot spots in the SAR pattern characteristic of the helical
antenna. A number of temperature distribution plots were measured from which multiple
temperature profiles were determined.
Figures 3.16 (a) and (b) show the average
temperature profile widths corresponding to 90%, 70%, 50% and 30% of the maximum
temperature rise measured for each case.
The X-axis temperature profile measurements showed insignificant statistical
deviations between the control and seed case scenarios. In general, 90% contour widths
ranged in size from 8.5mm to 9.5mm. The maximum width variation was therefore equal
to approximately 1mm. The 70% and 50% width measurements had similar degrees of
variability, where the maximum width variation was still about 1mm. There was some
statistical variability however associated with the profile width measurements
corresponding to 30% of the maximum temperature rise for the various seed cases as
opposed to the no-seed case.
While there was also some variability inherent in the Y-axis temperature profile
width measurements, some trends were evident in seed Case-C. The 70%, 50% and 30%
width measurements for this case were consistently smaller than the other configurations.
For the other seed cases however, the differences in the profile widths were statistically
insignificant and no seed-induced trends were evident.
77
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
X-Axis Temperature Profile Width Characterization:
Extended Heating 8W / 5min
90%
70%
50%
Profile Width
30%
(a)
Y-Axis Temperature Profile Width Characterization
Extended Heating 8W / 5min
90%
70%
50%
Profile Width Level
30%
(b)
Figure 3.16. X and Y-axis Profile Width Characterization Bar Graph. Multiple 2-D temperature
distributions obtained from the antenna plane were used to measure the x-axis temperature profile widths
(shown in (a» and y-axis profile widths (shown in (b), corresponding to 90%, 70%, 50% and 30% of the
maximum temperature rise for the control and seed case scenarios (A-C). This figure characterizes the
average widths of the temperature profiles and the error bars denote the standard deviation from the mean.
78
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3.3.2
Three-Dimensional Extended Heating Measurements
Two-dimensional coronal temperature distributions were measured for the no­
seed and seed Case-A cases, and subsequently three-dimensional temperature
distributions were reconstructed. 8W was supplied to the helical antenna for Sminutes.
After heating, the phantom slice coincident with the antenna plane was exposed for
imaging. This procedure was repeated for multiple planes above the antenna plane until a
planar depth was reached at which no heating was evident. Figures 3.17 and 3.18 show
normalized 3-D temperature reconstructions for the no-seed and seed (Case-A) scenarios
respectively. Both the no-seed and Case-A results were normalized to their respective
maximum temperature rises. These 3-D temperature distributions did not indicate any
significant differences in the shape and size of the heating patterns on any plane. In
addition, there was no evidence of additional hot spots or regions of altered heating on
the antenna plane or throughout the heated volume.
79
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3 -0 Reconstruction of Experimentally Determined Extended Heating Temperature Distribution
No-Seetd&W, 5min
^ ^ h io o
6 0 s.
360
50
Normalized
Temperature
Rise
(c)
(b)
Figure 3.17. Normalized 3*0 Reconstruction of Experimentally Determined Extended Heating
Temperature Distribution: No*Seed. 3-D cuts in the temperature distribution both perpendicular to and
parallel to the axis of the antenna-through the hot spot, are shown in (b) and (c) respectively. The grid
delineations are representative o f millimeter units.
80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3 -0 Reconstruction of Experimentally Determined Extended Heating Temperature Distribution
Seed (Case.-A) 8W. 5min
1
Normalized
Temperature
Rise
(b)
(C)
Figure 3.18. Normalized 3-D Reconstruction of Experimentally Determined Extended Heating
Temperature Distribution: Seed Case-A. 3-D cuts in the temperature distribution both perpendicular to
and parallel to the axis o f the antenna-through the hot spot, are shown in (b) and (c) respectively. The grid
delineations are representative o f millimeter units.
81
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3.3.3
Extended Heating Luxtron Temperature Measurements
For comparison with thermographic measurements, Luxtron temperature probes
were used to measure temperatures at discrete locations during extended heating
experiments (the antenna was supplied with 8W for 5 minutes). Due to the increased
thermal conductivity and decreased heat capacity of the titanium seeds, one would have
expected increased temperatures at increasing distances from the source and at the seed
locations respectively. Therefore, to quantify the seed effects, temperature probes were
located beside the antenna as well as O.Scm or 1.0cm from the antenna (see figure 2.9 in
the Methodology chapter). Figure 3.19 shows temperature rise at the antenna hot spot as
a function of time, for each case. Figure 3.20 shows time-dependent temperature rise at a
distance 1cm from and parallel to the tip of the antenna, for each case. Figure 3.21 shows
the temperature rise at a distance O.Scm from and parallel to the tip of the antenna for
seed Case-A, where the seed is located 0.5cm from the antenna.
As seen in figure 3.19, at the antenna hot spot, no significant differences between
the control and seed cases A and B were observed. Lower temperatures were observed
for Case C, although this was not significant. There was no difference in the rates of
temperature rise for each case, including seed case-C. Similarly, there was no apparent
difference in the cooling rates of each case, following power shut off.
At a lateral distance of 1cm from the antenna tip, the temperature rise of the no­
seed, seed case-A and seed case-B probe I (Luxtron probe located 1cm from the antenna
and therefore right beside the seed line) scenarios were all within one standard deviation.
The average maximum temperature rise for the three scenarios at this probe location was
11°C.
In seed case-C however, the results again demonstrated somewhat lower
temperature rises. Interestingly, in seed case-B, when the probe was located 1cm from
the seed line positioned right beside the antenna (Probe II), the average maximum
temperature rise was about 12.5°C.
The large error bars associated with these
measurements however, overlapped with those error bars o f the no-seed case
82
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
and seed cases A and B (probe I)- It is also interesting to note that after about 100
seconds, the temperature lines of figure 3.20 were almost parallel to one another; that is,
their slopes were almost identical. This suggested negligible differences in the rates of
temperature rise between each case. The variability associated with the results of seed
Case-B probe II position may be attributed to the difficulty consistently locating the
temperature probe 1cm away from the antenna and seed line.
Average Temperature Rise at Antenna 'Hot Spot'
Luxtron Measurements: 8W, 5min
No-Seed Vs. Seed
No-Seed
Seed (Case-A)
Seed (Case-C)
Seed (Case-B)
100
200
300
400
Time (sec)
500
600
Figure 3.19. Average Temperature Rise at Antenna Hot Spot: Control Vs. Seed Case Scenarios.
These discrete temperature measurements were made with the use o f a Luxtron temperature probe. An 8W
power input was supplied to the antenna for Sminutes. Temperature data was acquired every second.
Multiple measurements (4) for each case were made from which the average and corresponding standard
deviation was calculated.
83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
Average Temperature Rise 1cm Lateral Diatance From
Antenna Hot Spot, 8W, 5min Luxtron Meaeurements
No-Seed Va. Seed
No-Seed
Seed (Case-A)
Seed (Case-C)
Seed (Case-B) I
Seed (Case-B) II
Time (sec)
Figure 3.20. Average Temperature Rise 1cm Away from and Parallel to Antenna Hot Spot: Control
Vs. Seed Case Scenarios. These discrete temperature measurements were made with the use of a Luxtron
temperature probe. An 8W power input was supplied to the antenna for Sminutes. Temperature data was
acquired every second. Multiple measurements (4) for each case were made from which the average and
corresponding standard deviation was calculated. For the seed (Case-B) I, the temperature probe was
located beside the top seed located icm from the antenna. For seed (Case-B) n, the temperature probe was
located icm from the antenna, and hence -icm from the seeds.
In the results of figure 3.21, there was minimal variation between the temperature
rises incurred 5mm from the antenna with and without the presence of seeds, when
configured according to the case-A protocol. At this location, the average maximum
temperature rise was about 27°C. In addition, there were negligible differences in the
rates of temperature rise and the cooling rates at this distance from the source.
84
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
Average Temperature Riee 5mm Lateral Distance from Antenna Hot
Spot 8W, 5min Luxtron Measurements
No-Seed Ve. Seed
No-Seed
Seed
0
100
200
300
Time (see)
400
500
600
Figure 3.21. Average Temperature Rise 0.5cm Away from and Parallel to Antenna Hot Spot: Control
Vs. Seed Case-A Scenario. These discrete temperature measurements were made with the use of a
Luxtron temperature probe. An 8W power input was supplied to the antenna for 5minutes. Temperature
data was acquired every second. Multiple measurements (4) for each case were made from which the
average and corresponding standard deviation was calculated.
Generally, in the presence of brachytherapy seeds, the Luxtron temperature
measurements did not indicate any significant differences in the magnitudes of
temperature rise nor in the rates of temperature rise or decline, for each case. While seed
case-C did tend to demonstrate lower temperature rises at the hot spot and i cm from the
antenna, these differences were still on the order of a few degrees only.
85
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
3.4
Finite Element Extended Heating Simulations
Extended heating simulations were employed to provide some understanding as to the
theoretical effects of the brachytherapy seeds when subject to a high temperature
environment for extended periods and in tum the reasons for which the suspected seed
effects were experimentally undetectable, and the subsequent clinical implications of
these results. Specifically, the finite element method was employed to solve a threedimensional heat conduction equation in a finite sized domain, with and without seeds in
the model. The seeds were modeled as solid cylinders and configured according to the
case-A protocol.
An experimentally determined three-dimensional SAR distribution
served as the input power source to the model. Due to a limitation of the thermal solver,
Kaskade, the model with seeds was prematurely terminated resulting in a heating time of
only 116seconds instead of the desired heating time of 300 sec, implemented
experimentally. Therefore, a no-seed model was run with a heating time of 116 seconds
thus enabling a relative comparison of the no-seed and seed temperature distributions.
Figures 3.22 and 3.23 are 2-D images taken from the no-seed and seed (Case-A)
theoretical simulations run for 116seconds, on the antenna plane and perpendicular to the
antenna plane, respectively.
From the temperature distributions, X and Y-axis
temperature profiles were used to quantify the differences in the no-seed and seed case.
These results are shown in figure 3.24.
Figure 3.22 shows the theoretical temperature distributions on a coronal plane, for
the no-seed and seed Case-A cases. As both images are plotted on the same colour bar
scale, it is evident that the model with seeds resulted in maximum temperature rises 5
degrees higher than in the no-seed case. In addition, there appears to be more spreading
of the thermal energy in the presence of the seeds and thus increased temperatures
throughout the heated volume.
The magnitude of the temperature rises at discrete
locations, specifically at locations coincident with the antenna hot spot, as well as Smm
and 10mm laterally from this hot spot are given in table 3.1. These results indicated
higher temperature rises at the hot spot and peripheral locations, in the phantom with the
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
brachytherapy seeds. Figure 3.23 identifies the locations at which these temperature rises
were obtained.
-0.01
0
0.01
0.02
m
(a)
0.03
0.04
0.05
-0.01
0
0.01
0.02
m
(b)
0.03
0.04
0.05
Figure 3.22. 2-D Antenna Plane Coronal Images: Theoretical Extended Heating Simulations 8W,
116sec. Figure (a) on the left hand side denotes the no-seed case, and figure (b) on the right side denotes
the seed (Case-A) scenario. Both images are depicted on the same colour bar scale. Each slice was taken
through the location of maximum temperature rise.
Location of Tem perature
Rise
Antenna Hot Spot (near tip
of antenna)
5mm Lateral Distance from
Antenna Hot Spot
10mm Lateral Distance
from Antenna Hot Spot
No-Seed Tem perature
Rise(°C)
51.75
Seed (Case-A)
T em perature Rise (°C)
56.64
24.75
31
5.35
14.01
Table 3.1. No-Seed Vs. Seed (Case-A) Temperature Rise Values. The temperature rises at locations
coincident with the antenna hot spot, as well as 5mm and 10mm from the hot spot were determined from
the extended heating simulation results in both the seed and no-seed case. Figure 3.21 identifies the
locations at which these temperature rises were determined.
87
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
105
0
I i Mhn111
105
.01
I 1i
’Mn m ) M ^ ()
115
.02
125
.03
-0.01 -0.005
0.005 0.01 0.015 0.02 0.025 0.03 0.035
0 0.005 0.01 0.015 0.02 0.025 0.03
m
m
(b)
(a)
Figure 3.23. 2-D Antenna Plane Axial Images: Theoretical Extended Heating Simulations 8W,
116sec. The no-seed case (a) and the seed Case-A (b) are depicted on the same colour bar scale. Each slice
was taken perpendicular to the antenna plane and axis through the location of maximum temperature rise.
To further quantify the differences, X and Y-axis temperature profiles were
obtained through the point of maximum temperature rise and the normalized profile
widths of each case compared. Figure 3.24 shows the temperature profiles for each case.
Tables 3.2 and 3.3 denote the X and Y-profile width measurements obtained from these
theoretical heat conduction simulations.
Specifically, normalized profile widths
corresponding to 90%, 70%, 50% and 30% of the maximum temperature rise were
measured from the antenna plane of the seed and no-seed cases and the results tabulated.
These results indicated larger profile widths in both the x and y directions for the seed
case as opposed to the no-seed case. The Y-profile widths of the seed case were 14% and
22% larger than the no-seed case at the 90% and 30% profile width locations
respectively. In the same way, the X-profile widths were 15.4% and 11.1% larger at the
90% and 30% profile width locations respectively for the seed case scenario. Although
direct comparisons of simulated and experimental extended heating results is not suitable
(as they were generated from different heating times), a relative comparison of the
88
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3. RESULTS
effects can be made.
The simulated results indicated that the seeds increased the
magnitude and extent o f the temperature distribution; experimentally however, there was
no evidence of any seed effects.
N o -S e e d
S ee d CaseN o -S e e d
too
-to .
mm
(a)
too
(b)
Figure 3.24. Normalized X and Y-Axis Temperature Profiles. The following temperature profiles
represent those profiles measured from the simulated results, for the control and seed (Case-A) scenario.
These Y-Axis profiles are shown in (a) and the X-Axis temperature profiles are shown in (b), both obtained
through the location of maximum temperature rise. The horizontal axis is representative o f distance in
millimeter units. The vertical axis identifies the normalized temperature rise.
Y-Profile Width Location
(% of Maximum)
90%
70%
50%
30%
No-Seed Profile Width
(mm)
3.1724
6.2069
8.9655
13.2414
Seed (Case-A) Profile
Width (mm)
3.6207
6.8966
10.3448
16.2045
X-Profile Width Location
(% of Maximum)
90%
70%
50%
30%
No-Seed Profile Width
(mm)
4.1270
15.4603
18.9206
22.8571
Seed (Case-A) Profile
Width (mm)
4.7619
16.0634
19.9365
25.3968
Tables 3.2 and 3 3 . Y-Proflle and X-Proflle Width Measurements: No-Seed Vs. Seed Case-A At
locations corresponding to 90%, 70%, 50% and 30% of the maximum temperature rise, the width o f the
profile was measured.
89
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 4
DISCUSSION
4.1
Introduction: SAR Experiments and Plane Wave Simulations
The issue which first motivated the investigation of thermal therapy as a salvage
treatment for failed Brachytherapy, was the potential for the titanium brachytherapy seeds
to significantly perturb the electromagnetic field created by the 915MHz interstitial
helical antennae.
Due to the contrast in electrical properties of tissue and titanium
brachytherapy seeds, intuitively, one might have expected the metal seeds to scatter a
relatively significant portion of the incident electromagnetic energy, thereby resulting in
areas with a net increase in the total electric field and/or areas with a net decrease in the
total electric field deposited.
Over extended heating periods the seeds may cause
excessive and/or inadequate heating, in turn resulting in excessive and/or inadequate
thermal dose to regions of the target volume. However, experimentally measured SAR
patterns failed to demonstrate evidence of significant seed-induced perturbations to the
power deposition pattern.
Various seed configurations were tested and the SAR
experiments repeated to ensure statistical confidence. The experimental results indicated
that the titanium brachytherapy seeds had very little effect on the size, shape and
magnitude of the SAR pattern of this interstitial microwave applicator.
90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
Plane wave models, using an edge-based finite element program, were simulated
in the presence of single seeds and seed lines to provide a general understanding of the
interactions of the microwave energy with the metal seeds in three dimensions, and to
provide an understanding of why the interactions between the seeds and the microwave
radiation were experimentally undetectable.
4.1.1
SAR Experiments
The experimental SAR results did not demonstrate any perturbations to the SAR
patterns indicative of seed scattering. SAR patterns were measured on the antenna plane
and planes 2.5mm and 5mm above the antenna plane, in phantoms with and without
brachytherapy seeds.
Visual inspection of these SAR contour patterns as well as
quantification of contour widths and lengths demonstrated little if any changes in the size,
shape and magnitude of the SAR patterns, in spite the inclusion of brachytherapy seeds
configured in three different arrangements.
The presence of inhomogeneities in an
electric field suggests that some scattering should occur.
The lack of experimental
evidence of seed scattering phenomena may be explained then by the limited detectability
of the measurement system, that is, inadequate display pixel size and/or temperature
resolution of the IR thermographic imaging system. Specifically, a net deposition of seed
scatters inducing a rise in temperature distributed over an area less than 1mm2 and/or
with a difference in magnitude less than 0.1°C from its neighbouring pixel temperatures,
would not be detectable with this measurement system. The maintenance of the overall
size and shape of the helical antenna’s SAR pattern in the presence of brachytherapy seed
lines suggests then that scattering events were too small, either in size or magnitude or
both, to be detected by the IR imaging system. Therefore, plane wave simulations were
run to provide some understanding as to the nature of the seed scattering events and
substantiate the hypothesis for which they were empirically undetectable.
91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISSCUSION
4.1.2
Plane Wave Simulations
The plane wave simulations demonstrated scattering of the incident microwave
energy by the seeds. The most significant effects observed in the plane wave simulations
occurred when the plane wave was polarized in the direction of the seed axis and
propagated perpendicular to it (Case I). This scenario was simulated in the presence of a
single seed and seed lines. In these simulations, ‘hot spots’ with an intensity 13 times
that of the incident energy (at that same location), resulted at the edges of the seed(s),
while ‘cold spots’ were evident around the middle of the seed(s). The location of these
hot and cold spots did not concur with the expected location of antinodes and nodes
respectively, as nodes and antinodes occur at quarter-wavelengths (~11mm). These hot
and cold spots were significantly closer to one another and did not occur in the direction
of propagation of the plane wave but rather perpendicular to it, as would be expected for
a standing wave pattern. Therefore, these results may be explained by the scattering of
the incident microwaves from the seeds. At the seed edges, the scattering events together
with the non-perturbed incident energy resulted in a net increase in the power deposited,
subsequently creating hot spots covering an area l-2mm2 and about 13 times that of the
incident energy.
Scattering events in the region surrounding the middle of the seed
resulted in a 10% decrease in the net power deposited suggesting destructive interference
between the scattered and the incident energy.
For those simulations where the axis of the seed was oriented parallel to the
direction of wave propagation (Case II), the perturbing effects induced by the seeds were
much less significant. In this case, the wave was incident on the smallest dimension of
the seed (1mm diameter-as opposed to the 4mm seed length of the first case). Due to the
smaller cross-section of the scatterer (seed), it would seem that a smaller proportion of
the incident energy was scattered, resulting in a net deposition of power 3 times larger
than the incident radiation at that location, but almost 4.S times less than that which
occurred at the edges of the seed in the first case that is, when the seed axis was oriented
perpendicular to the direction of wave propagation (see figure 4.1 (a) and (b)).
92
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
Thus it would seem that the orientation o f the seed relative to the direction of wave
propagation and polarization dictates the extent of the interference effects.
W/m3
0041 Net Deposition
of Scatters:
13x Incident
ooa- Ener«y
Net Deposition
004™ of Scatters:
^
3x Incident
Energy
-0.005
Figure 4.1. Plane Wave Simulations for seed axis oriented perpendicular to wave propagation (a) and seed
axis oriented parallel to direction of wave propagation (b).
93
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
4.1.3
Correlation: Simulations and Experiments
The plane wave simulations suggested that the brachytherapy seeds could perturb
the electromagnetic field. In the worst case, that is, when the plane wave was polarized
in the direction of the seed axis and propagated perpendicular to it, the seeds were able to
scatter a significant portion of the incident radiation, leading to areas with a net increase
in power deposition.
These scattering effects would be detectable by the IR
thermography system, as their intensity is about 13 times the incident radiation, which is
itself within the limits of detectability. In addition, the hot spots are distributed over an
area slightly larger than the display pixel size and therefore, would not be washed out by
averaging effects.
Thus the lack of experimentally determined hot and cold spots
suggests that the scattering events resulting from seeds immersed in a circularly polarized
electric Held (as opposed to a linearly polarized field induced by a plane wave) were too
small to be detectable with the thermographic measurement system. The orientation of
the seeds relative to the propagation direction of this complex radiation pattern may have
minimized the proportion of energy scattered, resulting in an undetectable net deposition
of scatters. Thus, while the nature of the circularly polarized helical antenna makes it
difficult to understand its modes of propagation and scatter, it is this complex nature that
seems to make it suitable for use in salvage therapy, as it seems to minimize the
perturbing effects of the brachytherapy seeds, resulting in the maintenance of an SAR
pattern characteristic of the helical antenna.
The experimental results are promising from a therapeutic perspective, as the
inability to quantify seed scattering events is indicative of their insignificant contribution
to the power deposition pattern of the helical antenna. Over extended heating periods,
typical of TIMT then, the localization of heating to the desired treatment volume and
preservation of surrounding critical tissue structures does not appear to be compromised
by the scattering nature of the brachytherapy seeds.
94
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. D ISCU SSION
4.2
Future Work: Modeling the Helical Antenna
The challenges of making small-scale SAR measurements thermographically with high
spatial resolution, and the complexities modeling circularly polarized radiation, make it
difficult to fully understand how scatterers (as in metal seeds) perturb the field pattern of
a non-linear radiator, like the helical antenna. An accurate electromagnetic model for the
helical antenna would be useful for quantifying and providing an understanding of such
electromagnetic interactions. If a successful model for the antenna could be developed,
then theoretically, one would simply have to model the brachytherapy seeds with the
helical antenna, as a means of quantifying their scattering effects on the SAR pattern.
Mirotznik et al (IEEE, 1993) developed an analytical model of a biomedical interstitial
helical antenna, which they demonstrated to be in good agreement with SAR data
determined experimentally. In this model, the helix is removed, but is represented in the
outer conductor, which is modeled as an anisotropically conducting cylindrical tube.
This model generates a complex analytical solution. We have begun preliminary work
modeling the helical antenna using a numerical approach, specifically an edge based
finite element method, based on Maxwell’s equations. In order to simulate the complex
nature of this radiator, the antenna was modeled using Rhinoceros, a 3-D Nurbs modeling
program, and meshed using the ICEM mesh generator, see figures 4.2-4.6.
The geometrical model of the antenna was based on the description of the helical
antenna provided in Chapter 2, section 2.2. In this model, an annular-type surface was
created at the non-helix end of the antenna. This surface was specified to be a dirichlet
surface, whereby the electric Held solution was forced to be radially polarized. This
dirichlet surface or ‘feed’ surface was coincident with the outer surface of the inner
conductor and the inner surface of the outer conductor. These coincident points in space
permitted the propagation of the radially polarized E-field from the feed surface to the
inner and outer conductors. The connection of the helix to the inner conductor and the
outer conductor at the top and bottom of the antenna respectively, further propagated the
E-field to the helix. Therefore, the inner conductor extended from this feed surface to the
95
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
tip of the helix at the opposite end of the antenna. The outer conductor also extended
from this feed surface and terminated at the base of the helix. The section of antenna
including the inner and outer conductors and the feed surface modeled the coaxial feed
cable present in the real antenna. The helix and the inner and outer conductors of the real
antenna are copper and were therefore modeled as perfect conductors. The dielectric
insulator (which surrounds the outer conductor) and the plastic insulating sheath (that
surrounds the helical region of the antenna) were modeled as a single insulating unit
because they possessed the same electromagnetic properties. While this model is still a
work in progress, it promises to be a useful tool for characterizing the power deposition
pattern of the antenna in the presence of electromagnetically perturbing structures as in
the metal brachytherapy seeds.
Figure 4.2 Rhinoceros Model: Full Antenna. The inner conductor is denoted in red, the outer conductor
is denoted in green, the feed surface is denoted in grey (seen in the top figure at the right hand side of the
antenna) and the helix is denoted in purple. The insulting unit consisting of the plastic sheath and the
dielectric has not been included in this figure, so that it would not interfere with the visualization of the
helix and the inner and outer conductors. The ‘top o f the antenna’ refers to the helix end o f the antenna, as
this is the location of maximum power deposition.
96
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
(b)
Figure 4 J Top of Antenna: Rhinoceros Solid Model (a) and Mesh (b). The helix, the inner conductor,
and the dielectric are shown.
97
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
(a)
(b)
Figure 4.4 Junction: Rhinoceros Model (a) and ICEM Mesh (b). The outer conductor, the inner
conductor and the helix are shown. The dielectric insulator surrounds and is contained within the gaps and
spaces in between each component.
98
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
(b)
Figure 4.5: End of Helix: Rhinoceros Solid Model (a) and ICEM Mesh (b): The outer conductor, the
inner conductor, the helix and the dirichlet feed surface are shown.
99
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
4J
Extended Heating
Although the electromagnetic scattering effects of the titanium brachytherapy seeds were
experimentally observed to be negligible, concern still existed as to the thermal impact of
the seeds on the resulting 3-D temperature distribution.
The increased thermal
conductivity and decreased heat capacity of the seeds suggested perturbation of the
temperature distribution, and therefore potential for the seeds to compromise the
localization of heating to the target volume. However, these anticipated effects were
difficult to detect experimentally. Comparison of relative thermographical temperature
measurements in phantoms with and without seeds indicated little if any seed-induced
changes to the temperature distribution during extended heating, as did measurements
obtained using fluoroptical based point thermometry.
In order to provide some
understanding as to the expected theoretical response of the titanium seeds and the
reasons for which it was experimentally undetectable, thermal models based on the heat
conduction equation were simulated in the presence of seed lines configured according to
the Case-A seed arrangement.
4.3.1
Experimental Temperature Measurements
Following extended heating, temperature measurements were made in phantoms
with and without brachytherapy seeds, using two measurement systems: IR thermography
and point thermometry, in order to quantify the contribution o f the seeds to the resulting
temperature rise distribution.
The IR camera was used to capture thermographical
images for comparison of relative temperature distributions of the no-seed and seed cases
A-C.
In order to quantify any seed-induced differences, temperature profiles were
obtained from the thermographic images and profile widths corresponding to 90%, 70%,
50% and 30% of the maximum temperature rise were measured.
100
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
Examination of these results indicated statistically insignificant differences in
profile width size, in spite the presence of the brachytherapy seeds.
Due to the
conductive and convective cooling artifacts inherent in this thermographic measurement
technique, thermal losses sufficient to wash out the effects of the seeds may have
resulted.
Consequently, a fluoroptical based thermometry system, which measures
temperatures in real time and which is not subject to thermal losses due to cooling, was
employed to serve as an additional means of measuring seed-induced thermal effects.
Subsequently, it was used to aid in determining the reason for which the suspected seed
effects were thermographically undetectable.
Temperature rises were measured at
specific locations at and beyond the seeds during the course of heating and following
power shut off.
These results indicated statistically insignificant differences in the
temperature rises measured at the seeds, or 5mm or 10mm beyond the seeds. In addition
to the magnitude of temperature rise, there was no statistically significant evidence of
variations in rates of temperature rise or in the cooling rates.
Both IR thermography and discrete thermometry measurements indicated
negligible differences in the temperature distributions measured in phantoms with and
without seeds. This suggests that the total volume of titanium contained in 30 seeds was
insufficient to induce a detectable change in the temperature distribution. The proportion
of the total heated volume that belongs to 30 seeds (Case-A and B) is approximately
0.25% (Ft), while the phantom material makes up the remainder, which is approximately
99.75% (Fp). Therefore, if we consider the bulk thermal properties of the phantom
material together with the brachytherapy seeds (which are actually titanium shells as
opposed to solid cylinders), through weighted fractions, this volume of titanium increases
the thermal conductivity of the phantom material by approximately 10%. One may
hypothesize then that this increase was too small to significantly perturb the temperature
distribution. In addition to the insufficient volume of titanium, the air contained inside
the seeds may have also limited the conductive effects of titanium by acting as an
insulator. Air has little vibrational contact between its molecules due to the increased
space between them and is thus a poor conductor of thermal energy.
101
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
The undetectability of seed effects with both types of temperature measurement tools
suggests then that the effects are insignificant, which is promising from a therapeutic
perspective as it implies that TIMT is a viable salvage treatment option, as the seeds do
not compromise the localization of heating to the target volume.
4.3.2
Multiple Seed Lines Case
As was previously indicated, both the thermographical and thermometry
measurements indicated statistically insignificant differences between the no-seed and
various seed cases. While this was also true for the results of the multiple seed lines case
(Case-C), a trend was consistently evident in the results. At all measured Y-axis profile
width levels, the multiple seed lines case consistently indicated smaller temperature
profile widths. Similarly, the temperature measurements obtained using thermometry
also consistently demonstrated smaller temperature rises for this particular seed case.
Surprisingly, these results contradicted theoretical expectations, as the titanium seeds
were expected to perturb the temperature distribution in the opposite direction, that is,
increase the widths of the temperature profiles and the magnitude of the temperature rise.
While statistically insignificant, these results may be due to a change in the electrical
loading of the antenna in the presence of seeds. In seed cases A and B, 30 seeds were
distributed on three planes each separated by 1cm. In this case, 30 seeds were distributed
on the same plane, coincident with the source. This particular seed density may have
been sufficient to induce some loading on the antenna. As a result of antenna loading,
power is reflected back through the antenna feed line, to the power generator. The BSD
power generator is equipped with a feedback circuit to detect this reflected power.
During the multiple seed line experiments however, there was no indication of reflected
power.
This may be the result of a very small amount of loading, which was
subsequently difficult to detect due to losses associated with the feedback circuit of the
generator. In the therapeutic environment then, where similar seed densities are likely
(due to the large numbers of seeds implanted into the prostate for a Brachytherapy
102
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
treatment), the effect of antenna loading will not compromise the coverage and intensity
of the thermal dose administered to the tumour.
4.3.3
Thermal Simulations
While the extended heating simulations were not designed to predict antenna
loading, they were used to provide an understanding of the expected theoretical response
of the seeds and the reasons for which these anticipated effects were not empirically
detectable. It was shown in section 3.4 that the theoretical thermal model was useful for
predicting temperature distributions. Simulated results were generated for both the no­
seed and seed Case-A scenarios, such that a relative comparison of the temperature
distributions could be made.
These transient heat conduction simulation results did
correlate with those theoretical expectations. The increased thermal conductivity of the
titanium and the decreased heat capacity of the titanium seeds resulted in an increase in
the magnitude and distribution of temperature rise, as was initially hypothesized. These
results are depicted in the temperature distribution plots and temperature profiles of
figures 3.22 to 3.24. However, while this model does provide an understanding of the
theoretical thermal response of the titanium seeds in three dimensions, it does not
adequately model the experimental scenario for two reasons. First, the brachytherapy
seeds were modeled as solid titanium cylinders as opposed to titanium shells with a
hollow center. Secondly, the heating time was limited to 116seconds, whereas a heating
time of 300 seconds was implemented experimentally.
The truncated heating time
resulted from a limitation in the maximum computation time allowable by the thermal
solver.
Although the cylinder model is not an accurate representation of the titanium
shell, this model was useful in demonstrating that 30 seeds of solid titanium were
sufficient to induce a thermal response that would be detectable by our measurement
tools.
The total volume of titanium contained in 30 solid seeds is 68mm3.
Experimentally however, we were unable to detect a seed-induced thermal response,
103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4. DISCUSSION
which may be due to the insufficient volume of titanium contained in 30 titanium shells
(~30mm3). In addition, this model did not include the insulating air cavity contained
within the seeds, which may also impede the conductive effects of the seeds. Thus the
thermal model helped to substantiate the hypothesis that the limited volume of titanium
associated with the shell-like structure of the real brachytherapy seeds may have been
insufficient to cause a detectable, seed-induced thermal effect.
4.3.4
Clinical Implications
Temperature measurements obtained using thermography and point thermometry
indicated that brachytherapy seeds consisting of a titanium shell surrounding a hollow air
space, had a negligible effect on the temperature distribution resulting following extended
heating periods. In addition, regardless of the number of seeds, or their location and
configuration relative to the single antenna, their impact on the temperature distribution
was negligible.
These results suggest that TIMT can be implemented as a salvage
treatment for failed Brachytherapy, as the brachytherapy seeds do not compromise the
localization of heating to the target volume.
From a therapeutic perspective, the
undetectable electromagnetic and thermal impact of the seeds suggests minimal changesif any, to the current treatment plan employed in TIMT of prostatic carcinoma following
failed radiation therapy.
104
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
1.
AGEMA Infrared Systems, 1991, Thermovision®400 Series Operating Manual
2.
Andreuccetti D, Bini M, Ignesti A, Olmi R, Rubino N, and Vanni R, 1988, Use of
Polyacrylamide as a Tissue-Equivalent Material in the Microwave Range. IEEE
Transactions on Biomedical Engineering, Vol.35 pp.275-277,1988
3.
Beyer D C, 2001, The Evolving Role of Prostate Brachytherapy. Cancer Control,
Vol.8 pp. 163-170, 2001
4.
Beyer D C, Brachman D G, Thomas T and Hilbe J, 1998, Failure Free Survival
Following Brachytherapy Alone or External Beam Irradiation Alone for T1/T2
Prostate Tumors in 2222 Patients: results from a Single Practice. Proceedings from
the 40fh Annual Meeting o f ASTRO, Phoenix, 1998
5.
Bohris C, Schreiber W G, Jenne J, Simiantonakis I, Rasteri R, Zabel H J, Huber P,
Bader R and Brix G, 1999, Quantitative MR temperature monitoring of highintensity focused ultrasound therapy. Magnetic Resonance Imaging, Vol. 17
pp.603-610,1999
6.
Brawer M K, Stamey T A, Fowler J, Droller M, Messing E, Fair W R, 2001,
Perspective on prostate cancer diagnosis and treatment: a roundtable. Journal o f
Urology, Vol.58 pp.135-149,2001
7.
Brehonnet A, Drossos A, Sinisalo P and Santomaa V, 2000, An automated E-field
scanner for the evaluation of specific absorption rate (SAR) of mobile
telecommunication equipment (MTE) in homogeneous phantoms. Proceedings o f
the 2000 International Symposium on Antennas and Propagation (ISAP2000),
Vol.3 pp.1375-1378, 2000
105
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
8.
Butler W M, Merrick G S, Lief J H, Dorsey A T, 1999, Comparison of seed
loading approaches in prostate brachytherapy. Med. Phys., Vol.27 pp.381-392,
2000
9.
Chan K W, Chou C K, McDougall J A, Luk K H, Vora N L, and Forell B W, 1989,
Changes in heating patterns of interstitial microwave antenna arrays at different
insertion depths. International Journal o f Hyperthermia, Vol.5 pp.499-507, 1989
10.
Cherry P C, Iskander M F, 1993, Calculations of Heating Patterns of an Array of
Microwave Interstitial Antennas. IEEE Transactions on Biomedical Engineering,
Vol.40 pp.771-779,1993
11.
D’Amico el al, 1999, Prostatectomy, External Beam Radiation Therapy, or
Brachytherapy for Localized Prostate Cancer. JAMA, Vol.281 No. 17, 1999
12.
Duck F A, 1991, Physical Properties of tissue. Academic Press limited, New York
13.
Fenn A J, Sathiaseelan V, King G A, Stauffer P R, 1996, Improved Localization of
Energy Deposition in Adaptive Phases-Array Hyperthermia Treatment of Cancer.
The Lincoln Laboratory Journal, Vol.9, 1996
14.
Gladman A S, 1996, Infrared Thermographic Measurement of the SAR Patterns of
Interstitial Hyperthermia Applicators, Master’s Thesis: Graduate Department of
Electrical and Computer Engineering-The University of Toronto
15.
Graham S J, Chen L, Leitch M, Peters R D, Bronskill M J, Foster F S, Henkelman
R M and Plewes D B, 1999, Quantifying tissue damage due to focused ultrasound
heating observed by MRI. Magnetic Resonance in Medicine, Vol.41 pp.321-328,
1999
106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
16.
Grimm P D, Blasko J C, Sylvester J E, Meier R M and Cavanagh W, 2001, 10-year
biochemical (prostate-specific antigen) control of prostate cancer with (125) I
brachytherapy. International Journal o f Radiation Oncology Biol. Phys., Vol.5l
pp.31-40, 2001
17.
Grasso A, Watkinson A F, Tibballs J M, Burroughs A K, 2000, Radiofrequency
Ablation in the treatment of heaptocellular carcinoma-a clinical viewpoint. Journal
o f Hepatology, Vol.33 pp.667-672, 2001
18.
Gross E J, Cetas T C, Stauffer P R, Liu R L and Lumori M L D, 1990,
Experimental assessment of phased-array heating of neck tumours. International
Journal o f Hyperthermia, Vol.6 pp.453-474, 1990
19.
ICEM CFD Engineering, 2000, ICEM CFD Tutorial manual: Meshing Modules
version 4.0. ICEM CFD Engineering, Berkeley, CA
20.
Incropera F, DeWitt D, 1996, Fundamentals of Heat and Mass Transfer 4th Ed.,
John Wiley & Sons Inc., New York
21.
Iskander M F and Tumeh A M, 1989, Design optimization of interstitial antennas.
IEEE Transactions on Biomedical Engineering, Vol.36 pp.238-246,1989
22.
Jia X, Paulsen K D, Buechler D N, Gibbs F A and Meaney P M, 1994, Finite
element simulation of Sigma 60 heating in the Utah phantom: computed and
measured data compared. International Journal o f Hyperthermia, Vol.10 pp.755774,1994
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
23.
Kazantsev Y N, Apletalin V N, Solosin V S, Zubov A S, 2000, Study of
Electromagnetic
Wave
Backscattering
From
Structures
with
Anisotropic
Conductivity. Journal o f Radioelectronics, No. 4, 2000
24.
Khizhnyak E P and Ziskin M C, 1994, Heating Patterns in biological tissue
phantoms caused
by
millimeter wave electromagnetic
irradiation. IEEE
Transactions on Biomedical Engineering, Vol.41 pp.865-873, 1994
25.
King R W P, Smith G S, 1981, Antennas in Matter. Fundamentals, Theory, and
Applications. The MIT Press, Massachusetts
26.
King R W P, Trembly B S and Strohbehn J W, 1983, The Electromagnetic Field of
an Insulated Antenna in a Conducting or Dielectric Medium. IEEE Transactions on
Microwave Theory and Techniques, Vol.MTT-31 pp.574-583,1983
27.
Kumaradas J. Carl, 2002, An Edge-element based Finite Element Model of
Microwave Heating in Hyperthermia: Method and Verification. International
Journal o f Hyperthermia, 2002
28.
LeCarpentier G L, Motamedi M, McMath L P Rastegar S and Welch A J, 1993,
Continuous Wave Laser Ablation o f Tissue: Analysis of Thermal and Mechanical
Events. IEEE Transactions on Biomedical Engineering, Vol.40. pp.188-200,1993
29.
Lee K F, 1984, Principles of Antenna Theory. John Wiley & Sons, Toronto
30.
Lide David R. Editor-in-Chief, 2001, CRC Handbook of Chemistry and Physics
82nd Ed. CRC Press LLC, Boca Raton
108
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
31.
Lorrain P, Corson D, 1962, Electromagnetic Fields and Waves. W.H. Freeman and
Company, New York
32.
LUXTRON, 1993, LUXTRON Fluoroptic® Thermometer Model 3100 Operator’s
Guide. LUXTRON, Santa Clara
33.
Mason P A, Hurt W D, Walters T J, D’Andrea J A, Gajsek P and Ryan K L, 2000,
Effects of frequency, permittivity, and voxel size on predicted absorption rate
values inn biological tissue during electromagnetic-field exposure. IEEE
Transactions on Microwave Theory and Techniques, Vol.48 pp.2050-2058, 2000
34.
McGahan J P, Dodd III G D, 2001, Radiofrequency Abaltion of the Liver: Current
Status. American Journal o f Radiology, Vol. 176,2001
35.
Mettlin C J, Murphy G P, McDonald C J, Menck H R, 1999, The National Cancer
Data base Report on increased use of brachytherapy for the treatment of patients
with prostate carcinoma in the U.S. Cancer, Vol.86 pp. 1877-1882
36.
Mirotznik M S, Engheta N, Foster K R, 1993, Heating Characteristics of Thin
Helical Antennas with Conducting Cores in a Lossy Medium- I: Non-insulated
Antennas. IEEE Transactions on Microwave Theory and Techniques, Vol.41
pp.1878-1886,1993
37.
Morrison H I, MacNeill I B, Miller D, Levy I, Xie L and Mao Y, 1995, The
impending Canadian prostate cancer epidemic. Canadian Journal o f Public Health,
Vol.86 pp.274-278,1995
38.
Neumann B, 1991, Electromagnetic Modeling and Measurements for Analysis and
Synthesis Problems. Kluwer Academic Publishers, The Netherlands
109
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
39.
dzi§ik M Necati, 1993, Heal Conduction 2nd Ed. John Wiley & Sons Inc., New
York
40.
Quinn M, and Babb P, 2002, Patterns and trends in prostate cancer incidence,
survival, prevalence and mortality. Part I: international comparisons. BJU Int.,
Vol.90 pp.162-173, 2002
41.
Raaymakers B W, Van Vulpen M, Lagendijk J J W, De Leeuw A A C, Crezee J
and Battermann J J, 2001, Determination and validation of the actual 3D
temperature distribution during interstitial hyperthermia of prostate carcinoma.
Physics in Medicine and Biology, Vol.46 pp.3115-3131, 2001
42.
Ragde H, Korb L J, Elgamal A A, Grado G L, Nadir B S, 2000, Modem prostate
brachytherapy. Prostate specific antigen results in 219 patients with up to 12 years
of observed follow-up. Cancer, Vol.89 pp. 135-141, 2000
43.
Rine G, Dewhirst M W, Samulski T V and Wallen A, 1987, Modeling of SAR
values in tissue due to slab loaded waveguided applicators. Proceedings o f the
Ninth Annual Conference o f the IEEE Engineering in Medicine and Biology
Society, Cat. No.87CH2513-0
44.
Robert McNeel and Associates, 1993, Rhinoceros® NURBS modeling for
Windows Version 1.0 User’s Guide, USA
45.
Roggan A, Muller G J, Laser-Induced Interstitial Thermotherapy. SPIE Press-The
International Society for Optical Engineers, Bellingham, Washington © 1995.
110
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
46.
Rossetto F, Diederich C J and Stauffer P R, 2000, Thermal and SAR
characterization of multielement dual concentric conductor microwave applicators
for hyperthermia, a theoretical investigation. Medical Physics, Vol.27 pp.745-753,
2000
47.
Rossetto F and Stauffer P R, 1999, Effect of complex bolus-tissue load
configurations on SAR distributions from dual concentric conductor applicators.
IEEE Transactions on Biomedical Engineering, Vol.46 pp.1310-1319, 1999
48.
Ryan T P, Mechling J A, Strohbehn J W, 1990, Absorbed Power Deposition for
Various Insertion Depths for 915MHz Interstitial Dipole Antenna Arrays:
Experiment Versus Theory. International Journal o f Radiation Oncology Biol.
Phys., Vol. 19 pp.377-387,1990
49.
Ryan T P, 1991, Comparison of Six Microwave Antennas For Hyperthermia
Treatment of Cancer: SAR results for Single Antennas and Arrays. Radiation
Oncology Biol. Phys., Vol. 19 pp.377-387,1991
50.
Satoh T, Stauffer P R, Fike J R, 1988, Thermal Distribution Studies of Helical Coil
Microwave Antennas for Interstitial Hyperthermia. International Journal o f
Radiation Oncology Biol. Phys., Vol.lS pp. 1209-1218,1988
51.
Schulman C C, Altwein J E and Zlotta A R, 2000, Treatment options after failure
of local curative treatments in prostate cancer: a controversial issue. BJU
International, Vol.86 pp.1014-1022, 2000
111
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
52.
Sherar M D, Gertner M R, Yue C K, O’Malley M E, Toi A, Gladman A S,
Davidson S R, Trachtenberg J, 2001, Interstitial microwave thermal therapy for
prostate cancer: method of treatment and results of a phase I/n trial. Journal o f
Urology, Vol. 166 pp.1707-1714, 2001
53.
Sherar M D, Gladman A S, Davidson S R H, Trachtenberg J, Gertner M R, 2001,
Helical Antenna arrays for interstitial microwave thermal therapy for prostate
cancer: tissue phantom testing and simulations for treatment. Physics in Medicine
and Biology, Vol. 46 pp.1905-1908, 2001
54.
Sherar M D, Trachtenberg J, Davidson S R H, McCann C, Yue C K K and Gertner
M R, Interstitial Microwave Thermal Therapy for Prostate Cancer, Journal o f
Endourology, In Press.
55.
Shrivastava P, Luk K, Oleson J, Dewhirst M, Pajak T, Paliwal B, Perez C, Sapreto
S, Saylor T, and Steeves R, 1989, Hyperthermia quality assurance guidelines.
International Journal o f Radiation, Oncology, Biol, and Phys., Vol. 16 pp.571-587,
1989
56.
Skinner M G, Iizuka M N, Kolios M C and Sherar M, 1998, A theoretical
comparison of energy sources-microwave, ultrasound and laser-for interstitial
thermal therapy. Phys. Med. Bio., Vol.43 pp.3535-3547,1998
57.
Smith I M and Griffiths D V, 1982, Programming The Finite Element Method 2nd
Ed. John Wiley & Sons, Toronto
58.
Society of Photo-Optical Instrumentation Engineers, 2000, Matching the Energy
Source to the Clinical Need. Proceedings o f SPIE Conference on Thermal
Treatment o f tissue with Image Guidance, 2000.
112
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
59.
Stauffer P R, Satoh T, Suen S A and Fike J R, 1987, Thermal Dosimetry
Characterization of Implantable Helical Coil Microwave Antennas. IEEE Ninth
Annual Conference o f the Engineering in Medicine and Biology Society-1633,
1987
60.
Suroweic A, Shrivastava P N, Astrahan M, and Petrovich Z, 1992, Utilization of a
multi-layer polyacrylamide phantom for evaluation of hyperthermia applicators.
International Journal o f Hyperthermia, Vol.8 pp.795-807, 1992
61.
Ter Haar G., 2001, Acoustic Surgery, Physics Today, Vol.54, 2001
62
Ter Haar G., 1999, Ultrasound Focal Beam Surgery, Ultrasound in Medicine and
Biology, Vol.21 pp. 1089-1100,1999
63
Thomsen S, 1999, Mapping of Thermal Injury in Biologic Tissues Using
Quantitative Pathologic Techniques. SPIE Conference on Thermal Treatment o f
tissue with Image Guidance, SPIE Vol. 3594,1999
64
Vitkin A, Moriarty J A, Kolios M C, Gladman A S, Chen J C, Hinks R S, Hunt J
W, Wilson B C, Easty A C, Bronskill M J, Kucharcyzk W, Sherar M D and
Henkelman R M, 1996, Magnetic resonance imaging of temperature changes
during interstitial microwave heating: A phantom study. Medical Physics, Vol.24
pp269-277,1996
65
Wan H, Aarsvold J, O’Donnell M and Cain C, 1999, Thermal Dose Optimization
for
Ultrasound
Tissue
Ablation.
IEEE
Transactions
on
Ferroelectrics, and Frequency Control, Vol.46 pp.913-928,1999
113
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ultrasonics,
References
66.
Zhang Y, Samulski T V, Joines W T, Mattiello J, Levin R L, LeBihan D, 1992,
On the accuracy of noninvasive thermometry using molecular diffusion magnetic
resonance imaging. International Journal o f Hyperthermia, Vol.8 pp.263-274,
1992
114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
5 397 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа