HIGH FIELD MAGNETIC RESONANCE IMAGING-BASED GEL DOSIMETRY
FOR SMALL RADIATION FIELDS
BY
XUANFENG DING
A Dissertation Submitted to the Graduate Faculty of
WAKE FOREST UNIVERSITY GRADUATE SCHOOL OF ARTS AND SCIENCES
in Partial Fulfillment of the Requirements
for the Degree of
DOCTOR OF PHILOSOPHY
Physics
May, 2012
Winston-Salem, North Carolina
Copyright Xuanfeng Ding 2012
Approved By:
J. Daniel Bourland, Ph.D. Advisor
Carnell Hampton, Ph.D. Chair
David L. Carroll, Ph.D.
Martin Guthold, Ph.D.
Richard Williams, Ph.D.
ACKNOWLEDGMENTS
This work would not exist without my supervisor, Dr. J. Daniel Bourland. His
guidance, invaluable advice, suggestions and ideas are absolutely essential for my every
step of progress in this gel dosimetry world. I also want to thank Dr. Bourland for
introducing me into this wonderful medical physics society and training me from a
curious student to a knowledgeable and determined young medical physicist. I really
appreciate his encouragement and patience in proofreading and correcting abstracts and
publications. His optimism, ability and responsibility set an excellent example for every
one of his students. Dr. Bourland is not only my supervisor in academic research, but the
lighthouse of my life as well.
I also wish to thank the following people for their strong support and guidance
during this project.
I would like to thank John Olson, M.S., the core coordinator of 7T MRI facility,
for his endless help and assistance in running 7T MRI unit. It is really enjoyable to do
research in the Center for Biomolecular Imaging. We have developed many ideas in this
project together. Also I wish to thank Wendy Dolesh, 3T MRI technologist in Radiation
Oncology for her support of the 3T MRI unit.
I really appreciate Dr. David Carroll, Dr. Carnell Hampton, Dr. Martin Guthold
and Dr. Richard Williams for taking the time to serve on my committee. I would like to
thank Dr. David Carroll who led me to the frontier of nanophysics and material science
which gave me the fundamental skills and knowledge to make and improve the quality of
my gels. I would like to thank Dr. Carnell Hampton for his guidance in training me as a
young medical physicist to perform IMRT quality assurance and other clinical duties
ii
independently. Ivery much thank Dr. Martin Guthold for his guidance during my
graduate studies for his help not only in academic study but in real life as well. I would
like to thank Dr. Richard Williams for his guidance in my optics class. His ideas of
experimental design, data analysis and critical thinking have benefited me in my whole
graduate study.
I am so proud of being one of the physics graduates from Wake Forest. I would
like to express my gratefulness to all the co-workers in the Department of Physics and the
Department of Radiation Oncology who contributed to this work and gave me valuable
suggestions and ideas.
At last but not the least, I really appreciate my family for their support, my mother,
my father and especially, my wife Jingying. Her unconditional love, from China to the
U.S.A. continually supports me walking through the difficulties. Thank you.
iii
Contents
Acknowledgements
ii
List of Figures
vii
List of Tables
xv
Abbreviations and acronyms
xvii
Abstract
xx
Chapter I Introduction
1.1 Nature of Ionizing Radiation
1.1.1 Ionization chambers
1.1.2 Radiochromic film
1.1.3 Thermoluminescent dosimeters (TLDs)
1.2 Small field dosimetry: Importance and Challenges
1.3 Gel dosimeter: History and development
1.4 Thesis Outline
1
1
2
3
4
5
9
16
Chapter II Polymer gel dosimeter
2.1 Chemistry mechanisms of polymer gel dosimeter
2.2 Fabrication of polymer gel dosimeter
2.3 Glucose-based MAGIC gel: Sensitivity and dose-response curve
2.4 Evaluation of polymer gel dosimeter
2.4.1 Optical CT
2.4.1 X-ray CT
2.4.2 Ultrasound
2.4.3 Magnetic Resonance Imaging
2.5 Temperature sensitivity and other important constraints
2.6 Summary
18
18
21
29
33
33
35
36
37
38
40
Chapter III Magnetic resonance imaging
3.1 Basic principles of MRI
3.1.1 Nuclear Spin and the Magnetic Moment
3.1.2 Radiofrequency Excitation
3.1.3 T1 and T2 relaxation
3.2 MR pulse sequence for T2 map
3.2.1 Single spin-echo sequence
3.2.2 Multi-spin Multi-echo sequence
42
42
43
46
47
50
51
52
iv
3.3 MRI resolution and Signal-to-Noise Ratio (SNR)
54
Chapter IV Small field application: linear accelerator
4.1 Introduction
4.2 Small radiation field relative output factor measurements
4.2.1 Methods and Materials
4.2.1.1. MAGIC Gel preparation
4.2.1.2. MR imaging and parameters for gel dosimeter
4.2.1.3. Ionizing chamber setup and measurements
4.2.1.4 Radiochromic film measurements and calibration
4.2.1.5 TLD measurements and calibration
4.2.2 Results
4.2.2.1 MAGIC gel
4.2.2.2 Ionizing chamber
4.2.2.3 Radiochromic film
4.2.2.4 TLDs
4.2.2.5 Summary
4.3 Sub-millimeter radiation field dose profile measurement (Ding 2010)
4.3.1 Methods and Materials
4.3.1.1. MAGIC gel preparation
4.3.1.2. Polymer gel and radiochromic film irradiation
4.3.1.3. MR imaging and parameters
4.3.2 Results
4.3.3 Conclusion
58
58
59
60
60
61
63
64
65
66
66
67
68
69
70
72
72
72
72
75
77
80
Chapter V Small field application: Gamma Knife Radiosurgery
5.1 Introduction
5.2 Verification: multiple small targets
5.2.1. Methods and Materials
5.2.1.1. Gel preparation
5.2.1.2. Treatment planning and Gamma Knife irradiation
5.2.1.3. MR Gel scanning
5.2.1.4 Analysis methods
5.2.2 Results
5.3 Verification: single shot, anthropomorphic phantom
5.3.1. Methods and Materials
5.3.1.1. Gel preparation
5.3.1.2. Treatment planning and Gamma Knife irradiation
5.5.1.3. MR Gel scanning
5.3.2 Results
81
81
82
82
82
83
85
86
86
89
89
89
90
91
93
v
5.4 Localization error: incorrect stereotactic imaging due to fiducial
chemical shift
5.4.1 Chemical shift in MRI
5.4.2 Methods and Materials
5.4.3 Results
5.4.4 Conclusion
95
96
100
101
104
Chapter VII Conclusion
108
References
109
Appendix
117
Curriculum Vita
132
vi
List of Figures
Figure 1.1.1
(a) 0.6 cm3 Cylindrical PTW Farmer chamber with removable
buildup cap (b) schematic of the PTW farmer chamber.
Sensitive volume within outer electrode is 0.6 cm3. (c, d) are
the
parallel
plate
chamber
diagram
(Advanced
®
Markus Electron Chamber). Gas between the electrodes is
ionized by radiation beam. The ions and dissociated electrons
move to the electrodes of the opposite polarity, shown by small
circles and arrows, and the accumulated charge is measured by
electrometer. (Courtesy of PTW, Inc.)
2
Figure 1.1.2
A photograph of radiochromic films exposed to 10 x 10 mm2
and 5 x 5mm2 fields (details in Chapter Four)
4
Figure 1.1.3.
Photograph of 3 TLDs (Lithium Fluoride -100) with dimension
3 x 3 x 1mm3, centered on the surface of a solid water phantom
in a large field radiation beam.
5
Figure 1.2.1
(a) ImRT MatriXX from IBA. (b) Dose profile comparison
between treatment plan and dose measured in the ImRT
MatriXX. (c) Gel capsule in a human skull phantom was
irradiated in Gamma Knife using 4mm collimator. 3D
localization errors were observed in coronal, sagittal and axial
planes (c-D,E,F) due to the wrong fiducial filling liquid.
9
Figure 2.1.1
(a) Customized block with “WF” shaped aperture for electron
cone. (b) “WF” shaped electron beam setup schematic using 10
x 10 cm2 electron cone in Varian 2100c linear accelerator.
SSD@100cm(c) MAGIC gel dosimeter was irradiated using
“WF” shaped custom block. The lighter gel color in (c) is the
optically viewable polymerization area due to the radiation
energy deposition. (d) Dose distribution calculated from MR
images in a process that is discussed later in this chapter.
19
Figure 2.1.2
Radiation-induced radolysis of water by high-energy x-ray
occurs in “spurs” The radiolytic products diffuse from the site
of creation while recombination processes take place
20
Figure 2.2.1
Chemicals used for MAGIC gel manufacture.
22
Figure 2.2.2
Spherical gel container cover with plastic wrap and sealed
tightly for gamma radiosurgery quality assurance.
23
vii
Figure 2.2.3
(a)T2 weighted MR images from 3T MR scanner. The dark
area is the polymerised region, but apparently the polymerized
region was not in the irradiated area (b) expected polymerized
area after x-ray irradiation.
25
Figure 2.2.4
MR images using minimum TE to maximize the signal
intensity. High density ‘bone gel’ was irradiated using 4cm x
4cm, 8cm x 8cm, 12cm x 12cm X-ray beam at the same time.
‘Bone gel’ has a three-step dose distribution from 15Gy to 5
Gy. However, due to the low signal intensity or SNR and low
radiation sensitivity, the step dose gradient is not visible in this
MRI image.
26
Figure 2.3.1.
Gel dosimeter setup in solid water for 6MV x-ray calibration.
(a) a vial 1.8 cm diameter x 12 cm in length was filled with
MAGIC gel. (b) Bolus was wrapped around the gel vial to
eliminate the air gap between solid water and vial. (c) the gel
vial was positioned in the center of solid water (6cm in depth)
(d) scheme of the gel dosimeter and X-ray beam setup (e)
farmer chamber setup for verify the dose delivered to the
isocenter.
30
Figure 2.3.2.
(a) T2 weighted MR images for calibration gel dosimeter.
From left to right gel dosimeter were irradiated (0, 2, 4, 6, 8, 10
Gy). ROIs (over 100 voxel) are manually selected near central
axis. (b) R2 calibration curve as a function of dose absorbed in
the gel dosimeter. Two groups of MAGIC gel dosimeters (six
gel dosimeters each group) were irradiated from 0 to 10 Gy.
32
Figure 2.4.1.1
Schematic diagram of the different types of optical CT
scanners: (a) first-generation laser system; (b) fast laser
scanner ; (c) cone-beam CCD scanner; (d) parallel-beam CCD
scanner. (C Baldock 2010) Reproduced with permission.
34
Figure 2.4.2
(a) Density of PAG as a function of dose. (b) Linear
attenuation coefficient (μ) of PAG as a function of density for
X-ray CT gel readout (Trapp J V 2002) Reproduced with
permission.
35
Figure 2.4.3
(a) Ultrasound tomography system. Transmission (b) and timeof-flight image (c) of a PAG irradiated with a 4 × 4 cm2 square
photon beam. Scan lines are acquired at a translational
resolution of 1 mm and a rotational resolution of 2◦ with a total
rotation of 360◦. The reconstructed image resolution was 1.45 ×
1.45 mm2 (Mather and Baldock 2003) Reproduced with
37
viii
permission.
Figure 2.5.1
nMAG gel R2-dose response curve. nMAG gel dosimeter dose
sensitivity will be severely decreased at a temperature of 28 ◦C
(De Deene 2006) Reproduced with permission.
39
Figure 2.6.1
Gel dosimeter protocol. Day 1 fabrication; Day 2 irradiation;
Day 3 MRI scan and data analysis
41
Figure 3.1.1.1
(a) Hydrogen nuclei (protons) or nuclear spin behaves like tiny
rotating magnets, represented by vectors. (b) Normally,
the direction of nuclear spins is randomly distributed. Thus, the
sum of all the spins gives a null net magnetization. (c) Within
a large external magnetic field, B0, spins aligned with the field
with a net magnetic moment along the longitudinal direction.
44
Figure 3.1.1.2
With an external magnetic field B0, spins precess at a certain
frequency called the Larmor frequency
45
Figure 3.1.2.1
A signal detector measures the component of magnetic
momentum M0 which is tipped down on the xy-plane, Mxy. The
flip angle of the M0 depends on the strength and the duration of
B1 (RF)
47
Figure 3.1.3.1
T1 relaxation (spin-lattice relaxation). M0 returns to thermal
equilibrium position along z-axis
48
Figure.3.1.3.2
T2 relaxation (spin-spin relaxation). Mxy dephase
49
Figure 3.1.3.3
MR image slices through the head of a patient with a brain
tumor. (a) T1 weighted image with enhancement of the blood
vessels which appear very bright; (b) T2 weighted image which
is very useful for showing pathological abnormalities. Grey (G)
and white (W) matter of the brain and cerebrospinal fluid (CSF)
can be easily distinguished, in particular in the T1 and T2
images.
50
Figure 3.2.1
Gamma Knife (GK) radiosurgery irradiated gel phantom
scanned on a 3T clinical scanner with two spin-echo pulse
sequences, with calculated R2 maps according to Eq. 3.2.1.
51
Figure 3.2.2.1.
Spherical gel phantom attached to GK is irradiated using the
4mm collimator.
52
Figure 3.2.2.2
Gamma Knife radiosurgery irradiated gel phantom scanned at
7T with the Multi-spin Multi-echo (MSME) pulse sequence
53
ix
(greyscale images)and calculated 3D R2 maps (color image).
Signal intensities (red arrow) of each voxel taken from different
echo time (TE) were fit into exponential decay Eq. 3.2.2
Figure 3.3.1
7T Bruker micro-MR gradiant coil configuration for different
animal or gel sizes.
56
Figure 3.3.2
3T GE Signa® EXCITETM MR scanner and coil configuration
(a) 8 channel head coil (b) 4 channel head coil
57
Figure 4.1.1
1cm x 1cm radiation field compared to the traditional 0.6cc
volume PTW farmer chamber. Partial volume irradiation leads
to dosimetric error.
59
Figure 4.2.1.1
Gel dosimeter setup in solid water for 6MV x-ray output factor
measurement.
61
Figure 4.2.1.2
MAGIC gel dosimeter protocol for MR measurement of
radiation dose. Seven gel dosimeters were irradiated under
different field size (1x1, 2x2, 3x3, 4x4, 6x6, 8x8 and10x10
cm2)
62
Figure 4.2.1.3
Ionization chamber setup in solid water for 6MV x-ray output
factor measurement. (a) Experiment setup scheme (b). Ionizing
chamber PTW:TN30013 in solid water. (c) Capintec PR-05P
setup in solid water.
64
Figure 4.2.1.4.
(a)Radiochromic film setup in solid water for 6MV x-ray
output factor measurement. SAD =100cm SSD = 94cm. (b)
10cm x 10cm radiation field and region of interest (ORI) for
output measurement.
65
Figure 4.2.1.5
(a) TLD experiment setup for output factor measurements and
calibration. (b) TLD setup in the center of the radiation field.
(c) HARSHAW 5500 multi-chip reader
66
Figure 4.2.2.3
EBT2 radiochromic film dose calibration curve. Signal intensity
changes as a function of absorbed dose.
68
Figure 4.2.2.4
TLD (LiF-100) dose calibration curve in 10 x 10 cm2 6MV
photon (3, 4, 5Gy).
69
Figure 4.2.2.5
Relative Output Factor versus field size measured with various
available radiation detectors listed in the paper for 6 MV
71
x
beams. The data are plotted without the consideration of any
corrections or perturbations and are simply the ratio of reading
in a field size to the reference field (10x10 cm2).
Figure 4.3.1.2
Two small fields 10 x 10 mm2 and 5 x 5 mm2 experiment setup.
(a) a cylindrical gel dosimeter is placed vertically in the middle
between solid water. (b) Gafchromic film setup.
73
Figure 4.3.2.2
(a) 2mm diameter custom collimator. (b) Experiment setup for
2mm diameter field irradiation.
74
Figure 4.3.1.3
(a) 7T Bruker micro-MR scanner setup for gel dosimeter. (b)
Gamma Knife radiosurgery irradiated gel phantom scanned at
7T with the MSME pulse sequence for T2 maps calculation. (c)
Gel dosimetry protocol for dose profile measurement in 3T and
7T MRI scanner.
77
Figure 4.3.2.1
(a) Gel phantom R2 map, axial plane, calculated from 7T MRI
scan, pixel size = 0.156 mm. (b) Gel phantom R2 map, axial
plane, calculated from 3T MRI scan, pixel size = 0.254mm. (c)
Radiochromic film irradiated by 5 x 5 mm2 and 10 x 10 mm2
fields. d) dose profile of 10 x 10 mm2 6MV beam at depth of
1.5cm. (e) dose profile of 5 x 5mm2 6MV beam at depth of
1.5cm.
78
Figure 4.3.2.2
Dose profile of 2mm round shaped x-ray beam field measured
using 3T clinic GE SIGNA scanner, 7T micro-MRI Bruker
scanner and radiochromic film.
79
Figure 5.2.1.1.
(a) Spherical Gamma Knife gel phantom attached to the Leksell
GK head frame to simulate the target. (b) GK gel phantom
attached with GK MRI fiducial box for stereotactic MRI scan
and treatment planning.
83
Figure 5.2.1.2.1
Leksell Gamma Plan (LGP): five-single shots using the 4mm
collimator.
84
Figure 5.2.1.2.2
Gel phantom irradiated in GK Perfexion
84
Figure 5.2.1.3
(a)T2- weighted image of GK spherical gel phantom with 4mm
collimator shots delivering five different doses. (b) R2-map of
GK spherical gel phantom calculated in MATLAB. (c) Dose
distribution in Leksell Gamma Knife Planning system with a
0.5mm dose grid size.
85
xi
Figure 5.2.2.1
GK gel phantom R2 calibration curve
87
Figure 5.2.2.2
A comparison of the central profile along the X direction
(10Gy, 15Gy, 20Gy shots) with calibrations made by LGP
treatment planning and measurements by gel dosimeter.
87
Figure 5.2.2.3
A comparison of the central profile along the Y direction (5Gy,
2Gy, 20Gy shots) with measurements made by LGP treatment
planning and gel dosimeter.
88
Figure 5.3.1.1
(a) Human skull phantom filled with water with gel insert to
simulate a target in the brain. (b) Axial MRI structure image
using T1 weighted pulse sequence for GK treatment planning.
Bright circle indicates the gel capsule position relative to
fiducial box and skull.
90
Figure 5.3.1.2
Screenshot from LGP with CT images in the axial (a),
coronal(b), and sagittal(c) planes. The plan shows a single shot
using 4mm collimator, prescribed in the center of the gel target
with the coordinate X:125.8mm, Y:57.5mm, Z: 85.4mm
91
Figure 5.3.1.3
The MR images (50% intensity) overlay on top of CT
images (50% intensity) after automated fusion and
registration process in LGP. Water in CT images is black
because of HU = 0. Bone is black in MRI because of no
water proton.
92
Figure 5.3.2.1
Prescribed dose in Leksell GammaKnife Plan (LGP)
combined with T2-weighed gel MR images. Gel MR
images were first fused with original CT structure images.
Then CT image intensity was adjusted to 0%. Direct
comparison could be done directly in LGP.
94
Figure 5.3.2.2
A comparison of the central profile along central X
direction with measurements made by LGP treatment
planning and gel dosimeter.
95
Figure 5.4
(a) MR fiducial box attached to a spherical-shaped gel
phantom. Fiducial marker channel is highlighted (yellow).
(b)Human skull phantom filled with water with gel insert to
simulate a target in the brain. (c) Axial MRI structure image
using T1 weighted pulse sequence for GK treatment planning.
Bright dots show position of fiducial marker.
xii
96
Figure 5.4.1.1
A frequency encoding gradient was applied throughout the
object. According to the Larmor equation, the proton precess in
a higher frequency locates in high gradient area.
98
Figure 5.4.1.2.
An axial image of the skull phantom. The baby oil filled
capsule (bright circle) is located in the center of the water filled
skull phantom. Six white dots on the left and right are the
fiducial marks indicating the accurate position of each image.
The capsule is 2.45mm shifted towards the frequency encoding
direction (Y direction).
98
Figure 5.4.1.3.
Relative position of fiducial marks between “baby oil” and
copper sulfate MR images. Blue dots are the actual fiducial
mark position. Red dots are the apparent fiducial mark positions
with 2.45cm shift in frequency encoding direction. After
registering MR images in the LGP, we gave a dose in the center
of the capsule. However, the actual object (light blue) is higher
than the apparent object (dark blue), so the dose would be
delivered to the incorrect location.
99
Figure 5.4.2
Protocol used to evaluate the positioning and localization error
using grid phantom and skull phantom with gel capsule insert.
101
Figure 5.4.2.1
(a) Axial MR images of vertical grid phantom, MR images
fused with CT images. Direct measurement from the Teflon
rods in the MR and CT grid phantom images indicate an
average 2.7mm shift in frequency encoding direction in Gamma
Knife treatment planning. (b) Coronal image of horizontal grid
phantom MR images fused with CT images indicating a 2.7mm
shift in slice encoding direction. White dots in CT images are
Teflon rods. Black dots in the MR images are Teflon rods.
102
Figure 5.4.2.2
Sagittal plane of vertical grid phantom MR images fused with
CT images. Direct measurement from the Teflon rods MR and
CT grid phantom images indicate average 3.8mm shift in
frequency encoding and slice encoding direction in Gamma
Knife treatment planning. White line is the CT Teflon rod.
Black line is Teflon rod in MR images.
102
Figure 5.4.2.3.
(a,b,c) Comparison between gel MR images and LGP on the
correct CT fiducial box coordinates. (d,e,f) 3.8 mm shift in
frequency and slice encoding directions is observed.
103
xiii
Figure 5.4.2.4
Dose profile measurements indicate a 3.8 mm shift in slice and
frequency encoding direction.
xiv
104
List of Tables
Table 1.3.1
Summary of anoxic polymer gel dosimeter
13
Table 1.3.2
Summary of normoxic polymer gel dosimeter published in
2000s
MRI dose sensitivity of different gel (Lepage M 2001)
14
Table 2.1.1
Radiochemical yields of intermediates (number of particles per
100eV of absorbed energy) in the radiolysis of pure neutral
water with hard x-rays, gamma-rays or fast electrons. e-aq is the
hydrated election(J 1964)
20
Table 2.2.1.
The whole MRI-based gel dosimetry protocol and required time
to perform each step
24
Table 2.2.2
Comparison of elemental composition, electron densities and
average atomic numbers for MAGIC gel, human muscle tissue
and water (weight fractions denoted as wk) (Fong 2001).
27
Table 2.2.3
MAGIC gel dosimeter formulations
28
Table 2.2.4
Physical component of MAGIC Gel
28
Table 3.1.1.1
Intrinsic magnetic moments and spins of some elementary
particles (from NIST Fundamental Physical Constants)
45
Table 3.2.2
Imaging pulse sequence for GK irradiation spherical phantom
53
Table 3.3.1
List of 7T Bruker MRI coil size for gel dosimetry applications
56
Table 3.3.2
List of 3T MRI gradient coil for gel dosimetry applications
57
Table 4.2
Summary of the devices used for small radiation field relative
output factor measurements.
60
Table 4.2.1.2
List of MRI scanning parameters used to image gel dosimeter
irradiated with different field size.
63
Table 4.2.2.1
Gel dosimetry R2 measurements using 3T clinical scanner and
7T micro canner
67
Table 4.2.2.2
Ionizing chamber measurements for small field relative output
67
Table 1.3.3
xv
14
factors (ROFs)
Table 4.2.2.3
EBT2 radiochromic film measurements for small field relative
output factors (ROFs)
68
Table 4.2.2.4
TLD measurements for small field relative output factors
(ROFs)
69
Table 4.2.2.5
Summary of the ROFs measured in different devices.
70
Table 4.3.1.3.1
List of MRI scanning parameters used to image 1cm x 1cm and
0.5cm x 0.5cm square fields.
76
Table 4.3.1.3.2
List of MRI scanning parameters used to image 2mm diameter
fields.
76
Table 4.3.2.1
Full width at half maximum measurements in 7T.
78
Table 4.3.2.2
Full width at half maximum measurements in 3T.
79
Table 4.3.2.2
Measurements of Full width at half maximum (FWHM) of
2mm diameter beam profile.
79
Table 5.2.1.2.
Summary of the dose plan created for irradiation in GK
phantom
85
Table 5.2.1.3
List of MRI scanning parameters used to image the spherical
GK gel phantom.
86
Table 5.2.2.1
Full width at half maximum measurements along central X axis
89
Table 5.2.2.2
Full width at half maximum measurements along central Y axis
89
Table 5.3.1.2
A single shot for human skull phantom irradiation
91
Table 5.3.1.3
List of MRI scanning parameter used to image the skull
phantom
92
Table 5.3.2.1
Full width at half maximum measurements along central X
direction
94
xvi
Abbreviations and acronyms
Aa – Acrylamide
BANANA – BIS, Acrylamide, Nitrous oxide, and Agarose
BANG – (a.k.a. BANG-1) BIS, Acrylamide, Nitrogen and Gelatine
BIS – N’ N-methylene-bis-acrylamide
CT – Computed Tomography
Dmax – Maximum dose below the surface of a sample/patient from a radiotherapy beam
is delivered.
dmax – Depth below the surface of a sample/patient at which the maximum value of dose
from a radiotherapy beam is delivered.
DNA – Deoxyribonucleic Acid
EPI – Echo Planar Imaging
FOV – Field of View
FSE – Fast Spin Echo
GSR – Gamma Stereotactic Radiosurgery
GK – Gamma KnifeTM
IMRT – Intensity-Modulated Radiation Therapy
LGP – Leksell Gamma Plan
NMR – Nuclear Magnetic Resonance
MAc – Methacrylic acid
MAG – Methacrylic Acid based Gel
MAGIC – Methacrylic and ascorbic acid in gelatin initiated by copper
MAGAS – Methacrylic Acid Gelatin with Ascorbic acid
xvii
MAGAT – Methacrylic acid, Gelatine And Tetrakis
MLC – Multileaf Collimator
MRI – Magnetic Resonance Imaging
MSME – Multiple Spin Multiple Echo
MU – Monitor Units
nMAG – Normoxic Methacrylic acid based Gel
NMR – Nuclear Magnetic Resonance
nPAG – Normoxic Polyacrylamide Gel
OSLD – Optically Stimulated Luminescent Dosimeter
PAG – Polyacrylamide Gel
PAGAS – Polyacrylamide Gel with Ascorbic acid
PAGAT – Polyacrylamide Gel And THPC
PVA – Poly Vinyl Acohol
R1 – spin-lattice relaxation rate
R2 – spin-spin relaxation rate
RF – radiofrequency
ROI – Region of Interest
SRS – Stereotactic Radiosurgery
SSD – Source-Surface Distance
T1 – spin-lattice relaxation time
T2 – spin-spin relaxation time
TE – Echo Time
THPC – Tetrakis (hydroxymethyl) Phosphonium Chloride
xviii
THPS – bis[Tetrakis(Hydroxymethyl) Phosphonium]Sulfate
TLD – Thermoluminescent Dosimeter
TPS – Treatment Planning System
TR – Repetition Time
xix
ABSTRACT
Xuanfeng Ding
HIGH FIELD MAGNETIC RESONANCE IMAGING-BASED GEL DOSIMETRY
FOR SMALL RADIATION FIELDS APPLICATIONS
Dissertation under the direction of
J. Daniel Bourland, Ph.D., Professor of Radiation Oncology, Physics and Biomedical
Engineering
Small field megavoltage photon beams (< 3cm diameter) are used in advanced
radiation therapy techniques, such as intensity modulated radiotherapy and stereotactic
radiosurgery Basic physics and radiobiology studies may use very small fields (<1 mm
diameter). Radiation dose characteristics for these small fields are difficult to determine
in multiple dimensions because of steep dose gradients (30-40% per mm) and conditions
of electronic disequilibrium. Conventional radiation dosimetry techniques have
limitations for small fields because detector size may be large compared to radiation field
size and/or dose acquisition may be restricted to one or two dimensions. Polymer gel
dosimetry, is a three-dimensional (3D) dosimeter based on radiation-induced
polymerization of tissue equivalent gelatin. Polymer gel dosimeters can be read using
magnetic resonance imaging (MRI), which detects changes in relaxivity due to gel
polymerization. Spatial resolution for gel-dose readout is limited to 0.25-0.5mm pixel
xx
size because of the available magnetic field strengths (1.5T and 3T) and the reduced
stability of polymer gelatin at elevated room temperatures.
A reliable glucose-based MAGIC (methacrylic and ascorbic acid in gelatine
initiated by copper) gel dosimeter was formulated and evaluated for small field 3D
dosimetry using 3T and 7T high field MRI for dose readout. The melting point of the
original recipe MAGIC gel was increased by 4oC by adding 10% glucose to improve gel
stability. Excellent spatial resolution of 79um (1.5 hr scan) and 39um (12 hr scan) was
achieved using 7T MRI, proving gel stability for long scan times and high resolution 3D
dosimetry. Two studies demonstrate small field application: 1) 6MV photon field output
factor and dose profile measurements, and 2) localization accuracy and dose profile
measurements for small fields used in gamma radiosurgery.
These results indicate the reliability and versatility of the glucose gel for high
resolution 3D dose measurements using high field MRI ( = 4.8% (3T) and  = 3.6%
(7T)) as the dose readout technique, provided gel manufacturing and handling aspects are
appropriately accommodated.
xxi
CHAPTER ONE
INTRODUCTION
1.1 Nature of Ionizing Radiation
Radiation therapy with ionizing radiation uses high energy photons and charged
particles e.g. electrons and protons that transfer their energy in the medium. Charged
particles as direct ionizing radiation release their energy in the medium directly while
photons as indirect ionizing radiation create electrons or positrons in the medium first
that then ionize. The ability of a photon to ionize an atom or molecule depends on its
energy e.g. X-ray and gamma rays of energy exceeding approximately 10eV. Lowerenergy radiation, such as visible light, microwaves and radio waves are not ionizing.
Ionizing radiation may produce free radicals, such as reactive oxygen species. Free
radical production is a primary basis for the particular danger to biological systems which
easily damage DNA by ionizing or breaking DNA molecules. Radiation therapy, as a
modality of cancer treatment, is the medical use of ionizing radiation to control or kill
malignant cells by damaging the DNA of cancer cells. In order to spare normal tissues or
important healthy organs which radiation must pass through, shaped radiation beams are
aimed from several angles to deposit a much larger dose in the target rather than
surrounding healthy tissues.
1.1.1
Ionization chambers
In order to measure the dose in the target area accurately, the ionization chamber
was first invented to measure the radiation dose at a point. It measures charge from the
1
number of ionizations produced by ionizing radiation within a medium. Normally the
medium is air, but it can also be solid or liquid materials. Figure 1.1.1 (a) is a typical
Farmer chamber which consists of a gas filled enclosure between two conduction
electrodes (anode and cathode) (Boag 1950). The shape of chamber may be in the form of
parallel plates or coaxial cylinders like the Farmer chamber pictured which includes a
central linear electrode and an outer cylindrical electrode that forms the chamber wall
itself. Figure 1.1.1 (c) shows the parallel plate geometry. Bias voltage is needed e.g.
300V. The accumulated charge produced in the gas is proportional to the number of
charged ions. Since then, the ionization chamber has been the gold standard for treatment
plan and machine quality assurance (Peter R. Almond 1999).
a
b
c
d
Figure 1.1.1 (a) 0.6 cm3 Cylindrical PTW Farmer chamber with removable buildup cap
(b) schematic of the PTW farmer chamber. Sensitive volume within outer electrode is 0.6
cm3. (c, d) are the parallel plate chamber diagram (Advanced Markus® Electron
Chamber). Gas between the electrodes is ionized by radiation beam. The ions and
dissociated electrons move to the electrodes of the opposite polarity, shown by small
circles and arrows, and the accumulated charge is measured by electrometer. (Courtesy of
PTW, Inc.)
2
For accurate patient treatment, knowledge of the absolute dose in a reference
beam is required. This is performed through the TG-51 protocol in North America ( P. R.
Almond 1999). This protocol provides the methodology to perform dosimetry in a
reference field, usually 10 x10 cm2 for medical linear accelerator. However, while
ionization chambers are well established and work well for beams that are collimated to
large field sizes, they have difficulties in measurements of small beams, where the
irradiated field size becomes comparable to the maximum electron range (B. E. Bjärngard
1990; C. McKerracher 1999, X. R. Zhu 2000; D.A. Low 2003). When the size of the
irradiated field is relatively large in comparison to the size of the irradiated field,
problems in measurement arises. Challenges in small field dosimetry are discussed in the
next section 1.2.
In this study, MAGIC gel dosimeter was first calibrated using 6MV reference
field 10 x 10 cm2. R2 - dose calibration curve was plotted for future applications. Small
fields’ Relative output factors (ROFs) measurements are compared to the reference field
10 x 10 cm2. Chapter Two described the detailed gel dosimetry calibration process.
1.1.2
Radiochromic film
Radiochromic films are a type of 2D dosimeter in radiotherapy. The film changes
color as a function of the incident dose (Figure 1.1.2). Compared to traditional films,
radiochromic films do not require wet chemical processing, as they are self-developing
and change color automatically in response to radiation dose (A. Niroomand-Rad 1998; P.
Francescon 1998). However, sunlight and other UV sources may have adverse effects on
3
the radiochromic film if the film is exposed for a long time. Radiochomic film is
relatively energy independent, dose-rate independent and it is considered to have a high
intrinsic resolution, over 1200line/mm (W. L. McLaughlin 1991). In this work,
radiochromic film (EBT2, ISP Technologies, Wayne, NJ) was used as a comparison
method to gel dosimetry for measuring the dose profile of small fields produced by jaw
collimators (10 x 10mm2 and 5 x 5 mm2) (Figure 1.1.2) and custom block (2mm diameter)
(Figure 4.3.2.2).
Figure 1.1.2 A photograph of radiochromic films exposed to 10 x 10 mm2 and 5 x 5mm2
fields (details in Chapter Four)
1.1.3
Thermoluminescent dosimeters (TLDs)
TLDs are crystals that can be used as a dosimeter for radiotherapy. The electrons
excited by radiation in TLDs will become trapped in a higher energy state due to
impurities in the crystal. Electrons in the higher energy state will fall back to their ground
state and emit light in proportion to the dose deposited in the TLD when it is heated (G. C.
Taylor 1978). TLD is also almost a point detector with dimensions of 3 x 3 x 1 mm3 or 1
4
x 1 x 1mm3 . By placing several of TLDs within an irradiated region, a measurement of
the dose distribution over the covered volume of the TLD can be obtained (Figure 1.1.3).
In this study, TLD-100 (Lithium Fluoride) 3 x 3 x 1mm3 was one of the methods used to
measure the 6MV small field relative output factors (ROFs).
Figure 1.1.3. Photograph of 3 TLDs (Lithium Fluoride -100) with dimension 3 x 3 x
1mm3, centered on the surface of a solid water phantom in a large field radiation beam.
1.2 Small field dosimetry: Importance and Challenges
In order to deliver enough dose to targets with complex shapes, including
concavities and adjacent normal and critical structures, modern radiotherapy techniques
use high dose gradients between the tumor and healthy tissue delivered with very small
fields (< 3 cm2) (Das 2008). As a result, it is very important to find a suitable tool to
verify such complex and small radiation dose distributions. In the clinic, the most
5
common tools of dosimetry are ion chambers, thermoluminescent dosimetry (TLDs),
radiochromic film and diodes. However, quantifying and accounting for the associated
field-size dependent output factors for such small fields poses several challenges
(Institute of Physics and Engineering in Medicine (IPEM) Report 103).
First, planning system accuracy for small fields is more sensitive to modeling
parameters than for large fields. In particular, effective source size (Aspradakis 2005,
Scott 2008, Sham 2008) and multi-leaf collimator (MLC) modeling (Lydon 2005) impact
dose calculation for small fields more so than for larger fields. As IMRT and stereotactic
radiotherapy typically use many small beamlets, segments or fields to achieve the desired
dose distribution and target coverage, accurate small field size dependent output factors
are required to be modeled within the treatment planning system (TPS).
Second, the challenges of penumbra size versus detector size (Haryanto 2002, Scott
2008), and the impact of changes in the energy spectrum on detector response (Haryanto
2002, Yin 2004, Scott 2008) all complicate the measurement process. Consequently,
there have been multiple incidents recently of incorrect small field size output factors
being measured resulting in the mistreatment of patients (Derreumaux 2008). In these
small fields in millimeter range dosimetry, the most notable are the dose-averaging effect
which arises from the physical scale compared to the field and the averaging of the dose
in a pre-determined size volume (E. Pappas 2006). In such cases, there will be a
significant variation of the dose over the volume that the ionization chamber occupies
resulting in an overestimation of the dose in the penumbra regions and an
underestimation of dose in the central regions.
6
Third, several dosimetric challenges due to this trend are lack of charged particle
equilibrium (CPE). Cavity theory holds when the following two conditions are met: 1.
The cavity must be small compared with the range of charged particles incident on it so
that its presence does not perturb the fluence of charged particles in the medium; 2.
Photon interactions in the cavity are assumed negligible so that the absorbed dose in the
cavity is deposited solely by charged particles crossing it (Faiz M. Khan, The Physics of
Radiation Therapy, 2003). Under these two conditions, the electron fluence in the
absorbing medium and that in the cavity are considered the same, and the ratio of the
dose to medium to the dose in cavity is then reduced to the mass collision stopping power
ratio of medium-to-cavity. This simplified relationship is valid only when the charged
particle equilibrium (CPE) condition is established in the absorbing medium. Li et al
described the lateral range of electrons is the critical parameter to the CPE, rather than the
forward range of the electrons (Li 1995). When the range of charge particles exceeds the
size of the cavity, CPE condition is not established. The presence of a detector can
change the local level of the CPE, adding more perturbations to complicate the problem.
Due to the condition of nonequibrium and partial volume irradiation, a traditional
0.6 cc volume Farmer ionization chamber (thimble design) could not be used for accurate
dose determination for field sizes smaller than 3 x 3 cm2. Instead, a pinpoint ionization
chamber or diodes (P. Metcalfe 1993; C. Martens 2000), diamond detectors (C. Martens
2000) and TLDs are possible instrumentation for these small field measurements. Film
dosimeters (Thwaites 1999; C. Kirisits 2002) can be used for high-resolution 2D
measurement. However, in order to acquire a 3D dose verification, a stack of films must
be irradiated and then individually scanned which is quite labor intensive.
7
The ionization chamber (Figure 1.1.1) is a 1D dosimetry method which provides
accurate point dose measurements. For quality assurance of complex treatment field such
as IMRT, chamber array matrix can be used for 2D dosimetry. Figure 1.2.1 shows the
I’mRT MatriXX, a 2D array dosimeter, used in the clinic for routine IMRT treatment
quality assurance. This device enables the comparison of 2D dose profiles and isodose
lines generated by TPS. Polymer based gel dosimetry, a unique and interesting 3D
dosimetry method, provides not only the 2D dose profile comparison, but also 3D dose
comparisons. Figure 1.2.1 (c) is the example of the Gamma Knife quality assurance using
a polymer gel dosimeter. Localization error was observed in sagittal, axial and coronal
plan as the same time due to the wrong fiducial filling liquid (details in Chapter Five). 3D
dosimetry techniques are becoming more popular in research dosimetry and may be a
future method for routine clinical applications. This dissertation is focused on clinical
applications of gel dosimetry for measuring small radiation field outputs and dose
distributions for linear accelerators and gamma stereotactic radiosurgery(GSR) .
8
a
b
c
Figure 1.2.1 (a) ImRT MatriXX from IBA. (b) Dose profile comparison between
treatment plan and dose measured in the I'mRT MatriXX. (c) Gel capsule in a human
skull phantom was irradiated in Gamma Knife using 4mm collimator. 3D localization
errors were observed in coronal, sagittal and axial planes (c-D,E,F) due to the wrong
fiducial filling liquid.
1.3 Gel dosimeter: History and development
The first gel dosimetry was performed over half a century ago, when Day and
Stein observed color changes in gels containing dyes such as methylene blue during
9
irradiation (Stein 1950). In 1957 Andrews et al investigated depth doses using a
spectrophotometer and pH probe measurements(Andrews H L 1957). Due to the
technology limitation, e.g. lack of readout tools and gel stability issue, accurate dose
mapping was not available until Gore et al in 1984 (Gore JC 1984) showed the ferrous
sulfate chemical dosimeter (Fricke dosimeter) could be probed by nuclear magnetic
relaxometry and imaging using an MRI scanner. The Fricke dosimeter is based on ferrous
ions (Fe2+) which are placed spatially in a gel matrix. Upon irradiation, ferrous ions Fe2+
are transformed into ferric ions Fe3+. The different amount of ferric ion Fe3+ changes the
MRI relaxation time which scales directly with the dose absorbed. However, ion
diffusion is a main disadvantage for the Fricke-type gel dosimeter (Olsson L E 1992), and
results in an unstable and blurred dose distribution. Although different gelling agents
were tested such as gelatin, agarose, sephadex and polyvinyl alcohol (PVA), the diffusion
issue was still significant.
The polymer based gel dosimeter, or polymer gel dosimeter was introduced first
in 1954 where Alexander et al discussed the interaction of ionizing radiation in
polymethylmethacrylate. He found that the molecular weight of the irradiated polymer
was inversely proportional to the radiation dose) (Alexander P 1954). In 1961 Boni used
polyacrylamide as a gamma dosimeter (Boni 1961). Rapid development of polymer gel
dosimeters began in 1993, when Maryanski et al invented the now famous BANG gel
(Bis acrylamide nitrogen gelatin) and PAG gel (Polyacrylamide gelatin). These new
polymer gel dosimeter systems solved the ion diffusion problem. The polymerization
consisted of the addition of monomers and of crosslinking of neighboring polymer chains
induced by the free radicals resulting from water radiolysis. The BANG gel formulation
10
was patented and became commercially available through MGS Research Inc. These
types of gel dosimeters can be read with MRI based on the T1 and T2 relaxation time
changes and with optical-CT based on the color intensity change. Image quality increases
with magnetic field strength for a better accuracy for high field MRI systems. Numerous
polymer gel studies of clinical applications for conformal radiotherapy treatment
verification were taken in 1990s (Maryanski M J 1993; Maryanski M J 1994; Ibbott G S
1997; De Deene Y 1998; Oldham M 1998; Low D A 1999). Applications of these gel
dosimeters with MRI readout continued in all clinical treatment techniques including
IMRT(De Deene Y 2000; Love P A 2003; Duthoy W 2004), linear accelerator and
gamma radiosurgery (Ertl A 2000; Grebe G 2001; Audet C 2002; Novotny J Jr 2002;
Watanabe Y 2002), brachytherapy (Farajollahi A R 1999; Wuu C-S 2003), high-LET and
proton therapy(Ramm U 2000; C 2002; Heufelder J 2003; Gustavsson H 2004) and boron
capture neutron therapy (Farajollahi A R 2000; Gambarini G 2004). However, a major
disadvantage for this kind of BANG and PAG gel is the oxygen contamination which
inhibits radiation induced polymerization(Y. De Deene; Maryanski M J 1993; Maryanski
M J 1994). As a result, these polymer gels have to be fabricated in a special environment
without any oxygen, such as a glove box with noble gas support (inert atmosphere) and
by using an excellent oxygen barrier container constructed of glass and plastics (C.
Baldock 1998).
In 2001, Fong et al. developed a new type of normoxic polymer gel dosimeter,
known as MAGIC, in which atmospheric oxygen was bound in a metallo-organic
complex thus solving the problem of oxygen impeded polymerization (Fong PM 2001).
This significant finding enabled the polymer gel dosimeter to be produced in a normal
11
laboratory. The PAG and BANG gel became known as anoxic gel dosimeters. The
MAGIC gel uses ascorbic acid (Vitamin C) as an oxygen scavenger which binds free
oxygen in the liquid gelatin matrix to metallo-organic coplexes in a process initiated by
copper sulfate (De Deene Y 2002). Since then, other antioxidants were also investigated,
such as tetrakis phosphonium chloride (THPC) (De Deene Y 2002; C 2006). With the
introduction of normoxic gel dosimeters, numerous authors have published results of
work investigating the characteristics and uses for different normoxic polymer gel
dosimeters(Senden R J 2006, De Deene Y 2002).
MRI based gel dosimetry was first introduced also by Gore who found that the
ferrous sulfate Fricke dosimeter (EJ 1927; EJ 1966) could be read through magnetic
resonance. (Gore JC 1984) These Fricke gelatins are able to capture the dose distribution
in three dimension and MRI can be used to quantitatively analyze the dose in these gel
matrices. Their study showed that the T1 relaxation parameter varied proportionally to
dose (Gore JC 1984). However, the ferric ions produced by radiation diffuse so fast,
leading to a decrease in signal intensity and a loss of spatial information.(Baldock C
2001). Schriener in 2002 suggested that the imaging must be performed within two hours
of irradiation to avoid serious degradation of the dosimetric detail. (LJ. 2002) Other
materials were also investigated to improve the gel dosimeter’s stability. Chu et al. in
2000 found that the diffusion process could be reduced by replacing the gelatin matrix
with a poly vinyl acohol (PVA) matrix (Chu KC 2000) . Table 1.3.1 and Table 1.3.2 are
the summaries of the important normoxic and anoxic polymer gels in history. Lepage in
2001 summarized current MRI based gel dosimeter properties (Table 1.3.3). Compared
to PAG based gel, the MAG gel dosimeter is a more sensitive gel dosimeter (De Deene
12
2006; Lepage 2001) which is very helpful in detecting low dose and for determining
small beam dosimetry. This dissertation will mainly focus on the application of high
field MRI-based Methacrylic and ascorbic acid in gelatin initiated by copper (MAGIC)
gel dosimeter.
Table 1.3.1 Summary of anoxic polymer gel dosimeter
Gel dosimeter abbreviations
Chemical components
BANANA (Maryanski et al 1993)
Bis, acrylamide, nitrous oxide, and agarose
BANG-1 (Maryanski et al 1994)
Bis, acrylamide, nitrous oxide, and, gelatin
BANG-2 (Ibbot et al 1997)
BIS, acrylic acid, sodium hydroxide, gelatin
BANG-3 (MacDougall et al 2008)
MAc, gelatine
PAG (De Deene et al 2002)
Primarily acrylamide (or variant) as the monomer,
+various others components
Most of the polymer gel dosimetry studies described in the literature have been
performed on 1.5T clinical MRI scanners which limited the spatial resolution as well as
the gel stability because of a long scan times (normal scan time is less than half hour
MRI). Small field (< 3cm) output factor and dose profile measurement remains a
challenge for the gel dosimeter. In order to improve readout image quality e.g. signal-tonoise ratio (SNR) and spatial resolution, a 3T clinical MRI scanner and a 7T micro-MRI
research scanner were used in this work.
13
Table 1.3.2 Summary of normoxic polymer gel dosimeter published in 2000s
Gel dosimeter abbreviations
Chemical components
MAGIC (Fong et al 2001)
Methacrylic acid, ascorbic acid, hydroquinone,
CuSO4·5H2O, gelatin
Methacrylic acid, ascorbic acid, gelatin
MAGAS (De Deene et al 2002)
MAGAT (De Deene et al 2002)
nMAG (De Deene et al 2006)
PAGAS (De Deene et al 2002)
nPAG (De Deene et al 2006)
PAGAT (Brindha et al 2004)
Methacrylic
acid,
tetrakis
(hydroxymethyl)
phosphonium chloride, gelatin
Methacrylic acid, Bis[tetrakis (hydroxymethyl)
phosphonium] sulfate, gelatin
Acrylamide,
N,N-methylene-bis-acrylamide,
Ascorbic acid, gelatin
Acrylamide,
N,N-methylene-bis-acrylamide,
Bis[tetrakis (hydroxymethyl) phosphonium] sulfate,
Acrylamide, N,N-methylene-bis-acrylamide, tetrakis
(hydroxymethyl)
phosphonium
chloride,
hydroquinone, gelatin
Table 1.3.3. MRI dose sensitivity of different gel (Lepage M 2001)
Monomer
D1/2 (Gy)
δR2/δD-1 (s-1 Gy-1)
AAm/Bis
5.5(±0.1)
0.33((±0.1))
Acrylic acid (AAc)
31.2(±0.1)
0.358(±0.006)
Methacrylic Acid (Mac)
12.5(±0.1)
1.19(±0.05)
1-Vinyl-2-pyrrolidone(VP)
23.6(±0.1)
0.082(±0.004)
2-Hydroxyethyl-methacrylate (HEMA)
41.6(±0.1)
0.046(±0.002)
N-Iso-propylacrylamide (NIPAM)
20(±0.1)
0.14(±0.01)
14
The thesis’s aims are:

Develop and characterize a reproducible, small field polymer gel MRI-compatible
dosimeter based on established recipes. Improve the stability of MAGIC gel
dosimeter for high resolution MRI scan.

Establish a gel readout protocol using 3T and 7T MRI and quantitative dose
conversion algorithms, including precision and accuracy analysis, for absolute
and relative dose readout of 3D polymer gel dosimeters.

Apply the gel dosimeter techniques to particular small field geometries:
o Use a gel dosimeter to quantitatively measure small field dose output
factors of a clinical linear accelerator. Gel dosimetry is scanned in a 3T
clinical MRI scanner and the results are compared with film, TLDs and
different ion chambers in the same experimental setup.
o Compare small field gel dosimeter readout accuracy and precision for MR
field strengths of 3T and 7T with results obtained using radiochromic film.
o Apply polymer gel dosimetry techniques for gamma radiosurgery using
multiple target shots for dose calibration and dose profile measurements.
Additionally, use the polymer gel dosimeter to evaluate dose discrepancies
and localization errors under conditions of stereotactic imaging errors.
15
1.4 Thesis Outline
This thesis includes six individual chapters. Each provides background knowledge
and review articles related to the project or the experimental design and results.
Chapter 1 is a brief introduction to small field dosimetry, background for polymer
gel dosimetry, different types of gel dosimeter as well as the research objectives. A
detailed outline of the thesis is included in this chapter.
Chapter 2 is a detailed review of polymer gel dosimetry. This chapter explains
further the details of the chemical reaction of the polymer gel, general manufacturing
processes, and a brief introduction of the common scanning techniques used to
interrogate gels for dose readout. Development of a glucose based MAGIC gel with
higher melting temperature and greater stability for long image scans, and R2 relaxivity vs.
dose curves are described in this chapter.
Chapter 3 discusses Magnetic Resonance Imaging (MRI) principles, pulse
sequences used for dose readout and the data analysis technique used for reading the gel
dose distribution in a 3T GE Signa® EXCITETM MRI scanner and a 7T Burker Biospin
micro-MRI scanner.
Chapter 4 investigates small radiation field measurements for a 6MV x-ray beam
from a clinical linear accelerator, including relative output factors (ROFs) and dose
profiles. This chapter reviews small beam dosimetry challenges with traditional methods.
Improvements to the resolution (micrometer scale) can be achieved by using a 7T Bruker
Biospin micro-MRI scanner which could be applied to micro-beam radiotherapy. Thus, a
2mm diameter radiation field was measured and analyzed using the techniques developed.
16
Chapter 5 describes the use of polymer gels for MRI based Gamma Knife (GK)
Radiosurgery quality assurance including 4mm collimator absolute dose output and dose
profile measurements. In this application, we also use 3D gel techniques to investigate
the localization and dose distribution errors possible if incorrect fiducial marking fluids
are used for GK stereotactic imaging on a 3T GE clinical scanner.
Chapter 6 presents a discussion and conclusions from all the experiments
performed in this dissertation.
17
CHAPTER TWO
POLYMER GEL DOSIMETER
2.1 Chemistry mechanisms of polymer gel dosimeter
Polymer gel dosimeters are radiation sensitive chemicals in a gel matrix. Different
from an ion chamber which measures electron charges created in the medium (gas), a
polymer gel dosimeter is based on a polymerization mechanism within the gel. The
unique property for these gel dosimeters is that an entire 3D dose distribution can be
acquired and then measured in one scan compared to 1D dosimeters, such as ion
chambers and 2D dosimeters, such as film and diode matrix. Polymer gel dosimeter is
tissue-equivalent with about the same physical density (Table 2.3.4), and as a result
correction factors are not needed to convert the dose from the dosimeter media to the
normal tissue. The advantages are particularly significant in high dose gradient dosimetry.
This chapter will cover the basic chemistry mechanisms, the manufacturing process of
these gel dosimeters and the gel dosimetry imaging protocol. As an example, Figure 2.1.1
shows the glucose-MAGIC dosimeter developed for this work after 18MV electron
irradiation through a “WF” shaped custom block.
18
a
b
c
d
Figure 2.1.1 (a) Customized block with “WF” shaped aperture for electron cone. (b)
“WF” shaped electron beam setup schematic using 10 x 10 cm2 electron cone in Varian
2100c linear accelerator. SSD@100cm(c) MAGIC gel dosimeter was irradiated using
“WF” shaped custom block. The lighter gel color in (c) is the optically viewable
polymerization area due to the radiation energy deposition. (d) Dose distribution
calculated from MR images in a process that is discussed later in this chapter.
Polymer gel dosimeters are based on ionizing radiation-induced polymerization of
monomer molecules. Upon irradiation, water molecules are dissociated into several
highly reactive free radicals and ions R● (Figure 2.1.2 and Table 2.1.1) which is
proportional to the absorbed dose(J 1964). The radiolytic products diffuse from the site of
creation within 1nm range.
19
Figure 2.1.2 Radiation-induced radolysis of water by high-energy x-ray occurs in “spurs”
The radiolytic products diffuse from the site of creation while recombination processes
take place
Table 2.1.1 Radiochemical yields (G-value) of intermediates (number of particles per
100eV of absorbed energy) in the radiolysis of pure neutral water with hard x-rays,
gamma-rays or fast electrons. e-aq is the hydrated election(J 1964)
The water radicals R● initiate the polymerization of monomers in the gel solution
to generate monomer radicals RM●. These monomer radicals react with other monomers
M or polymers Mn resulting in the formation of polymer radicals:
The parameter n indicates the length of the polymer chain.(Y De Deene 2002). The
growth of the polymer continues by chain propagation reactions until the polymerization
20
reaction is terminated by the reaction of two radicals by either combination or
disproportionation. Oxygen also terminates polymer formation. As the production of free
radicals is dependent on the dose delivered to the water, the final degree of
polymerization is also radiation dose dependent.
In Fong’s MAGIC formulation, the oxidation of the ascorbic acid is catalyzed by
the presence of copper (M. M. T. Khan and A. E. Martell 1967).The polymer chain
reaction is catalyzed by complex-derived radicals Rc● , which are produced by the
following reactions (Y De Deene 2002):
Additionally, polymer chain growth can also be terminated by the reduction of Cu2+
2.2 Fabrication of polymer gel dosimeter
Preparation of normoxic polymer gel dosimeters does not require an
atmospherically-controlled glove box. The entire gel manufacturing process in this work
was performed in the Clinical Physics laboratory of the Wake Forest University
Comprehensive Cancer Center. The MAGIC gel can be produced in suitable quantities
and quality once the following materials and accessories are ready:
1. A standard glass jar for mixing the chemicals
2. A water heater with temperature monitor used to heat up the water to 55oC.
21
3. A high precision electric balance.
4. A glass rod or stainless rod to mix the chemicals.
5. The following chemicals (Figure 2.2.1)
a. Methacrylic acid (Extra pure 99.5%, Fisher chemicals)
b. Gelatin (100 Bloom, Gelfoam, Fisher chemicals)
c. Copper sulfate
d. Ascorbic acid
e. Deionized water
6. Sealable plastic containers to store the manufactured polymer gel. The shape
of the container could be customized made, for instance, to match a desired
anthropomorphic or geometric phantom shape.
7. Plastic wrap is used to cover the container as an additional oxygen blocker.
Figure 2.2.1 Chemicals used for MAGIC gel manufacture.
22
Figure 2.2.2 Spherical gel container cover with plastic wrap and sealed tightly for
gamma radiosurgery quality assurance.
Detailed steps for preparing 1000ml MAGIC gel in this work are listed below.
1. Pour 870ml of deionized water into 2L glass container.
2. Add 90g gelatin into the 2L glass container.
3. Heat the gel solution slowly to 55oC using a heat pad. Temperature is
monitored using thermal couple during the whole process. Mix at high speed.
4. Add the 40ml Methacrylic acid in the 2L glass container.
5. Mix at high speed for 5 minutes.
6. Dissolve 0.3522g ascorbic acid and 0.025g cupric sulfate in the Methacrylic
acid gel solution.
7. Mix at high speed for 5 minutes.
23
8. Pour the MAGIC gel solution into phantom container and cool them down to
room temperature.
9. Seal the container using plastic wrap and store them in refrigerator for next
day irradiation.
Accumulated time for a researcher to perform one study is about 4.5 hours. 20 hours
storage time before and after irradiation are not included in the table. (Table 2.2.1).
Table 2.2.1. The whole MRI-based gel dosimetry protocol and required time to perform
each step
Step
Time required (hours)
2
Gel fabrication
Irradiation
0.5
MRI scan
0.5
Data analysis
0.5
Total time
3.5
The stability and low melting point of methacrylic acid based MAGIC gels is an
issue especially for long MR scan times needed for fine spatial resolution. Under these
conditions the MAGIC gel could melt due to ambient room temperatures and MRinduced heating. A pilot experiment using MAGIC gel demonstrated heating effects. T2
weighted MRI images acquired after 30 minutes showed that the gel melted (over 28oC)
during the scan or the process of transportation. (Figure 2.2.3) The dark area is the
polymerized region, but apparently the polymerized region was not in the irradiated area.
24
In 2010, C Wong et al reported similar MAGIC gel melting issues during 7T high
resolution MRI scanning where he found the serious distortion . (C Wong 2010)
a
b.
Figure 2.2.3 (a)T2 weighted MR images from 3T MR scanner. The dark area is the
polymerised region, but apparently the polymerized region was not in the irradiated area
(b) expected polymerized area after x-ray irradiation.
The initial motivation of adding glucose into the original MAGIC gel formulation
was to create a bone density equivalent ‘bone gel’ to investigate the dose absorbed in the
bone during the irradiation. By means of the large solubility of glucose in water, we are
able to create a gel dosimeter density up to 2.4g/cm3 by dissolving 244g glucose in 50oC
100ml gel solution. Due to the low free proton density inside the gel, it is very difficult to
image this high density ‘Bone gel’ and calculate the R2 map. Figure 2.2.4 is the T2
weighted MR images of “Bone density” MAGIC gel irradiated by 4cm x 4cm, 8cm x
8cm, 12cm x 12cm X-ray beams. The step dose gradient cannot be seen in this MRI
image clearly due to the low SNR. However, addition of glucose solved the MAGIC gel
stability and low melting point issues, especially for long scan times needed for micro
resolution.
25
Figure 2.2.4 MR images using minimum TE to maximize the signal intensity. High
density ‘bone gel’ was irradiated using 4cm x 4cm, 8cm x 8cm, 12cm x 12cm X-ray
beam at the same time. Thus ‘Bone gel’ has a three-step dose distribution from 15Gy to 5
Gy. However, due to the low signal intensity or SNR and low radiation sensitivity, the
step dose gradient is not visible in this MRI image.
After seeing the effect of added glucose, we started to develop our own glucosebased MAGIC gel recipe to stabilize the gel dosimeter and increase its melting point by
adding a certain amount of 10% w/w glucose. To test the melting points of the gels, small
quantity of gels from the formulation were set in a small glass jar. The jar was put in a
heated water bath at a small angle (~30o), such that the gel surface was at an angle to the
heated water surface. A thermometer was used to monitor the water temperature. When
the gels started melting, the side wall of the glass jars can no longer hold the solid gels,
and the gel surface would slowly change to a new level and angle within the container.
26
Thus, the melting point temperature was determined (Table 2.2.4). Tissue-equivalent
density is also very important, because dose distribution and actual particle range e.g.
proton, cannot be directly measured in the phantoms whose density is very different from
soft tissue. In other words, dose responding curves measured in the phantom with
different physical density are very different from expected tissue response (Parodi 2008).
As shown in the table 2.2.2, the elemental composition for carbon, hydrogen, nitrogen
and oxygen in MAGIC is very close to the elemental composition for muscle tissue. The
electron density and mean atomic number are closer to those for muscle than for water.
(Fong 2001)
Table 2.2.2 Comparison of elemental composition, density and average atomic numbers
for MAGIC gel, human muscle tissue and water (weight fractions denoted as wk) (Fong
2001).
Material Wc
WH
WN
WO
WS
WCu
ρ(kg m-3)
Za
MAGIC
0.0751
0.1062 0.0139 0.8021 2.58x10-6
5.08x10-6 1060
7.07
Muscle
0.1230
0.1020 0.0350 0.7298 0.00
0.00
1030
6.93
Water
0.00
0.1111 0.00
0.00
1000
7.22
0.8889 0.00
In order to balance the physical density and melting temperature for the MAGIC
gel, a small amount of 10% glucose was added into the original MAGIC gel formulation.
In this study, the densities of the gels were calculated from the weight of the gel inside
the vial divided by the known vial volume.
Two kinds of 4% Methacrylic acid MAGIC gel (with glucose and without glucose)
were produced based on the recipe published in 2001 (Fong et al. 2001). (Table 2.2.3).
Physical density and melting point were measured according the methods described
27
above (Table 2.2.4). The result shows that 10% (w/w) glucose increases the melting point
by about 4oC (from 28oC to 32oC). In most air-conditioned room, a 32 oC melting point
ensures the gel will stay solid even after over-night MRI scan. There is only a very small
change (+2.8%) in the density which is considered very acceptable in our study.
Moreover it didn’t introduce any additional toxicity.
Table 2.2.3 MAGIC gel dosimeter formulations
Chemical
Glucose based MAGIC MAGIC Gel
Component
Gel
Water
77%(w/w)
87%(w/w)
Methecrylic Acid
4% (w/w)
4% (w/w)
Copper sulfate
0.1mM
0.1mM
Ascorbic Acid
2mM
2mM
Gelatin
9%(w/w)
9%(w/w)
Glucose
10%
-
Table 2.2.4 Physical component of MAGIC Gel
Gel type
Density (g/cm3)
Melting point(oC)
Glucose based MAGIC Gel
1.09
32
MAGIC Gel
1.06
28
Soft tissue (ICRU report 44)
1.06
-
28
2.3 Glucose-based MAGIC gel: Sensitivity and dose-response curve
Before using the glucose based MAGIC gel, the basic dose-respond curve and its
radiation sensitivity was studied and compared with traditional MAGIC gel results. Two
batches of MAGIC gel with and without glucose described above were filled in a series
of small vials. Irradiations were performed using a 6MV X-ray beam from a 2100c
clinical linear accelerator (Varian Medical Systems, Palo Alto, CA) with following
parameters: square field 10cm x 10cm, SAD = 100cm, SSD = 94cm, 6MV, Dose rate
=600MU/min, vial irradiation doses of 2Gy, 4Gy, 6Gy, 8Gy and 10Gy (Figure 2.3.1.) A
very thin layer of bolus was wrapped around the gel vial to fill the air gap. Gantry
direction is marked on the vial to indicate the beam direction. These gel vials were kept
in the cooler before and after irradiation. A PTW farmer chamber connected to a
electrometer (INOVISION model:35040 S/N: 99353) was used to verify the dose
delivered to the isocenter in the same setup as gel dosimeter.(Figure 2.3.1.e.)
29
a
c
b
d
e
Figure 2.3.1. Gel dosimeter setup in solid water for 6MV x-ray calibration. (a) a vial 1.8
cm diameter x 12 cm in length was filled with MAGIC gel. (b) Bolus was wrapped
around the gel vial to eliminate the air gap between solid water and vial. (c) the gel vial
was positioned in the center of solid water (6cm in depth) (d) scheme of the gel dosimeter
and X-ray beam setup (e) farmer chamber setup for verify the dose delivered to the
isocenter.
30
The gel phantom was scanned in the 7T MR using a linear birdcage volume coil
(Bruker Corporation, Billerica, MA) with these parameters: Field of View (FOV) =
6.8cm; Slice thickness=2mm; interslice distance = 14mm; Matrix size = 128 x 128; pixel
size = 0.53mm; TE = 30, 60 … 300 ms and TR = 3330ms. Total scan time was about
20mins. Temperature was monitored using a thermal couple. The 7T MRI room
temperature is kept at 21 +/- 1oC, same as the LINAC room. The detailed MR imaging
technique and data analysis are described in the next chapter.
The dose response curves are shown in Figure 2.3.2. The slope increases by a
factor of 3.6 when 10% glucose is added into the original MAGIC gel. The R2 relaxivity
of the gel dosimeter without irradiation increases from 5.23 +/- 0.15s-1 to 6.56 +/- 0.2s-1.
Thus, besides increased temperature stability, the sensitivity of the MAGIC gel is also
improved by adding 10% glucose.
31
a
b
Figure 2.3.2. (a) T2 weighted MR images for calibration gel dosimeter. From left to right
gel dosimeter were irradiated (0, 2, 4, 6, 8, 10 Gy). ROIs (over 100 voxel) are manually
selected near central axis. (b) R2(1/T2) calibration curve as a function of dose absorbed in
the gel dosimeter. Two groups of MAGIC gel dosimeters (six gel dosimeters each group)
were irradiated from 0 to 10 Gy.
32
2.4 Evaluation of polymer gel dosimeter
Gel dosimetry readout can be obtained from MRI, optical CT, X-ray CT and
ultrasound techniques of which optical CT and MRI are the most popular methods. This
section will briefly introduce the techniques mentioned above and the basic principle
behind each.
2.4.1 Optical CT
Unirradiated gel is virtually transparent to visible light, but irradiated gel becomes
opaque due to the polymerization. In 1996, Gore et al first introduced Optical computed
tomography (Optical CT) to scan polymer gel dosimeters (Gore J C 1996). The optical
attenuation of the polymerized gel is dose dependent. Similar to X-ray CT, 3D images
can be acquired by using visible light instead of X-ray (Figure 2.4.1.1) (Gore J C 1996).
According to Beer’s law, the relationship between the measured signal intensity I to the
signal in the absence of the sample I0 is :
(2.4.1)
Where, u is a quantity known as the optical attenuation coefficient and s is the distance
along the selected ray-path through the sample. It showed that optical-CT could provide a
low cost and attractive alternative to MRI for polymer gels applications (Gore J C 1996).
Figure 2.4.1.1 is the development of different optical CT system since 1996 (Baldock
2010). The disadvantage of using optical CT is that the water-bath has to be well matched
to the refractive index of the gel, otherwise a radial compression distortion was observed.
33
Figure 2.4.1.1 Schematic diagram of the different types of optical CT scanners: (a) firstgeneration laser system; (b) fast laser scanner ; (c) cone-beam CCD scanner; (d) parallelbeam CCD scanner. (C Baldock 2010) Reproduced with permission.
34
2.4.2 X-ray CT
The linear attenuation coefficient of polymer gel dosimeters also changes upon
irradiation. This phenomenon is mainly attributed to a change in electron density
originating from the expulsion of water in the polymer cluster. CT images are expressed
as CT numbers (NCT), in Hounsfield units (H). NCT are measures of the linear attenuation
coefficient of the sample (μ) relative to that of water (μ w) Figure 2.4.2 shows the density
of PAG as a function of dose and linear attenuation coefficient (μ) of PAG as a function
of density (Trapp J V 2002).
NCT = 1000 . (μ - μ w)/ μ w
(2.5.2)
Figure 2.4.2 (a) Density of PAG as a function of dose. (b) Linear attenuation coefficient
(μ) of PAG as a function of density for X-ray CT gel readout (Trapp J V 2002).
Reproduced with permission.
35
However, CT readout of polymer gel has a main limitation, in order to achieve better
SNR from the CT, 16 to 32 image averages in the region of interest have to be taken.
Thus, only low sensitivity gels (dose insensitive gels) can be scanned in X-ray CT.
2.4.3 Ultrasound
In 2002, Mather et al found that the acoustic speed of propagation changes with
the amount of radiation-induced polarization (Mather M L 2002). The changes in
ultrasonic speed with absorbed dose for PAG and MAGIC gel dosimeter are related to
both changes in the elastic modulus on absorbed dose and on changes in mass density. A
prototype tomographic ultrasound system is constructed that consists of a translation and
rotation table onto which the phantom is suspended, an ultrasound transducer and a
needle hydrophone Figure (2.4.3). The phantom and ultrasound components are
immersed in a water bath to avoid attenuation in air. Ultrasonic pulses are transmitted by
the transducer and received by the needle hydrophone. However, due to an acoustic
impedance mismatch between water and gel phantom, artifacts in the transmission image
can dominate. There is much room for improvement of image quality by matching the
acoustic properties of the matching fluid, by the use of more sophisticated acoustics
components and improvements in geometric alignment of the system.
36
.
Figure 2.4.3 (a) Ultrasound tomography system. Transmission (b) and time-of-flight
image (c) of a PAG irradiated with a 4 × 4 cm2 square photon beam. Scan lines are
acquired at a translational resolution of 1 mm and a rotational resolution of 2◦ with a total
rotation of 360◦. The reconstructed image resolution was 1.45 × 1.45 mm2 (Mather and
Baldock 2003) Reproduced with permission.
2.4.4 Magnetic Resonance Imaging
The fundamental basis of MRI is the nuclear spin. Spin is an intrinsic property of
elementary particles. A pool of proton spin in a certain volume is called proton pools
which can be thought of hydrogen nuclear precessing along the z axis similar to a
circulating electric current. After the excitation of the radiofrequency (RF), the process in
which the net magnetization returns to its original equilibrium position M0 is called
Relaxation. Spin-spin relaxation of the different proton pools is governed by the rate of
molecular “tumbling” and Brownian motion of the molecules that contain these protons.
Different proton pools are thus characterized by different relaxation rates. MRI based
polymer gel dosimeter readout is a measurement of the gel’s R2 (1/T2) relaxivities. Upon
37
irradiation, gel polymer molecular mobility is significantly reduced by polymerization in
the irradiated volume. The amount of the mobility of the bound water molecules is
reduced, spin-spin relaxation is more effective, which is observed by an increase in R2.
Detailed MR pulse sequences such as the multi-spin multi-echo (MSME) and spin echo
sequences and imaging parameters such as echo time (TE), repetition time (TR) and flip
angle (FA) are discussed in Chapter III.
2.5 Temperature sensitivity and other important constraints
Temperature dependence during irradiation was first found by De Deene in 2006
(De Deene et al. 2006). In their study, they found that nMAG gel dosimeter dose
sensitivity at the time of irradiation was severely decreased at 28 ◦C (above the room
temperature in WFU cancer center). The change in dose sensitivity with temperature
during irradiation amounted to 0.034 s−1 Gy−1 K−1 (Figure 2.5.1). Also nMAG gel
dosimeters are sensitive to the temperature of the gel during scanning. The influence of
the scanning temperature on the R2 values of PAG or nMAG gel dosimeters has been
described (Maryanski et al 1994, 1997, De Deene et al 1998, Spevacek et al 2001). In
order to minimize the effect from surrounding temperature, all the MAGIC gel
dosimeters in this work were irradiated and scanned in Wake Forest University School of
Medicine at the same room temperature (21+/- 1oC). Also, Bayreder et al found out that
the R2 relaxation of an irradiated gel dosimeter will be increased as a function of time.
Bayreder in 2010 shows the R2 relaxation increases as a function of time between
preparation and irradiation (Bayreder 2010). For absolute dose measurements, calibration
38
gel dosimeters and gel dosimeter samples should follow the exact protocol timeline and
be scanned at the same room temperature to minimize the temperature and time effects.
Figure 2.5.1 nMAG gel R2-dose response curve. nMAG gel dosimeter dose sensitivity
will be severely decreased at a temperature of 28 ◦C (De Deene 2006) Reproduced with
permission.
The sensitivity of polymer gels to oxygen has been discussed extensively (Fong
2001), and several investigators have responded by developing gels that contain oxygen
scavengers, such as the MAGIC gel (Fong 2001). Although it can remove additional
small amounts of oxygen, the scavenger ultimately will be overwhelmed if exposure to
normal atmosphere is ongoing. While this problem has been addressed by using plastic
containers which works efficiently as an oxygen barrier, it still creates a minor
inconvenience that might limit the successful introduction of gels into routine clinical use.
39
2.6 Summary
Polymer gel dosimetry contains three steps: Fabrication, Irradiation and Scanning.
(see Figure 2.6.1). In order to obtain accurate and constant results and minimize the gel
manufacture, time and temperature effect, a strict protocol should be followed. In this
study, all the gel dosimeters are measured one day after irradiation. Dosimeters were
placed in the MRI and LINAC treatment room over 30min before scanning and
irradiation to get thermal equilibrium. Room temperature in the LINAC and MRI scanner
room is the same at about 21+/-1 oC.
Nowadays, research in the field of polymer gel dosimetry has investigated a
variety of clinical techniques. Routine clinical use of polymer gels has certain challenges
for handling. Acrylamide based PAG is a severe neurotoxin, and a suspected human
carcinogen and teratogen that is readily absorbed through the skin (MSDS acrylamide,
version 1.11 Sigma-Aldrich Canada, Ltd) . In MAGIC gel formulation, Methacrylic Acid
(99.5%) is a kind of strong acid. Small spills would cause personal injury. Personnel
should be trained and follow the gel fabrication protocol carefully so that they know how
to react to minimize the impact of an accident. Currently, researchers are investigating
more user friendly and less toxic gel as well as special imaging read-out techniques to
help and improve the accuracy of gel dosimetry (R J. Senden 2006). For more clinical
aspect, one major disadvantage of the current MRI-based MAGIC gel dosimetry is that it
is very labor intensive and the costs include an expensive MR unit.
40
Figure 2.6.1 Gel dosimeter protocol. Day 1 fabrication; Day 2 irradiation; Day 3 MRI
scan and data analysis
41
CHAPTER THREE
MAGNETIC RESONANCE IMAGING
3.1 Basic principle of MRI
The concept of Nuclear Magnetic Resonance (NMR) was found by Rabi et al. in
the 1930’s. With this foundation, Bloch et al. (Bloch 1946) and Purcell et al. (E. M.
Purcell 1946) began investigating an effect of the precession of the spins in a magnetic
field. The famous Bloch equation paved the road for the powerful 3D imaging
technique—Magnetic Resonance Imaging (MRI). Both Dr. Lauterbur (Wüthrich 1986)
and Dr. Mansfield (A. Kumar 1975) received the 2003 Nobel Prize in Medicine for their
contribution of obtaining the first images based on NMR and first human body MRI cross
section. Compared to other imaging modality techniques like CT and Ultrasound, MRI
offers noninvasive high resolution, high contrast images of soft tissue structures in the
human body without the use of ionizing radiation. As a result, MRI is a popular imaging
modality for both clinicians and researchers. There are about 3000 MRI units located in
the United States and 20,000 all over the world.
This chapter covers general MRI physics principles, relaxation properties and
imaging techniques relative to gel dosimetry as well as detailed pulse sequences and
algorithm for T2 map calculation which is used to convert to dose map. For general
information about MRI, two websites and two text books are recommended. The website
www.e-mri.org is a very good introduction for beginners. It explains MRI physics
42
principles in an interesting and simple way without equations including animation tools
to simulate k-space reconstruction, T1 and T2 weighted contrast image acquisition and
MRI unit operation. A second website, www.revisMRI.com, has excellent Q&A sections
which cover most of the fundamentals of MRI. MRI Physical Principles and Sequence
Design by Haacke et al (Haacke). covers classical and quantum mechanical principles,
signal detection and acquisition, image artifacts, image reconstruction methods, k-space
trajectories, flow issues, and MR equipment issues; Another reference book, Handbook of
MRI Pulse Sequence (Matt A. Bernstein) is an advanced, in-depth treatise of MRI theory.
It is a first choice for scientists and engineers who have fundamental knowledge of MRI
physics and pulse sequence design.
3.1.1 Nuclear Spin and the Magnetic Moment
The fundamental basis of MRI is the nuclear spin. Spin is an intrinsic property of
elementary particles. It is a very abstract concept. However, proton spin can be thought of
hydrogen nuclear precessing along the z axis similar to a circulating electric current and
it’s associated magnetic moment (Fig. 3.1.1.1.a). The sum of all the tiny magnetic
moments of each spin is called the net magnetization or macroscopic magnetization.
Normally, such as in a bottle of water, the proton spins point in random directions.
Without an external magnetic field the sum of the random spin is equal to zero, or in
other words, null net magnetization. When a strong magnetic field is applied, some of the
spins align with the field (parallel) and some align against the field (anti-parallel).
Nuclear spin will "wander" towards an orientation with a lower magnetic energy (parallel)
43
than an orientation with a higher magnetic energy (anti-parallel). Over time, a net
magnetization forms. (Fig. 3.1.1.1.c)
a
b
c
Figure 3.1.1.1 (a) Hydrogen nuclei (protons) or nuclear spin behaves like tiny rotating
magnets, represented by vectors. (b) Normally, the direction of nuclear spins is randomly
distributed. Thus, the sum of all the spins gives a null net magnetization. (c) Within
a large external magnetic field, B0, spins aligned with the field with a net magnetic
moment along the longitudinal direction.
The equation to describe the magnetic moment μ is
μ=γћІ
(3.1.1.1)
where γ is the gyromagnetic ratio, ћ is the Planck´s constant (6.606 ×10-34 J/s) and I is
the angular momentum. In clinical imaging, the most commonly studied nucleus is 1H
because the human body is mainly composed of water and it has a detectable magnetic
moment. 13C, 19F, 23Na and 31P are other nuclei with magnetic moments that can be used
for biomedical research as well. Table 3.1.1.1 is the list of magnetic moments of some
elementary particles.
44
Table 3.1.1.1 Intrinsic magnetic moments and spins of some elementary particles (from
NIST Fundamental Physical Constants)
Particles
Magnetic Dipole moment in SI unit (10-27J.T-1)
Spin quantum number
electron
-9284.76
1/2
proton
14.106
1/2
neutron
-9.6623
1/2
muon
-44.90448
1/2
deuteron
4.3307346
1
triton
15.046094
1/2
helion
-10.746174
1/2
A wobbling of spin about the axis of the B0 field is called precession (Figure
3.1.1.2). Spinning precession corresponds to the gyration of the rotating axis of a
spinning body about an intersecting axis. For a particular magnetic field strength B0, the
Resonance frequency ω0 of precession can be calculated according to the Larmor
equation.
ω0 =γ B0
(3.1.1.2)
where γ is the gyromagnetic ratio, and B0 is external magnetic field strength.
Figure 3.1.1.2 With an external magnetic field B0, spins precess at a certain frequency
called the Larmor frequency
45
This equation is fundamental to MRI physics. In MRI, resonance corresponds to
the energetic interaction between spins and electromagnetic radiofrequency (RF). Protons
that precess with the same, ω0, frequency as the electromagnetic RF pulse will respond to
that RF pulse.
3.1.2 Radiofrequency Excitation
The energy absorption of a photon by radiofrequency (RF) excitation only occurs
at the resonance frequency (ω0). RF excitation changes energy levels and spin phases.
Magnetic momentum M can be split into z-axis, (longitudinal Mz) and xy-plane
(transverse Mxy) components. Without RF excitation, the proton spin has net
magnetization M0 = Mz. When a spin is excited by a 90 degree RF pusle, the net
magnetization vector from z direction tips down to the xy-plane (transverse plan). The
vector component Mxy creates an oscillating magnetic field and induces a signal in the
detection equipment (Fig. 3.1.2.1). The flip angle is determined by the strength and
duration of the RF.
46
Figure 3.1.2.1 A signal detector measures the component of magnetic momentum M0
which is tipped down on the xy-plane, Mxy. The flip angle of the M0 depends on the
strength and the duration of B1 (RF)
3.1.3 T1 and T2 relaxation
After the excitation of the RF, the process in which the net magnetization returns to
its original equilibrium position M0 is called Relaxation. Relaxation combines by two
different mechanisms: First longitudinal relaxation corresponds to longitudinal
magnetization recovery (T1 relaxation); Second transverse relaxation corresponds to
transverse magnetization decay (T2 relaxation). Both relaxations are the intrinsic
properties of tissue and its surroundings and are described by the phenomenological
Bloch equations:
Longitudinal relaxation
(3.1.3.1)
Transverse relaxation
(3.1.3.2)
47
Where Mz is the longitudinal component of relaxation, and Mx,y is the transverse
component. Longitudinal relaxation is due to energy exchange between the spins and
surrounding lattice (spin-lattice relaxation, T1), coming back to thermal equilibrium state.
RF energy is released back into the surrounding lattice, as spins go from a high energy
state back to a low energy state. The recovery of longitudinal magnetization follows an
exponential curve. The recovery rate is characterized by the tissue-specific time constant
T1. After time T1, longitudinal magnetization has returned to 63 % of its final value. With
a 1.5 T field strength, T1 values are about 200 to 3000 ms. T1 values are longer at higher
field strengths.
Figure 3.1.3.1 T1 relaxation (spin-lattice relaxation) . M0 returns to thermal equilibrium
position along z-axis.
Transverse relaxation is due to magnetic field interaction (spin-spin interaction,
T2), which changes spins’ Mxy precession rates. Spin-spin relaxation causes a cumulative
loss in phase resulting in transverse magnetization decay. Transverse magnetization
48
decay is described by an exponential curve, characterized by the time constant T2. After
time T2, transverse magnetization has lost 63 % of its original value. T2 is tissue-specific
and is always shorter than T1. Transverse relaxation is faster than longitudinal relaxation.
T2 values are unrelated to field strength.
Figure. 3.1.3.2 T2 relaxation (spin-spin relaxation). Mxy dephase
There are three most commonly acquired MRI image types: T1-weighted, T2weighted and proton density weighted. By changing the pulse sequence parameters, like
the echo time (TE) and repetition time (TR), the same image with different contrast can
be generated according to the T1 and T2 differences between tissues. The basic expression
for image contrast is described as :
CAB=SIGNALA-SIGNALB
(3.1.3)
In this way it is possible to distinguish between different types of tissue (white
and gray matter and cerebrospinal fluid in the brain (Fig. 3.1.3.3), to detect and delineate
49
pathological tissue alterations which occur for instance in tumors, and to measure many
properties of tissue both qualitatively and quantitatively, such as blood perfusion, oxygen
concentration, blood throughput of the heart, diffusion, and the concentration of different
metabolites. MRI techniques therefore play a key role in medical diagnostics today.
Figure 3.1.3.3 MR image slices through the head of a patient with a brain tumor. (a) T1
weighted image with enhancement of the blood vessels which appear very bright; (b) T2
weighted image which is very useful for showing pathological abnormalities. Grey (G)
and white (W) matter of the brain and cerebrospinal fluid (CSF) can be easily
distinguished, in particular in the T1 and T2 images.
3.2 MR pulse sequence for T2 map
This thesis is focused on the T2-weight images that are used for calculating R2
maps. There are different MRI pulse sequences to obtain quantitative R2 maps. The most
commonly used are the single spin-echo and multi-spin sequences. By changing the TE,
the signal intensity of T2 weighted images can be varied. The R2 (1/T2) can be calculated
for each voxel.
50
3.2.1 Single spin-echo sequence
This spin-echo sequence is available in most of clinical scanner. A simple R2 map
can be calculate from two spin-echo sequence with different TE, according to
(3.2.1)
Figure 3.2.1 Gamma Knife (GK) radiosurgery irradiated gel phantom scanned on a 3T
clinical scanner with two spin-echo pulse sequences, with calculated R2 maps according
to Eq. 3.2.1.
51
3.2.2 Multi-spin Multi-echo sequence
In order to get better image quality and improved signal-to-noise ratio in a limited
scanning time, De Deene (De Deene Y 1998) suggested using Multi spin-echo. In our 7T
micro-MRI unit, a commercially provided Multi-spin Multi-echo sequence is available.
Figure 3.2.2.1 is a pilot study of Gamma Knife irradiation phantom using MAGIC gel
formula described in Chapter Two. To calculate R2 map, signal intensity of each
individual pixel in a series echoes were fit exponentially according to Eq. 3.2.2 (Figure
3.2.2.2).
S(t) = S0 e –t/T2
(3.2.2)
Figure 3.2.2.1. Spherical gel phantom attached to GK is irradiated using the 4mm
collimator.
52
Table 3.2.2. MR Imaging pulse sequence for GK irradiation spherical phantom
Imaging modality
Pixel
size SNR
MRI parameters
(mm)
7T MRI
0.53 x 0.53
89
FOV = 6.4cm Slice Thickness=1mm
interslice distance = 1mm; Matrix size
= 128 x 128; pixel size = 0.53mm;
TE = 11, 22 … 330 ms and TR =
5000ms
Figure 3.2.2.2 Gamma Knife radiosurgery irradiated gel phantom scanned at 7T with the
Multi-spin Multi-echo (MSME) pulse sequence (greyscale images)and calculated 3D R2
maps (color image). Signal intensities (red arrow) of each voxel taken from different echo
time (TE) were fit into exponential decay Eq. 3.2.2
53
3.3 MRI resolution and Signal-to-Noise Ratio (SNR)
For polymer gels irradiated using very small and microbeam fields, an increased
image resolution is required. For MR imaging, the signal from a small voxel is obtained
to calculate the T2 signal decay. One limitation of selection of voxel (3D) or pixel (2D)
size is the hardware of the scanner, especially the precision and strength of the magnetic
field gradient of the scanner and RF coil. The stronger the gradient of the magnetic field,
the more precise the localization is to allow MR imaging with micrometer-size pixel (X
Ding 2010). In general, 1.5T clinical scanners are capable of a resolution on the order of
half a millimeter. 3T clinical scanners are capable of a resolution in the order of quarter a
millimeter. Besides spatial resolution, SNR, another factor of image quality, is critical to
accurately calculate the value of T2. Noise generates uncertainty in the T2 value
measurement, leading to uncertainty in dose resolution (C Baldock 2001). SNR in terms
of voxel size, number of measurements and receiver bandwidth is defined below:
(3.3.1)
Besides voxel size, number of measurements and receiver bandwidth, there are
many other factors which can control the SNR e.g. RF screening (noise in the imaging
environment), RF noise from the receiver and transmit system, different flip angle and
TE, TR parameters, and especially the main magnetic field strength.
The hypothetical achievable benefit in SNR might be a factor of 4 after
optimization of imaging parameters when the magnetic field strength is changed from 1.5
to 3T in clinical situations. MR imaging at a higher magnetic field strength (> 3 Tesla)
54
will be opening a new arena in MRI based gel dosimetry, especially in the situation of
very small field irradiation.
The Wake Forest University Center for Biomolecular Imaging has a 7T microMRI scanner (Bruker Corporation, Billerica, MA) which is specifically designed to make
high resolution images for research with small animals. Furthermore, it is equipped with
a mouse gradient coil, a 35mm quad coil with max gradient 1000mT/m (Figure 3.3.1).
Until now, there are only 15 of these new RF coils being used in research. While a 7T
MR scanner has advantages of increased spatial resolution, disadvantages include the
high cost and its small RF coil size also limits the sample size to diameter less than 10cm.
To measure the SNR of each MRI image, two region of interest (ROI) are taken.
One is the mean pixel value of the object and second is the noise to be the standard
deviation (σ) in pixel value in background air. SNR is calculated using equation
SNR = 0.655 x S/ σ
(3.3.2)
The 0.655 factor is due to the Rayleigh distribution of the background noise in a
magnitude image (which arises because noise variations, which can be negative and
positive, are all made positive which artificially reduces σ a bit) (Hendee, Medical
Imaging Physics, 4th ed. ).
55
A summary of the gradient and RF coil configurations for the 3T GE clinical and
7T Bruker micro-MRI scanners are used in this study are listed below and shown in
Figure 3.3.1. and 3.3.2
Figure 3.3.1 7T Bruker micro-MR gradiant coil configuration for different animal or gel
sizes.
Table 3.3.1 List of 7T Bruker MRI coil size for gel dosimetry applications
Gradient Coil
Max gradient strength
RF coil size
Applications
Mouse Gradient Coil
~ 1000mT/m
3.5cm
Rat Gradient Coil
~600mT/m
7.2 - 5.0 cm
High resolution scan only
for small gel dosimeter
Medium resolution scan
Monkey Gradient Coil
~300mT/m
15cm
56
Medium resolution scan for
large gel dosimeter e.g. GK
spherical phantom, “WF”
shaped irradiation phantom
a
b
Figure 3.3.2 3T GE Signa® EXCITETM MR scanner and coil configuration (a) 8 channel
head coil (b) 4 channel head coil
Table 3.3.2 List of 3T MRI gradient coil for gel dosimetry applications
Gradient Coil
Max gradient strength
RF coil size
3T GE EXITE
~ 40mT/m
8 channel head Medium resolution scan
coil (24cm)
4 channel head Medium resolution scan for
coil (28cm)
GK head phantom and
spherical phantom attached
to GK head frame
57
Applications
CHAPTER FOUR
SMALL FIELD APPLICATION: LINEAR ACCELERATOR
4.1 Introduction
Development of advanced radiation therapy techniques has included intensity
modulated radiotherapy (IMRT), volumetric modulated radiotherapy (VMRT),
tomotherapy, robotically controlled radiotherapy, gamma radiosurgery and stereotactic
body radiation therapy (SBRT). These image-guided techniques allow more accurate
targeting of the tumor while sparing the healthy tissues. Each uses small radiation fields
(<3cm diameter and as small as 4mm diameter) either as individual fields or as subsets of
larger fields. The dosimetry for these small fields can result in substantial error because
of the electron disequilibrium, the large dose gradients in the radiation beams, and
mismatches in the beam size of the radiation field compared to the detector size (Figure
4.1.1). Accurate dosimetry of such small fields usually requires special small volume ion
chamber, diodes detectors, TLD and films (Das 2008; Zhu 2010). A growing number of
authors have reported the comparison of measured data with Monte Carlo simulation.
However, at the present time the Monte Carlo simulations alone can’t be assumed to
invariably provide a gold standard without appropriate experimental validation (F. Araki
2003; F. Sánchez-Doblado 2003; F. Sánchez-Doblado 2005; Araki 2006; F. Crop 2007).
Polymer gel dosimeters are tissue-equivalent, their response to radiation is relatively
energy independent especially at megavoltage energies and dose distribution in these gels
can be determined by MRI scanning (De Deene Y 2002). We investigated the use of a
58
glucose-based Methacrylic and Ascorbic acid in Gelatin Initiated by Copper (MAGIC)
gel dosimeter to quantitatively measure small field output factors (OFs) for 6MV x-rays.
Two high field MR scanners, a 3T GE clinical scanner and a 7T Bruker micro-MRI
scanner were used to measure the gel dose profile and output factors. Ionizing chambers ,
radiochromic film and TLDs were also used as a comparison.
Figure 4.1.1. 1cm x 1cm radiation field compared to the traditional 0.6cc volume PTW
farmer chamber. Partial volume irradiation leads to dosimetric error.
4.2 Small radiation field relative output factor measurements
Output measurements were performed using a 6MV X-ray beam from a 2100c
clinical linear accelerator (Varian Medical Systems, Palo Alto, CA). Square fields (1x1,
2x2 3x3, 4x4, 6x6, 8x8 and 10x10 cm2) defined by symmetrical collimator jaws. Three
ionization chambers (PTW:TN30013, Exradin A12, Capintec PR-05P) , radiochromic
film: EBT2, TLD: LiF-100 as well as glucose-based MAGIC were used to measure beam
output in a 30 x 30 x 12cm3 solid water. SAD = 100cm, SSD = 94cm, 6MV, dose rate
59
=600MU/min. 512MU was delivered to the isocenter, such that the dose (from 10 x 10
cm2 to 1 x 1 cm2) is within the linear dose response region of the MAGIC gel. Relative
output factors (ROFs) were calculated according to Eq. 4.2
ROF = DFS/ D10x10
(4.2)
where DFS is the dose in different field size, D10x10 is the reference dose in 10 x 10 cm2
field. Table 4.2 is the list of devices used for small radiation field relative output factor
measurements. It should be noted that the relative output factors measured (d = 6cm) are
not identical to quoted output factors used clinically (dmax = 1.5cm), however, the two
output factors can be related by using the tissue maximum ratio.
Table 4.2 Summary of the devices used for small radiation field relative output factor
measurements.
Device
PTW:TN30013
Exradin A12
Capintec PR-05P
LiF-100
EBT2
Glucose-base MAGIC
Type
Ionizing chamber
Ionizing chamber
Ionizing chamber
TLD
Radiochromic film
Gel dosimeter
Volume
0.6cc
0.6cc
0.05cc
3x3x1mm3
Readout area size (pixel)
Readout volume size (voxel)
4.2.1. Methods and Materials:
4.2.1.1. MAGIC gel preparation
As described in Chapter II, the formula of the normoxic MAGIC polymer gel is
based on Fong et al in 2001 (Fong 2001) modified for longer stability by adding 10%
glucose [Gelatin 9%; Methacrylic acid 4%; CuSO4 0.1mM; Ascorbic acid 2mM; Glucose
60
10% and deionized water]. Similar to X-ray dose calibration setup (Chapter II), the
MAGIC gel was instilled into plastic vials 1.8cm in diameter and 12cm in length and
sealed with plastic wrapper to prevent outside oxygen from entering. Then, the completed
MAGIC gel phantom was stored in a refrigerator (4oC) to set overnight before irradiation.
The gel dosimeter was wrapped with a thin layer of bolus to eliminate the air gap
between the solid water and gel dosimeter. Figure 4.2.1.1 shows the experiment setup
scheme (SAD = 100cm, SSD = 94cm). Each vial was positioned in turn, the correct field
size was set, and then 512 MU were delivered.
a
b
Figure 4.2.1.1 Gel dosimeter setup in solid water for 6MV x-ray output factor
measurement.
4.2.1.2. MR imaging and parameters for gel dosimeter
Before MR scan, the gel phantoms were positioned in a foam holder in two lines
in the MR scanner. The gel phantom was scanned at the same room temperature as
LINAC irradiation (21+/-1oC) using a 3T MRI GE Signa® EXCITETM clinical scanner
(GE Healthcare, Milwaukee, WI) with 4-channel head coil with parameters: FOV = 20cm;
61
matrix size = 512 x 512; TE = 10, 110; TR = 4000 ms; NEX = 1; slice thickness = 3 mm
and pixel size = 0.4mm x 0.4mm for 30mins. Then gel dosimeters were repositioned in
the foam holder in order to fit into the 7T scanner and scan at same time. 7T MR
Monkey gradient coil with 15cm diameter single channel RF coil (Bruker Corporation,
Billerica, MA) were used with these parameters: MSME: Field of View (FOV) = 12cm
Slice Thickness=3mm; Matrix size = 512 x 512; pixel size = 0.23mm; TE = 10, 20 …
160 ms and TR = 4000ms (room temperature 20.5oC). In the MAGIC polymer gel
dosimeter R2 map, 3 x 3 x 1mm3 ROIs were manually selected near central X-ray axis. In
the MAGIC polymer gel dosimeter R2 map, 3 x 3 x 1mm3 ROIs were manually selected
near central X-ray axis.
Figure 4.2.1.2 MAGIC gel dosimeter protocol for MR measurement of radiation dose.
Seven gel dosimeters were irradiated under different field size (1x1, 2x2, 3x3, 4x4, 6x6,
8x8 and10x10 cm2)
62
Table 4.2.1.2 List of MRI scanning parameters used to image gel dosimeter irradiated
with different field size.
Imaging modality
3T clinical MRI
Pixel size (mm)
0.4 x 0.4
SNR
48
7T micro MRI
0.23 x 0.23
56
Scanning parameters
FSE: FOV = 20cm; matrix size =
512 x 512; TE = 10, 110; TR =
4000 ms; NEX = 1; slice
thickness = 3 mm
MSME: Field of View (FOV) =
12cm Slice Thickness=3mm;
Matrix size = 512 x 512; pixel
size = 0.23mm; TE = 10, 20 …
160 ms and TR = 4000ms
4.2.1.3. Ionizing chamber setup and measurements
Three ionizing chambers (PTW:TN30013, Exradin A12, Capintec PR-05P) and
an electrometer (INOVISION Model: 35040 S/N:99353) were used in this study. Similar
to the gel dosimeter setup, ionizing chambers were place in 6mm depth of solid water
with SSD =94cm. Two chambers, the PTW:TN30013 and Exradin A12 have their unique
solid water inserts which match the chamber size. The smaller volume chamber, Capintec
PR-05P, which is primary designed for Gamma Knife quality assurance doesn’t have
such special designed insert for LINAC. In order to place the Capintec PR-05P in the
correct position, the chamber was first taped on the solid water. Then, carefully 6mm
thick solid water phantom was put on the right and left side of chamber and secured the
position using tape. Finally, 5.7 mm thick solid water phantom were carefully placed on
the top of the chamber (SSD =94cm SAD = 100cm).
measurements were taken for good measurement statistics.
63
For each field size, five
a
b
c
Figure 4.2.1.3. Ionization chamber setup in solid water for 6MV x-ray output factor
measurement. (a) Experiment setup scheme (b). Ionizing chamber PTW:TN30013 in
solid water. (c) Capintec PR-05P setup in solid water.
4.2.1.4 Radiochromic film measurements and calibration
Radiochromic films EBT2 were irradiated as a comparison to gel dosimeters
under the same experiment setup and condition. Each film was placed horizontally
between solid water slabs. (Figure 4.2.1.4.1) In this work, Radiochromic EBT2 film (ISP
Technologies, Wayne, NJ) was scanned in the EPSON 10000XL and handled using
gloves to reduce the possibility of damage and contamination according to AAPM TG64
55(A. Niroomand-Rad Chair 1998). Each piece of radiochromic film was marked at the
corner to identify the gantry direction and this orientation was maintained during
irradiation and film readout. A series of radiochromic films were first calibrated in 6MV
x-ray from 3Gy to 5Gy range. (Figure 4.2.1.4). The software package ImageJ was used to
manually select 3 x 3 mm2 square ROIs in each film for calibration and small field ROF
measurements. Green channel images are used in data analysis.
a
b
Figure 4.2.1.4. (a)Radiochromic film setup in solid water for 6MV x-ray output factor
measurement. SAD =100cm SSD = 94cm. (b) 10cm x 10cm radiation field and region of
interest (ORI) for output measurement.
4.2.1.5 TLDs measurements and calibration
As a comparison method to gel dosimeter measurements for small fields, LiF-100
TLDs (3mm x 3mm x 1mm) were used for radiation fields (1x1, 2x2 3x3, 4x4, 6x6, 8x8,
10x10 cm2). LiF-100 was measured using HARSHAW 5500 multi-chip reader (Figure
4.2.1.5). Tweezers were used to handle each TLD chips to avoid contamination. After
65
each measurement, TLD was annealed at 400oC for 60min and cooling to 100oC for 2
hour. Only about one to two measurements were taken every night. Since TLD
measurement is extremely labor intensity, only one TLD chip were calibrated in 10 x 10
cm2 6MV photon (3, 4 and 5Gy) for small field ROFs measurements. Each field was
measured three times using the same TLD chip. Elapsed time was over two weeks to
finish all small field ROFs measurements and calibration for TLDs.
a
b
c
Figure 4.2.1.5 (a) TLD experiment setup for output factor measurements and calibration.
(b) TLD setup in the center of the radiation field. (c) HARSHAW 5500 multi-chip reader
4.2.2 Results:
4.2.2.1 MAGIC gel:
R2 values were plotted as a function of field size. Standard deviations were also
calculated in the ROIs. ROFs for a 1x1 cm2 field is 0.612 (+/- 0.034) in 3T and 0.622(+/0.028) (Table 4.2.2.1). The results indicated that R2 relaxivity measured in 7T (7.02+/0.34s-1) is a little bit higher than R2 measured in 3T (6.55+/-0.46s-1). The average
standard deviation (STD) measured from 3T MRI is 4.8% compared to 3.8% for 7T MRI.
66
The main improvement in STD is due to the high signal-to-noise Ratio (SNR=56) for 7T
MR imaging compared to SNR (48) in 3T.
Table 4.2.2.1 Gel dosimetry R2 measurements using 3T clinical scanner and 7T micro
canner
Field Size (cm2) 10 x10
R2 (s-1)
12.23
measured in 3T +/- 0.60
8x8
11.99
+/-0.64
6x6
11.86
+/-0.56
4x4
11.68
+/-0.52
3x3
11.49
+/-0.59
2x2
11.31
+/-0.35
1x1
10.02
+/-0.56
control
6.55
+/-0.46
0.957
+/-0.051
12.57
+/-0.41
0.935
+/-0.050
12.44
+/-0.51
0.904
+/-0.041
12.26
+/-0.51
0.871
+/-0.044
12.07
+/-0.37
0.839
+/-0.026
11.88
+/-0.48
0.612
+/-0.034
10.63
+/-0.49
-
0.960
+/-0.031
0.997
0.937
+/-0.038
0.998
0.894
+/-0.037
1.011
0.861
+/-0.026
1.011
0.851
+/-0.034
0.985
0.622
+/-0.028
0.984
-
MRI
ROFs (3T)
1.000
+/-0.049
R2 (s-1)
12.82
measured in 7T +/-0.51
7.02
+/-0.34
MRI
ROFs (7T)
1.000
+/-0.040
1.000
ROF3T/ROF7T
-
4.2.2.2 Ionizing chamber:
Ionizing chamber results shows reproducible characteristics for small field
measurements. The standard deviation for the three ionizing chambers measurements
were less than 0.5%. Due to the partial volume irradiation, PTW:TN30013 and Exradin
A12 underestimate ROFs of field size smaller than 3 x 3 cm2. The 0.05 cm3 volume
Capintec PR-05P chamber’s result agree with the result published in 2008 (Das 2008).
Table 4.2.2.2 Ionizing chamber measurements for small field relative output factors
(ROFs)
Field Size (cm2)
ROFs (PTW:TN30013)
ROFs(Exradin A12)
ROFs(Capintec PR-05P)
10 x10
1
+/-0.004
1
+/-0.003
1
+/-0.005
8x8
0.975
+/-0.004
0.974
+/-0.003
0.976
+/-0.004
6x6
0.949
+/-0.003
0.939
+/-0.004
0.943
+/-0.005
67
4x4
0.892
+/-0.004
0.887
+/-0.003
0.896
+/-0.003
3x3
0.837
+/-0.003
0.832
+/-0.003
0.863
+/-0.003
2x2
0.665
+/-0.003
0.636
+/-0.002
0.811
+/-0.005
1x1
0.256
+/-0.001
0.266
+/-0.001
0.586
+/-0.002
4.2.2.3 Radiochromic film:
EBT2 radiochromic films were calibration using a range of six doses in 6MV
photon 10 x 10cm2 field. 3x3mm2 ROIs were selected manually in ImageJ which are
described in the previous section. Signal intensity was plotted as a function of dose. The
results (Figure 4.2.2.3) shows signal intensity as a function of dose is close to linear from
3Gy to 5Gy. Standard deviation of radiochromic film is about 4%. The same 3 x 3mm2
ROIs were manually selected for 10 x 10cm2 to 1 x 1cm2 field size. The averages of
signal intensity were transferred to dose according the calibration curve in Figure 4.2.2.3.
Figure 4.2.2.3 EBT2 radiochromic film dose calibration curve. Signal intensity changes
as a function of absorbed dose.
Table 4.2.2.3 EBT2 radiochromic film measurements for small field relative output
factors (ROFs)
Field Size (cm2)
ROFs (EBT2)
10 x10
1
+/-0.052
8x8
0.979
+/-0.038
6x6
0.941
+/-0.042
68
4x4
0.892
+/-0.031
3x3
0.852
+/-0.035
2x2
0.783
+/-0.032
1x1
0.578
+/-0.022
4.2.2.4 TLDs:
Light outputs were plotted as a function of dose. Three measurements were taken
for individual dose (3, 4 and 5Gy). Standard deviation of TLD measurements range from
5%-5.5% (Figure 4.2.2.4). ROFs measurements were plotted as a function of field size.
(Table 4.2.2.4) Minimum STDs 4.1% and maximum STD 5.4% were observed in TLD
small field measurements.
Figure 4.2.2.4 TLD (LiF-100) dose calibration curve in 10 x 10 cm2 6MV photon (3, 4,
5Gy).
Table 4.2.2.4 TLD measurements for small field relative output factors (ROFs)
Field Size (cm2)
ROFs (TLD)
10 x10
1
+/-0.046
8x8
0.955
+/-0.052
6x6
0.935
+/-0.046
69
4x4
0.874
+/-0.036
3x3
0.854
+/-0.040
2x2
0.824
+/-0.045
1x1
0.598
+/-0.032
4.2.2.5 Summary
6MV x-ray OFs versus field size for all detector systems were graphically
compared. Table 4.2.2.5 is the summary of the ROFs measured in different devices.
MAGIC gel ROFs are approximately 2-4% higher than small volume dosimeters
(TLD, EBT2 film and the Capintec PR-05P ionizing chamber). For larger volume ion
chamber (PTW:TN30013 and Exradin A12, both ~ 0.6cc) measurements, ROFs were
underestimated (OF = 0.26 for 1x1cm2) due to non-equilibrium and partial volume
conditions. For polymer gel dosimeter measurements, acquiring T2 weighted image with
higher SNR could improve gel dosimetry accuracy. However, as a trade-off, it will
double the scan time in order to reduce the STDs introduced by MRI scanner by a factor
square root of 2.
Table 4.2.2.5 Summary of the ROFs measured in different devices.
Field Size (cm2)
ROFs (3T)
ROFs (7T)
ROFs (PTW:TN30013)
ROFs(Exradin A12)
ROFs(Capintec PR-05P)
ROFs
(EBT2)
ROFs
(TLD)
10 x10
1.000
+/-0.049
1.000
+/-0.040
1
+/-0.004
1
+/-0.003
1
+/-0.005
1
+/-0.052
1
+/-0.046
8x8
0.957
+/-0.051
0.960
+/-0.031
0.975
+/-0.004
0.974
+/-0.003
0.976
+/-0.004
0.979
+/-0.038
0.955
+/-0.052
6x6
0.935
+/-0.050
0.937
+/-0.038
0.949
+/-0.003
0.939
+/-0.004
0.943
+/-0.005
0.941
+/-0.042
0.935
+/-0.046
70
4x4
0.904
+/-0.041
0.894
+/-0.037
0.892
+/-0.004
0.887
+/-0.003
0.896
+/-0.003
0.892
+/-0.031
0.874
+/-0.036
3x3
0.871
+/-0.044
0.861
+/-0.026
0.837
+/-0.003
0.832
+/-0.003
0.863
+/-0.003
0.852
+/-0.035
0.854
+/-0.040
2x2
0.839
+/-0.026
0.851
+/-0.034
0.665
+/-0.003
0.636
+/-0.002
0.811
+/-0.005
0.783
+/-0.032
0.824
+/-0.045
1x1
0.612
+/-0.034
0.622
+/-0.028
0.256
+/-0.001
0.266
+/-0.001
0.586
+/-0.002
0.578
+/-0.022
0.598
+/-0.032
Figure 4.2.2.5 Relative Output Factor versus field size measured with various available
radiation detectors listed in the paper for 6 MV beams. The data are plotted without the
consideration of any corrections or perturbations and are simply the ratio of reading in a
field size to the reference field (10x10 cm2).
71
4.3 Sub-millimeter radiation field dose profile measurement (Ding 2010)
For very small radiation fields (<1cm diameter) or micro-beams (<1mm diameter),
increased image resolution and signal-to-noise ratio (SNR) are critical and may be
afforded by using MR field strengths greater than 3T. In this initial work, we compare the
performance of 3T and 7T MR imaging for the dose readout of a MAGIC polymer gel for
two 6MV small field defined by symmetrical collimator jaw (5 x 5 mm 2 and 10 x 10 mm2)
and customized 2mm diameter collimator. This study has been cited by MB Kakakhel in
Medical Physics 2011 and IV Koptyug in Progress in Nuclear Magnetic Resonance 2011.
4.3.1Methods and Materials:
4.3.1.1 MAGIC gel preparation:
The same glucose-based MAGIC gel was produced as described previously
(gelatin swine skin [9% (w/w) 90g]; methacrylic acid [4%(w/w) 40g]; copper(II) sulfate
penta-hydrate[0.025g], ascorbic acid [0.3522g (2mM)] Glucose (10%(w/w) 100g, and
deionized water). The glucose-based MAGIC gel was instilled into 5cm diameter plastic
tubes and sealed with plastic wrapper to prevent outside oxygen from entering. Then, the
completed MAGIC gel phantom was stored in a refrigerator (4oC) to set overnight before
irradiation.
4.3.1.2 Polymer gel and radiochromic film irradiation
Small field irradiation was performed using a 6MV X-ray beam from a 2100c
clinical linear accelerator (Varian Medical Systems, Palo Alto, CA). Two small fields 10
72
x 10 mm2 and 5 x 5 mm2 were defined using symmetrical collimator jaws (Figure 4.3.1.2).
The gel phantom was vertically placed on the treatment table with the source-to-surfacedistance (SSD) set to 100cm. Film was placed in 1.5cm depth of solid water and dose
profile of gel dosimeter was measured at 1.5cm depth of the gel surface.
For the 2mm diameter irradiation, a custom collimator block with dimension (7.8
x 7.5 x 15cm3) was made to create a unique 2mm diameter aperture in the middle of the
block. The block was placed on the solid water while film and gel dosimeter were placed
1.5cm under the phantom surface. (4.3.2.2)
Radiochromic EBT2 (ISP Technologies, Wayne, NJ) is used as a comparison to
the gel dosimeter. The film was handled using gloves to reduce the possibility of damage
and contamination according to AAPM TG-55. The exposed radiochromic films were
scanned using an Epson 10000XL scanner 20 hours after irradiation. The readout
protocol is the same as previous section.
a
b
Figure 4.3.1.2 Two small fields 10 x 10 mm2 and 5 x 5 mm2 experiment setup. (a) a
cylindrical gel dosimeter is placed vertically in the middle between solid water. (b)
Gafchromic film setup
73
a
b
Figure 4.3.2.2 (a) 2mm diameter custom collimator. (b) Experiment setup for 2mm
diameter field irradiation
74
4.3.1.3. MR imaging and parameters:
A 7T micro-MR scanner (Bruker Corporation, Billerica, MA) was used for high
resolution gel dosimeter reading (Figure 4.3.3.1). The high magnetic field strength
provides images with very high resolution and signal to noise ratio (SNR) compared to
the 1.5 T and 3 T clinical scanners typically used in clinic. However, the 7T micro-MR
scanner is limited by its aperture diameter and special RF coil design. As a result the
cylindrical gel phantom size for the 7T MRI scanner could not exceed 15 cm in diameter.
Moreover, due to the limited amount of B0 homogeneity in the scanner isocenter, only a
single sample can be scanned at a time. So the gel container size is very important
consideration for use of the micro-MR high resolution scan.
The 10mm x 10mm and 5mm x 5mm beam irradiated gel phantom was first
scanned in the 7T MR using a Rat gradient coil combined with a linear birdcage volume
RF coil with 7.2cm diameter (Bruker Corporation, Billerica, MA) with these parameters:
Field of View (FOV) = 4cm Slice Thickness=2mm interslice distance = 5mm; Matrix
size = 256 x 256; pixel size = 0.156mm; TE = 30, 60 … 300 ms and TR = 3330ms
(Figure 4.3.1.3). Total scan time was about 30mins. Next, the gel phantom was scanned
at the same room temperature using a 3T MRI GE signa® EXCITETM clinical scanner
(GE Healthcare, Milwaukee, WI) with 8-channel head coil. An alternative pulse sequence
to MSME, Fast-Spin-Echo (FSE) was used with parameters: FOV = 6.5cm; matrix size =
256 x 256; TE = 50, 100; TR = 4000 ms; NEX = 1; slice thickness = 2 mm and pixel size
= 0.254mm x 0.254mm
75
Table 4.3.1.3.1 List of MRI scanning parameters used to image 1cm x 1cm and 0.5cm x
0.5cm square fields.
Imaging modality
Pixel size (mm)
SNR
MRI parameters
3T MRI
0.254x0.254
6
7T MRI
0.156 x 0.156
90
FOV = 6.5cm; matrix size = 256 x
256; TE = 50, 100; TR = 4000 ms;
NEX = 1; slice thickness = 2 mm and
pixel size = 0.254mm x 0.254mm
FOV = 4cm Slice Thickness=2mm
interslice distance = 5mm; Matrix
size = 256 x 256; pixel size =
0.156mm; TE = 30, 60 … 300 ms and
TR = 3330ms
For 2mm diameter beam irradiation, a Bruker mouse 35mm diameter quad coil
was used to provide a better image quality and improved SNR. (Pixel size = 0.079mm in
1.5 hours scan and 0.039mm in 12hours overnight scan). Both results were compared to
the Gafchromic film and 3T clinical MRI with pixel size = 0.156mm x 0.156mm.
Table 4.3.1.3.2List of MRI scanning parameters used to image 2mm diameter fields. γγ
Imaging modality
Pixel size (mm)
SNR
MRI parameters
3T MRI
0.156x0.156
3
7T MRI
0.079 x 0.079
76
0.039 x 0.039
87
FOV = 8cm; matrix size = 512 x 512;
TE = 50, 100; TR = 4000 ms; NEX =
1; slice thickness = 2 mm and pixel
size = 0.156mm x 0.156mm
Field of View (FOV) = 4cm Slice
Thickness=2mm interslice distance =
5mm; Matrix size = 512 x 512;
NAX= 4; pixel size = 0.079mm; TE =
30, 60 … 300 ms and TR = 8000ms
Field of View (FOV) = 4cm Slice
Thickness=2mm interslice distance =
5mm; Matrix size = 1024 x 1024;
NAX = 20; pixel size = 0.039mm; TE
= 30, 60 … 300 ms and TR = 8000ms
(lower resolution)
7T MRI
(higher resolution)
76
a
b
C
Figure 4.3.1.3 (a) 7T Bruker micro-MR scanner setup for gel dosimeter. (b) Gamma
Knife radiosurgery irradiated gel phantom scanned at 7T with the MSME pulse sequence
for T2 maps calculation. (c) Gel dosimetry protocol for dose profile measurement in 3T
and 7T MRI scanner.
4.3.2 Results:
Polymer gel MR images are shown for 7T and 3T field strengths in Figure 4.3.4.
Which demonstrate higher image fidelity and less noise at 7T (Figure 4.3.4.1 a) vs 3T
(Figure 4.3.4.1 b). Radiochromic film, obtained simultaneously at the time of gel
77
irradiation, documents the field sizes (Figure 4.3.4.1 c). Dose profiles are shown for the
7T and 3T readouts and the radiochromic film (Figure 4.3.4.1 d, e).
Figure 4.3.2.1 (a) Gel phantom R2 map, axial plane, calculated from 7T MRI scan, pixel
size = 0.156 mm. (b) Gel phantom R2 map, axial plane, calculated from 3T MRI scan,
pixel size = 0.254mm. (c) Radiochromic film irradiated by 5 x 5 mm 2 and 10 x 10 mm2
fields. d) dose profile of 10 x 10 mm2 6MV beam at depth of 1.5cm. (e) dose profile of 5
x 5mm2 6MV beam at depth of 1.5cm.
Table 4.3.2.1 Full width at half maximum measurements in 7T.
FWHM
EBT2
7T MRI gel readout
Difference (%)
5 x 5 mm2
5.35mm
5.23mm
-1.5%
78
10 x 10 mm2
10.32mm
10.12mm
-1.9%
Table 4.3.2.2 Full width at half maximum measurements in 3T.
5 x 5 mm2
5.35mm
4.89mm
-8.6%
FWHM
EBT2
3T MRI gel readout
Difference (%)
10 x 10 mm2
10.32mm
9.85mm
-4.6%
Figure 4.3.2.2 Dose profile of 2mm round shaped x-ray beam field measured using 3T
clinic GE SIGNA scanner, 7T micro-MRI Bruker scanner and radiochromic film.
Table 4.3.2.2 Measurements of Full width at half maximum (FWHM) of 2mm diameter
beam profile.
FWHM
EBT2
7T MRI (12-hour)
2.05 mm 2.15mm
4.8%
Error (%) -
79
7T MRI (1.5-hour)
2.15mm
4.8%
3T MRI (30mins)
2.43mm
18.5%
4.3.3. Conclusion:
7T micro-MRI gel imaging greatly improves the dose measurement in both
accuracy and spatial resolution compared to 3T clinical scanner. At 7T field strength, 3D
dose distribution images with voxel size of 0.156mm x 0.156mm x 2mm were achieved
within 30 minutes of scan time using a 150mm inside diameter linear birdcage volume
coil (max gradient 200mT/m). A higher resolution (~79um x 79um x 2mm) is achieved
within a 2 hour scan time using a 35mm inside diameter quandrature birdcage volume
coil (max gradient 1000mT/m). 35um x 35um x 2mm high resolution was achieved in 12
hours scan. This work suggests to use 7T MRI for radiation field smaller than 0.2 cm x
0.2cm. This approach could be potentially used to measure a micro-beam (~100um) dose
distribution.
80
CHAPTER FIVE
SMALL FIELD APPLICATION: GAMMA KNIFE
RADIOSURGRY
5.1 Introduction
GK stereotactic radiosurgery (Elekta, Stockholm, Sweden) is one of primary
methods used to treat intracranial lesions by administering and concentrating highintensity cobalt-60 gamma ray radiation over a small volume. The GK delivers highprecision, high-gradient dose to distribution to targets in the brain and is considered as
one of the most accurate devices for stereotactic radiosurgery (SRS). Nowadays Gamma
Knife has become increasingly attractive because of the excellent accuracy which spares
the nearby healthy brain tissue and the risks of gamma knife radiosurgery treatment are
very low (Chin LS 2000). The first GK device was invented by Lars Leksell and Börje
Larsson in 1967 at the Karolinska Institute in Stockholm, Sweden. The Wake Forest
University Comprehensive Cancer Center, has the latest GK model, Perfexion, with 192
cobalt-60 sources. In GK radiosurgery a stereotactic head frame is attached to a patient’s
head with a fiducial marking system which establishes a 3D coordinate system for target
localization. The GK has mechanical accuracy and precision of less than 1mm.
Localization errors for target coordinates should not exceed about 1mm and rely
exclusively on the stereotactic imaging process. Therefore, a careful quality assurance
program needs to be conducted to make sure the dose delivered is the dose prescribed in
the treatment plan. In this work, we use MAGIC polymer gel dosimeter as an alternative
81
method to evaluate the Gamma Knife dose distribution. It describes an experimental
procedure with potential to assess the overall accuracy associated with GK clinical
applications, from patient imaging, dosimetry planning to patient positioning and dose
delivery using the automated positioning system of GK Perfexion.
5.2 Verification: multiple small targets
In this study, we use MAGIC gel contained in a spherical plastic bottle and
deliver five different doses inside the gel phantom to simulate the entire gamma knife
treatment process. Multiple dose matrices are one of the unique planning methods in
which several dose matrices are used to produce isodose distribution for cases having
more than one tumor target that are widely spread. We present the evaluation of GK dose
calculation accuracy using the MRI-based MAGIC gel technique. Data analysis was
performed with custom MATLAB code. The dose calculated from Leksell Gamma Plan
(LGP) was used for comparison with the gel measurement.
5.2.1. Methods and Materials:
5.2.1.1. Gel preparation
Glucose-based MAGIC gel [Gelatin 9%; Methacrylic acid 4%; CuSO4 0.1mM;
Ascorbic ascid 2mM; Glucose 10%, and deionized water] is prepared according to the
formula described in Chapter II and IV. The MAGIC gel was instilled into a ball-shaped
plastic bottle (Figure 5.2.1.1). Then the GK gel phantom was stored in a refrigerator (4oC)
82
to set overnight. Before irradiation, the GK gel phantom was attached to the Leksell GK
headframe by fastening four fixation screws with pads so that the reference coordinate
system defined by fiducial markers lays aligned with the stereotactic coordinate system of
the GK unit.
a
b
Figure 5.2.1.1. (a) Spherical Gamma Knife gel phantom attached to the Leksell GK head
frame to simulate the target. (b) GK gel phantom attached with GK MRI fiducial box for
stereotactic MRI scan and treatment planning.
5.2.1.2. Treatment planning and Gamma Knife irradiation
The GK gel phantom with fiducial boxs were scanned in a 3T GE MRI scanner
resulting in an image dataset that was transferred to the Leksell Gamma Plan (LGP)
workstation (Figure 5.2.1.2). Five target doses using a 4mm collimator (1.0Gy, 2.5Gy,
5Gy, 7.5Gy, 10Gy @ 50% isodose line) were calculated for five different dose matrices
which were separated from each other by 25mm as shown in Table 5.2.1.2. Table 5.2.1.2
is the summary of the location, dose and time for each shot.
83
Figure 5.2.1.2.1 Leksell Gamma Plan (LGP): five-single shots using the 4mm collimator.
Figure 5.2.1.2.2 Gel phantom irradiated in GK Perfexion
84
Table 5.2.1.2. Summary of the dose plan created for irradiation in GK phantom
Target
Dose@50%
100% Gy
X[mm]
Y[mm]
Z[mm]
Grid[mm]
Collimator[mm]
A
10Gy
20
97
74
90
0.6
4
B
7.5Gy
15
122
74
90
0.6
4
C
5.0Gy
10
72
74
90
0.6
4
D
1.0Gy
2
97
99
90
0.6
4
E
2.5Gy
5
97
49
90
0.6
4
5.2.1.3. MRI gel scanning
The GK spherical gel phantom was scanned in a 3T MRI GE Signa® EXCITETM
clinical scanner (GE Healthcare, Milwaukee, WI) with 4-channel head coil using spinecho sequence with parameters: FOV = 20cm; matrix size = 512 x 512; TE = 10, 110; TR
= 4000 ms; NEX = 1; slice thickness = 3 mm and pixel size = 0.4mm x 0.4mm for
30mins.
a
b
c
Figure 5.2.1.3 (a)T2- weighted image of GK spherical gel phantom with 4mm
collimator shots delivering five different doses. (b) R2-map of GK spherical gel phantom
calculated in MATLAB. (c) Dose distribution in Leksell Gamma Knife Planning system
with a 0.5mm dose grid size.
85
Table 5.2.1.3 List of MRI scanning parameters used to image the spherical GK gel
phantom.
Imaging modality
3T MRI
Pixel size (mm)
SNR
0.4
81
MRI parameters
FSE: FOV = 20cm; matrix size = 512
x 512; TE = 10, 110; TR = 4000 ms;
NEX = 1; slice thickness = 3 mm and
pixel size = 0.4mm x 0.4mm for
30mins
5.2.1.4. Analysis methods
R2 maps were calculated using the two-echo method described in Chapter III
using an in-house MATLAB program (The MathWorks Inc., Natick, MA). The LGP
software saves calculated dose data in a 3D matrix form 145 x142 x 144 elements at
0.5mm dose resolution (Figure 5.2.1.3(c)). The dose matrices were exported in DICOMRT format and analyzed with MATLAB code. The dose response calibration curve was
plotted by taking the R2 value in the center of each shot (averaging 4-6 pixels value form
ROI diameter~1mm) as a function of corresponding dose @ 100% in LGP (Figure
5.2.2.1). A MATLAB program was used to combine the calculated dose data from each
slice into a single matrix. The matrix size was adjusted to match the MR images and LGP
dose. Dose profile measurements were compared in both X and Y directions in the center
slice between MRI gel readout and LGP.
5.2.2 Results
GK spherical phantom dose profiles were measured through the central axis along
the X and Y directions. Figure 5.2.2.1 shows the dose calibration curve obtained from the
methods described in the previous section and Figure 5.2.2.2. Compared to the 6MV X86
ray calibration curve (1.113 s-1.Gy-1), glucose-based MAGIC gel with Co-60 γ rays
(~1.25MeV) has very small energy dependent (1.1752 s-1.Gy-1). Figure 5.2.2.3 show the
dose profiles through the central X and Y axis.
Figure 5.2.2.1 GK gel phantom R2 calibration curve
Figure 5.2.2.2 A comparison of the central profile along the X direction (10Gy, 15Gy,
20Gy shots) with calibrations made by LGP treatment planning and measurements by
gel dosimeter.
87
Figure 5.2.2.3 A comparison of the central profile along the Y direction (5Gy, 2Gy,
20Gy shots) with measurements made by LGP treatment planning and gel dosimeter.
The biggest difference between the profiles (Figure 5.2.2.2) is in the low dose
regions near the edge of the penumbra,. For the Full width at half maximum (FWHM)
measurements, the biggest discrepancy between gel and LGP is about 4.5% (7.5Gy@50%
isodose line) and 3.6% (1.0Gy @ 50% isodose line) . There is no significant difference
between others targets and the treatment plan (Table 5.2.2.1).
88
Table 5.2.2.1 Full width at half maximum measurements along central X axis
FWHM
Target #1: 10Gy@50%
Target #2: 7.5Gy@50%
Target #3: 5.0Gy@50%
LGP (mm)
6.20
6.14
6.12
Gel readout (mm)
6.31
6.08
5.84
Error (%)
+1.7
-1.0
-4.5
Table 5.2.2.2 Full width at half maximum measurements along central Y axis
FWHM
Target #1: 10Gy@50%
Target #4: 2.5Gy@50%
Target #5: 1.0Gy@50%
LGP (mm)
6.10
6.22
6.10
Gel readout (mm)
6.11
6.11
5.88
Error (%)
+0.0
-1.8
-3.6
5.3 Verification: single shot, anthropomorphic phantom
In this study, we described an experimental procedure using human skull phantom
insert with a gel capsule phantom to simulate an intracranial target. This system is used to
verify GK dose delivery and to check for target localization errors.
5.3.1. Methods and Materials:
5.3.1.1. Gel preparation and stereotactic imaging
The same glucose-based MAGIC gel is prepared which was described in Chapter
II and IV. [Gelatin 9%; Methacrylic acid 4%; CuSO4 0.1mM; Ascorbic ascid 2mM;
Glucose 10% and deionized water], (all available from Fisher Scientific). Then gel is
instilled into a target capsule isolated from oxygen (Figure 5.3.1.1). The phantom is
filled with water except the target capsule is filled with glucose-based MAGIC gel. The
89
skull phantom was stored in a refrigerator (4oC) to set overnight. A cooler is used during
the transportation. Before irradiation, the skull phantom was attached to the Leksell GK
CT stereotactic head frame by fastening four fixation screws with pads so that the
reference coordinate system defined by CT fiducial markers lays aligned with the
stereotactic coordinate system of the Gamma Knife unit. The skull phantom was
stereotactically scanned in both CT and MRI for GK treatment planning.
a
b
Figure 5.3.1.1 (a) Human skull phantom filled with water with gel insert to simulate a
target in the brain. (b) Axial MRI structure image using T1 weighted pulse sequence for
GK treatment planning. Bright circle indicates the gel capsule position relative to fiducial
box and skull.
5.3.1.2. Treatment planning and Gamma Knife single shot irradiation
One single shot dose was given in the center of the gel capsule (Figure 5.3.1.2).
The summary of the location, dose and time is listed in Table 5.3.1.2. A single dose
matrix consists of 19 × 19 × 19 grid points.
90
Figure 5.3.1.2 Screenshot from LGP with CT images in the axial (a), coronal(b), and
sagittal(c) planes. The plan shows a single shot using 4mm collimator, prescribed in the
center of the gel target with the coordinate X:125.8mm, Y:57.5mm, Z: 85.4mm
Table 5.3.1.2 A single shot for human skull phantom irradiation
Target
Dose@50%
100% Gy
X[mm]
Y[mm]
Z[mm]
Grid[mm] Collimator[mm]
A
20Gy
40
125.8
57.5
85.4
0.6
4
5.3.1.3. MRI gel scanning
The GK spherical gel phantom was scanned in a 3T MRI GE Signa® EXCITETM
clinical scanner (GE Healthcare, Milwaukee, WI) with 4-channel head coil using spinecho sequence with parameters: FOV = 20cm; matrix size = 512 x 512; TE = 10, 110; TR
= 5000 ms; NEX = 1; slice thickness = 2 mm and pixel size = 0.4mm x 0.4mm. Then,
T2-weighted gel images were transported to LGP for automated fusion and registration
based on anatomy structures. Figure 5.3.1.3 display the MR images (50% intensity)
91
overlaid on top of CT images (50% intensity) after automated fusion and registration
process in LGP. An absolute dose profile was obtained by using the calibration curve in
section 5.2. Dose profile comparison between LGP and gel dosimetry along X direction
was done in MATLAB as described in the previous section. Full width at half maximum
(FWHM) was measured directly from the dose profiles.
Table 5.3.1.3 List of MRI scanning parameter used to image the skull phantom
Imaging modality
3T MRI
Pixel size (mm)
0.4
SNR
83
MRI parameters
FSE: FOV = 20cm; matrix size = 512
x 512; TE = 10, 110; TR = 4000 ms;
NEX = 1; slice thickness = 3 mm and
pixel size = 0.4mm x 0.4mm for
30mins
Figure 5.3.1.3 the MR images (50% intensity) overlay on top of CT images (50%
intensity) after automated fusion and registration process in LGP. Water in CT images is
black because of HU = 0. Bone is black in MRI because of no water proton.
92
5.3.2. Results:
Figure 5.3.2.1 shows a direct comparison between prescribed dose in LGP and
T2-weighed gel dose read-out MR images. The gel MR read-out images were fused with
CT structure images to match the bony structure and soft tissue. Then CT image intensity
was adjusted to 0%, so that gel MR read-out images with 100% intensity were clearly
shown on the LGP for direct comparison. This method could help physicist evaluate the
treatment plan and actual dose distribution in 3D more conveniently. Figure 5.3.2.2 is the
gel dose profile measurement of 4mm collimator single shot along X direction. The
measurement was taken in the center slice along X direction. The results indicate that the
profile matches in the high dose area but with differences in the low dose area near the
penumbra. The FWHM from 4mm collimator shot treatment plan is 6.20mm along X
direction compared to 6.13mm from gel dosimetry (Table 5.3.2.1).
Figure 5.3.2.1 Prescribed dose in Leksell GammaKnife Plan (LGP) combined with T2weighed gel MR images. Gel MR images were first fused with original CT structure
images. Then CT image intensity was adjusted to 0%. Direct comparison could be done
directly in LGP.
93
Figure 5.3.2.2 A comparison of the central profile along central X direction with
measurements made by LGP treatment planning and gel dosimeter.
Table 5.3.2.1 Full width at half maximum measurements along central X direction
FWHM
20Gy@ 50%
LGP
6.20
Gel readout
6.13
94
Error (%)
-1.1
5.4 Localization error: incorrect stereotactic imaging due to fiducial chemical
shift
The excellent treatment delivery accuracy of Gamma Knife stereotactic
radiosurgery stems from the increased accuracy and precision of each individual step in
the entire treatment procedure, which includes imaging, target localization, treatment
planning, dose calculation and patient positioning, in addition to high mechanical
accuracy and precise dose delivery. In the clinic, the use of MRI as the primary imaging
modality in radiosurgery has become increasingly attractive because of its excellent
depiction of the brain’s normal anatomy and demonstration of lesions not seen with other
imaging modalities e.g. CT. To assure the quality of the MRI-based treatment procedure,
an exhaustive quality control program is required. Polymer gel dosimeter quality
assurance protocol described above satisfies this requirement. In MRI-based GK
treatment, cupric sulfate solution is the liquid fiducial marker material used to indicate
image position. This liquid is filled in the hollow channels of the MR stereotactic fiducial
box(Figure 5.4.a). However, what will happen if the MR fiducial box is filled with the
‘wrong’ liquid? It is well known that MR images are sensitive to the chemical shift
(Hendee, Medical Imaging Physics, 4th Ed.). If ‘baby oil’ is used to replace cupric sulfate
solution, chemical shift could be induced for the localizer, displacing the MR-detected
coordinate system from the true coordinate system, resulting in GK target localization
errors. In this case how much error would occur for 3T MRI-based Gamma Knife
treatment? Also, could this dose localization error be detected using polymer gel
dosimeter protocol? In this study, we use the glucose-based MAGIC gel described above,
95
a grid phantom and the skull phantom to evaluate potential localization and positioning
errors due to the use of incorrect liquid filling material for a GK MRI fiducial box.
a
b
c
Figure 5.4 (a) MR fiducial box attached to a spherical-shaped gel phantom. Fiducial
marker channel is highlighted (yellow). (b)Human skull phantom filled with water with
gel insert to simulate a target in the brain. (c) Axial MRI structure image using T1
weighted pulse sequence for GK treatment planning. Bright dots show position of
fiducial marker.
5.4.1. Chemical Shift in MRI
Chemical shift artifact manifests as bright or dark outlines predominantly at fat /
water interfaces where the image of fat is shifted from its true position along frequency
encoding directions. In clinical practice, fatty tissue locations are anatomically incorrect
and can alter or even shield true anatomy. In a low magnetic field strength (<1.5T)
scanner, chemical shift artifacts can generally be ignored in image interpretation. It does
however, present serious problems in high magnetic field (> 1.5T) scanner, because shifts
can be 1mm or greater. Commonly, GK patients only receive MRI scan for treatment
planning. Since 3T or even higher field MRI clinic scanners are more and more popular,
a small error in fiducial marker materials could be a disaster that goes unnoticed unless
96
tested through a quality assurance procedure. MRI scanner uses the frequency of the
signal to indicate spatial position (Figure 5.4.1.1). Since water resonates at a slightly
different frequency than that of fat, the MRI scanner mistakes the frequency difference as
a spatial (positional) difference. This frequency difference results from the different
electron environments of the protons of water and of fat. The difference in chemical shift
is approximately 3.5 parts-per-million (ppm) which at 1.5 Tesla corresponds to a
frequency difference between that of fat and water of approximately 220 Hz. In 3T, it is
about 440Hz.
Figure 5.4.1.1 A frequency encoding gradient was applied throughout the object.
According to the Larmor equation, the proton precess in a higher frequency locates in
high gradient area.
We can calculate chemical shift using the equation below:
Chemical shift = FOV.Chemical shift frequency for magnet/Receiver Bandwidth
( 5.4.1)
As an example, in our current T1 Axial sequence for GK patient, FOV =25cm, Receiver
bandwidth = 45.5kHz, Frequency shift for 3T = 448Hz. Thus, possible chemical shift in
our GK MRI protocol = 25cm . 448Hz / 45.5kHz = 2.46mm.
97
The amount of chemical shift for our current GK treatment protocol was tested by
placing babyoil, a high fat liquid, in (1) a target capsule in skull phantom filled with
water and (2) the channels of a GK MR fiducial box attached to a grid phantom. Figure
5.4.1.2 shows the baby oil filled capsule’s relative position in skull phantom using our
current GK protocol. From the picture, we are able to measure the position error compare
its original position. The result shows it is 2.45mm higher than it supposed to be.
Figure 5.4.1.2. An axial image of the skull phantom. The baby oil filled capsule (bright
circle) is located in the center of the water filled skull phantom. Six white dots on the left
and right are the fiducial marks indicating the accurate position of each image. The
capsule is 2.45mm shifted towards the frequency encoding direction (Y direction).
For the fiducial box with baby oil, Figure 5.4.1.3 shows expected error from an
MR-based GK treatment plan. Blue dots are the actual fiducial marker positions. Red
dots are the apparent fiducial mark position with 2.45mm shift in frequency encoding
direction. After registering the MR image in the LGP, a GK shot was planned for the
98
center of the capsule. However, the actual object (light blue) is higher than the apparent
object (dark blue). As a result, the actual target in the object is lower than the localized
position.
Figure 5.4.1.3. Relative position of fiducial marks between “baby oil” and copper
sulfate MR images. Blue dots are the actual fiducial mark position. Red dots are the
apparent fiducial mark positions with 2.45cm shift in frequency encoding direction. After
registering MR images in the LGP, we gave a dose in the center of the capsule. However,
the actual object (light blue) is higher than the apparent object (dark blue), so the dose
would be delivered to the incorrect location.
Besides the localization error in Y direction (frequency encoding direction), the
same error happens in slice encoding direction, because the slice selection gradient is
very similar to the frequency gradients. It is a technique to isolate a single plane in the
object, by only exciting the spins in that plane (z-direction). It follows the same Larmor
frequency equation :
(5.1.1)
99
where B0 is the main magnetic field 3T, Gz is the gradient strength, z is the position
along slice selection direction.
Next, in order to test the chemical shift in both Y and Z directions (frequency and
slice encoding directions), the fiducial box channels were filled with baby oil instead of
cupric sulfate. A grid phantom with two insert (horizontal and vertical Teflon rods) was
used to evaluate the chemical shift in both direction.
5.4.2. Methods and Materials:
Two phantoms were used to evaluate geometric errors due to chemical shift for
“baby oil” fiducial markers: 1) a cylindrical grid phantom with two kinds of Teflon rod
inserts (vertical and horizontal grids); 2) a human skull phantom with a glucose-based
MAGIC (Gelatin 9%; Methacrylic acid 4%; CuSO4 0.1mM; Ascorbic acid 2mM;
Glucose 10%) gel capsule to simulate an intracranial target. Both phantoms were
stereotactically MR and CT scanned for GK treatment planning using a “baby oil” GK
MRI fiducial box and a standard CT fiducial box. The gel capsule was given 20Gy @ 50%
isodose line using the 4mm collimator based on “baby oil” fiducial box MR images. Then,
the skull phantom with capsule was scanned in a 3T GE Signa® EXCITETM clinical
scanner (Spin Echo pulse sequence) for gel dose distribution readout with parameter:
FOV = 20cm; matrix size = 512 x 512; TE = 10, 110; TR = 4000 ms; NEX = 1; slice
thickness = 3 mm and pixel size = 0.4mm x 0.4mm. Dose distribution errors were
analyzed based on gel results for the combined MR and CT plans. For comparison,
deviation of grid phantom teflon rods were directly measured from MR and CT
stereotactic images.
100
Figure 5.4.2 Protocol used to evaluate the positioning and localization error using grid
phantom and skull phantom with gel capsule insert.
5.4.3. Results:
Both CT and MRI images of the grid phantom were fused in the LGP for
comparison. The direct measurement from CT and MR images shows 2.7mm shift in the
frequency encoding direction (Figure 5.4.2.1.a) and 2.8mm shift in the slice encoding
direction (Figure 5.4.2.1.b). The combined chemical shift results in a total 3.8mm shift in
both Y and Z directions which shows on the Sagittal plan of the treatment planning
system (Figure 5.4.2.2) .
101
a
b
Figure 5.4.2.1(a) Axial MR images of vertical grid phantom, MR images fused with CT
images. Direct measurement from the Teflon rods in the MR and CT grid phantom
images indicate an average 2.7mm shift in frequency encoding direction in Gamma Knife
treatment planning. (b) Coronal image of horizontal grid phantom MR images fused with
CT images indicating a 2.7mm shift in slice encoding direction. White dots in CT images
are Teflon rods. Black dots in the MR images are Teflon rods.
Figure 5.4.2.2. Sagittal plane of vertical grid phantom MR images fused with CT images.
Direct measurement from the Teflon rods MR and CT grid phantom images indicate
average 3.8mm shift in frequency encoding and slice encoding direction in Gamma Knife
treatment planning. White line is the CT Teflon rod. Black line is Teflon rod in MR
images.
Gel dosimeter results were fused with original treatment plan based on the baby
oil fiducial marker. The results from gel dosimetry match the grid phantom results. The
102
dose profile of the apparent plan and actual plan indicates a 3.8mm shift in the same
direction as the grid phantom result due to inducing chemical shift in the fiducial box
markers, and thus, a systematic localization error for the entire GK coordinate system for
the images used.
Figure 5.4.2.3. (a,b,c) Comparison between gel MR images and LGP on the correct CT
fiducial box coordinates. (d,e,f) 3.8 mm shift in frequency and slice encoding directions
is observed.
103
Figure 5.4.2.4. Dose profile measurements indicate a 3.8 mm shift in slice and frequency
encoding direction.
5.4.4 Conclusion
The study shows that gel dosimeter is able to detect small position error for clinic
practice purpose. The results confirmed the localization error which was found in the
grid phantom. This simple MR-based dosimetry system for Gamma radiosurgery: image
-> plan -> irradiation ->image ->fuse could be widely used as an alternative quality
assurance procedure in routine clinic. Since Gamma radiosurgery focus on the very small
target (4mm diameter) in the brain, the wrong filling materials totally miss the target in
3T MRI based gamma treatment plan. Nowadays, 3T MRI is still not commonly seen in
radiation oncology.
However, with the fast development of image devices, more
institutions are interested in upgrading their MRI unit to a high magnetic field strength
e.g. 3T, 7T and even 10T body scanner. Institutions which are or will be using high
104
magnetic field strength unit for GK treatment, should be aware of the potential targeting
localization errors due to chemical shift that could be substantial.
105
CHAPTER SIX
CONCLUSION
This dissertation investigated the use of 3T and 7T high field MR imaging for
high resolution 3D glucose-based gel dosimetry, with applications shown for small field
megavoltage photon beams. A MAGIC gel recipe was formulated and characterized, then
applied for irradiation geometries using 6 MV x rays and Cobalt-60 γ rays. Specific
conclusions are the following:
1.
A new 10% glucose-based MAGIC formulation was found which
increases the melting point by 4oC, increasing gel stability during the irradiation and MRI
dose readout processes while maintaining gel tissue equivalency. Addition of glucose
also increased the R2 sensitivity by a factor of 3.6, enabling better discrimination of
different radiation doses. The increased R2 sensitivity may be due to the large glucose
molecule, which may limit free water proton movement to cause a decrease in T2
relaxation time and enhanced polymerization. The chemical mechanism behind this
phenomenon needs further investigation.
2.
The glucose-based MAGIC gel dosimeter has in small-sample vials was
shown to be effective for small field relative output factor measurements for 6 MV x rays.
Both 3T and 7T gel dosimetry output factors matched well with three other techniques (a
small volume ionization chamber, TLDs, and radiochromic film) with a range of
agreement of 2-4%. A special solid water insert for the gel dosimeter container may be
helpful to aid in positioning, to minimize any subtle changes in scattering conditions.
106
Although the ionization chamber is still the most reliable and reference instrument for
point dosimetric measurements, the 3D gel method can provide accurate point, 2D, and
3D dose measurements (dose profiles,
percent depth doses, isodose surfaces), all
obtained from one irradiation followed by a single MRI scan.
3.
Micro scale spatial resolution for 3D dose distribution readout is possible
using the gel dosimeter and high field 7T MRI. Enhanced temperature resistance in the
glucose-based MAGIC gel makes overnight scanning possible, as demonstrated by dose
profile assessment of a 2mm diameter 6 MV x-ray field. The high field strength of a 7T
MRI scanner is highly recommended in order to obtain high spatial resolution (40 – 80
µm) and suitable SNR (~50) for dose evaluation of sub-millimeter fields. It is strongly
recommend to obtain a higher SNR image with a lower resolution instead of a higher
resolution and lower SNR (SNR smaller than 10) image. Otherwise, the large noise will
introduce unexpected uncertainties in the MRI-based dose calculation. These results
show that the modified glucose-based MAGIC gel dosimeter with 7T MRI readout could
be potentially used for micro-beam verification. Further development of novel pulse
sequences or other techniques may improve the results and represent possible future work.
4.
A fast and simple gamma radiosurgery quality assurance procedure using
3D gel dosimetry was accomplished by fusing the gel 3T MR readout image back to the
radiosurgery treatment plan. Small localization errors could be easily detected in the
treatment plan. Compared to traditional film-based quality assurance, the use of MRI as
the primary imaging modality in gamma radiosurgery treatment planning and MRI as the
gel dose read-out enables a convenient 3D comparison of expected and measured dose. A
detailed dose profile comparison and analysis between treatment plans can be done in
107
MATLAB. The MRI-based gel technique can be used for single shot and multi-shots
dose verification as well as assessment of localization errors.
In summary, advantages of high field MRI-based gel dosimetry include: (a) a
glucose gel that can be configured into various geometries and is tissue-equivalent for
megavoltage photon energies; (b) super high resolution (up to 40 micrometer pixel size)
through the use of 3T and 7T MRI; (c) 3D verification characterization of individual
small field megavoltage photon beams; (d) 3D verification and quality assurance for
multi-field radiation treatment plans; And (e) simultaneous single- and multi-target dose
distribution verification for gamma radiosurgery plans. Disadvantages for the glucose gel
dosimeter may prohibit its application for routine clinical duty and include these aspects:
(a) labor intensive for manufacture, requiring a certain level of expertise;. (b) temperature
sensitivity that must be accounted for during the irradiation and imaging processes; (c)
time sensitivity relating to the propagation of polymerization as well as temperature
monitoring; and (d) oxygen sensitivity that also requires special attention to insure gel
quality and constancy of response. When these aspects are accommodated the glucose gel
dosimeter shows robust performance for multiple 3D dosimetry applications relevant to
small field radiation treatments.
108
References:
A. E. Nahum, (1996) “Perturbation effects in dosimetry. I. Kilovoltage x-rays and
electrons,” Phys. Med. Biol. 41, 1531–1580
A. Kumar, D. W. a. R. R. E. (1975). "NMR Fourier Zeugmatography." J. Magn. Reson
18.
Andrews H L, M. R. E. a. L. E. J. (1957). "Gel dosimeter for depth dose measurments."
Rev. Sci. Instrum. 28: 329.
A. Niroomand-Rad Chair, C. R. B., B. M. Coursey, K. P. Gall, J. M. Galvin, W. L.
McLaughlin, A. S. Meigooni, R. Nath, J. E. Rodgers, and C. G. Soares (1998).
"Radiochromic film dosimetry: Recommendations of AAPM Radiation Therapy
Committee Task Group 55." Med. Phys. 25.
Alexander P, C. A. a. R. M. (1954). "The degradation of solid polymethylmethacrylate by
ionizing radiations." Proc. R. Soc. A 223: 392.
Audet C, H. M., Jirasek A and Duzenli C (2002). "CT gel dosimetry technique:
comparison of a planned and measured 3D stereotactic dose volume." J. Appl. Clin. Med.
Phys. 3.
Araki, F. (2006). "Monte Carlo study of a Cyberknife stereotactic radiosurgery system."
Med. Phys. 33.
Baldock C, H. P., Piercy AR and Healy B (2001). "Experimental determination of the
diffusion coefficient in two-dimensions in ferrous sulphate gels using the finite element
method." Australas. Phys. Eng. Sci. Med. 24: 19.
Bloch, F. (1946). "Nuclear Induction." Physical Review 70.
BOAG, J. W., (1950). Ionization measurements at very high intensities. 1. Pulsed
radiation beams. British Journal of Radiology, 23, 601-611.
B. E. Bjärngard, J.-S. Tsai, and R. K. Rice, (1990)“Doses on the central axes of narrow 6MV x-ray beams,” Medical Physics, vol. 17, pp. 794-799,
Brahme, A. (1984). "Dosimetric precision requirements in radiation therapy." Acta
Radiol. Oncol. 23: 397.
C. Baldock. (2006). "Historical overview of the development of gel dosimetry." J. Phys.:
Conf. Ser. 56.
109
C. Baldock., Y De Deene, S Doran, G Ibbott, A Jirasek, M Lepage, K B McAuley, M
Oldham and L J Schreiner (2010). "Polymer gel dosimetry." Phys. Med. Biol. 55.
C, J. A. I. a. D. (2002). "Relative effectiveness of polyacrylamide gel dosimeters applied
to proton beams:Fourier transform Raman observations and track structure calculations."
Med. Phys. 29.
C, M. M. L. a. B. (2003). "Ultrasound tomography imaging of radiation dose
distributions in polymer gel dosimeters: preliminary study." Med. Phys. 30.
C. Baldock, R. P. B., N. C. Billingham, G. S. Wagner, S. Patval, R. D. Badawi and S. F.
Keevil, (1998). "Experimental procedure for the manufacture of polyacrylamide gel
(PAG) for magnetic resonance imaging (MRI) radiation dosimetry." Phys. Med. Biol. 43.
C. Kirisits, D. G., P. Wexberg, B. Pokrajac,D. Glogar and R. Pötter (2002).
"Determination and application of the reference isodose length (RIL) for commercial
endovascular brachytherapy devices." Radiother. Oncol 64: 309.
C. Martens, C. D. W. a. W. D. (2000). "The value of the PinPoint ion chamber for
characterization of small field segments used in intensity-modulated radiotherapy." Phys.
Med. Biol. 45: 2519.
C. McKerracher, and D. I. Thwaites,(1999) “Assessment of new small-field detectors
against standard-field detectors for practical stereotactic beam data acquisition,” Physics
in Medicine and Biology, 44, 2143-2160,
Chin LS, L. B., Biggins T, Amin P (2000). "Acute complications following gamma knife
radiosurgery are rare." Surg Neurol 53(5).
Chu KC, J. K., Battista JJ, Van Dyk J and Rutt BK (2000). "Polyvinyl alcohol Fricke
hydrogel and cryogel: two new gel dosimetry systems with low Fe3+ diffusion." Phys.
Med. Biol. 45: 955.
D.A. Low, P. Parikh, J.F. Dempsey, S. Wahab, S. Huq, (2003)“Ionization chamber
volume averaging effects in dynamic intensity modulated radiation therapy beams,”
Medical Physics, 30, pp. 1706-1711
Das, I. J. (2008). "Small fields:Nonequibrium radiation dosimetry." Medical Physics
35(1): 206.
De Deene Y, V. A., Hurley C, Healy B J and Baldock C (2002). "Dose–response stability
and integrity of the dose distribution of various polymer gel dosimeters." Phys. Med. Biol.
47.
110
De Deene Y, H. C., Venning A, Vergote K, Mather M, Healy B J and Baldock C. (2002).
"A basic study of some normoxic polymer gel dosimeters." Phys. Med. Biol. 47: 3441.
De Deene Y, D. W. C., Van Duyse B, Derycke S, Mersseman B, De Gersem W, Voet T,
Achten E and De NeveW (2000). "Validation ofMR-based polymer gel dosimetry as a
preclinical three-dimensional verification tool in conformal radiotherapy." Magn. Reson.
Med. 43.
De Deene Y, H. C., Venning A, Vergote K, Mather M, Healy B J and Baldock C. (2002).
"A basic study of some normoxic polymer gel dosimeters." Phys. Med. Biol. 47: 3441.
De Deene Y, V. A., Hurley C, Healy B J and Baldock C (2002). "Dose–response stability
and integrity of the dose distribution of various polymer gel dosimeters." Phys. Med. Biol.
47.
De Deene, K Vergote, C Claeys and C DeWagter (2006) The fundamental radiation
properties of normoxic polymer gel dosimeters: a comparison between a methacrylic acid
based gel and acrylamide based gels Phys. Med. Biol. 51, 653
De Deene Y, V. d. R., Achten E and DeWagter C (1998). "Mathematical analysis and
experimental investigation of noise in quantitative magnetic resonance imaging applied in
polymer gel dosimetry." Signal Process. 70.
De Deene Y, V. d. R., Achten E and DeWagter C (1998). "Mathematical analysis and
experimental investigation of noise in quantitative magnetic resonance imaging applied in
polymer gel dosimetry." Signal Process. 70(85).
Duthoy W, D. G. W., Vergote K, Boterberg T, Derie C, Smeets P, De Wagter C and De
Neve W (2004). "Clinical implementation of intensity-modulated arc therapy (IMAT) for
rectal cancer." Int. J. Radiat. Oncol. Biol. Phys. 60.
E. M. Purcell, H. C. T. a. R. V. P. (1946). "Resonance absorbtion by nuclear magnetic
moments in a solid." Phys. Rev. 69.
E. Pappas, T. G. Maris, A. Papadakis, F. Zacharopoulou, J. Damilakis, N. Papanikolaou,
and N. Gourtsoyiannis, (2006) “Experimental determination of the effect of detector size
on profile measurements in narrow photon beams,” Medical Physics, 33, pp. 3700-3710
EJ, F. H. a. H. (1927). "The chemical action of roentgen rays on dilute ferrosulphate
solutions as a measure of dose." Am. J. Roent. Radium Ther. Nucl. Med. 18: 430.
EJ, F. H. a. H. (1966). "Chemical Dosimetry." In Radiation Dosimetry, eds Attix FH and
Rowsch WC. Academic Press 2.
111
Ertl A, B. A., Zehetmayer M and Frigo P (2000). "High-resolution dose profile studies
based on MR imaging with polymer BANG gels in stereotactic radiation techniques."
Magn. Reson. Imaging 18.
F. H. Attix, (1986) Introduction to Radiological Physics and Radiation Dosimetry, Wiley
& Sons, New York
Farajollahi A R, B. D. E., Ratcliffe A J, Aukett R J and Mills J A (1999). "An
investigation into the use of polymer gel dosimetry in low dose rate brachytherapy." Br. J.
Radiol. 72.
Farajollahi A R, B. D. E., Tattam D and Green S (2000). "The potential use of polymer
gel dosimetry in boron neutron capture therapy." Phys. Med. Biol. 45.
F. Araki, T. I., T. Ishidoya, and H. D. Kubo (2003). "Measurements of Gamma-Knife
helmet output factors using a radiophotoluminescent glass rod dosimeter and a diode
detector." Med. Phys. 30.
F. Crop, N. R., G. Pittomvils, L. Paelinck, W. De Gersem, C. De Wagter, L. Vakaet, W.
De Neve, and H. Thierens (2007). "Monte Carlo modeling of the ModuLeaf miniature
MLC for small field dosimetry and quality assurance of the clinical treatment planning
system." Phys. Med. Biol. 52.
F. Sánchez-Doblado, P. A., R. Capote, A. Leal, M. Perucha, R. Arrans, L. Nunez, E.
Mainegra, J. I. Lagares, and E. Carrasco, (2003). "Ionization chamber dosimetry of small
photon fields: A Monte Carlo study on stopping-power ratios for radiosurgery and IMRT
beams." Phys. Med. Biol. 48.
F. Sánchez-Doblado, R. C., A. Leal, J. V. Rosello, J. I. Lagares, R. Arrans, and G. H.
Hartmann (2005). "Micro ionization chamber for reference dosimetry in IMRT
verification: Clinical implications on OAR dosimetric errors." Phys. Med. Biol. 50.
Fong PM, K. D., Does MD and Gore JC (2001). "Polymer gels for magnetic resonance
imaging of radiation dose distributions at normal room atmosphere." Phys. Med. Biol. 46:
3105.
G. C. Taylor, and E. Lilley,(1978) “The analysis of thermoluminescent glow peaks in LiF
(TLD-100),” Journal of Physics. D: Applied. Physics, 11, pp. 567-581
Gambarini G, C., Gaya S, Petrovichc C, Pirolaa L and Rosid G (2004). "In-phantom
imaging of all dose components in boron neutron capture therapy by means of gel
dosimeters." Appl. Radiat. Isot. 61.
112
Gore J C, R. M., Maryanski M J and Schulz R J (1996). "Radiation dose distributions in
three dimensions from tomographic optical density scanning of polymer gels: I.
Development of an optical scanner." Phys. Med. Biol. 41.
Gore JC, K. Y. a. S. R. (1984). "Measurement of radiation dose distributions by nuclear
magnetic resonance (NMR) imaging." Phys. Med. Biol. 29: 1189.
Gore JC, K. Y. a. S. R. (1984). "Measurement of radiation dose distributions by nuclear
magnetic resonance (NMR) imaging." Phys. Med. Biol. 29: 1189.
Grebe G, P. M., Roll M and Luedemann L (2001). "Dynamic arc radiosurgery and
radiotherapy: commissioning and verification of dose distributions." Int. J. Radiat. Oncol.
Biol. Phys. 49.
Gustavsson H, B. a. S. A. J., Medin J, Grusell E and Olsson L E (2004). "Linear energy
transfer dependence of a normoxic polymer gel dosimeter investigated using proton beam
absorbed dose measurements." Phys. Med. Biol. 49.
Haacke, B., Thompson, Venkatesan "MRI Physical Principles and Sequence Design."
Heufelder J, S. S., Pfaender M, L¨udemann L, Grebe G and Heese J (2003). "Use of
BANG polymer gel for dose measurements in a 68 MeV proton beam." Med. Phys. 30.
Ibbott G S, M. M. J., Eastman P, Holcomb S D, Zhang Y, Avison R G, Sanders M and
Gore J C (1997). "Threedimensionalvisualization and measurement of conformal dose
distributions using magnetic resonance imaging of BANG polymer gel dosimeters." Int. J.
Radiat. Oncol. Biol. Phys. 38.
J, S. J. W. T. a. W. R. (1964). "An Introduction to Radiation Chemistry." New York:
Wiley.
K. Parodi, T. Bortfeld, and T. Haberer (2008), “Comparison between in-beamand offline
positron emission tomography imaging of proton and carbon ion therapeutic irradiation at
synchrotron-and cyclotron-based facilities,” Int. J. Radiat. Oncol., Biol., Phys. 71, 945–
956 .
LJ., S. (2002). "Fricke gel dosimetry." Proceedings of the 2nd International Conference
on Gel Dosimetry, DOSGEL 2001: 15.
L, B. A. (1961). "A polyacrylamide gamma dosimeter." Radiat. Res. 14: 374.
Lepage M, J. P. M., Back S A and Baldock C (2001). "Dose resolution optimization of
polymer gel dosimeters using different monomers." Phys. Med. Biol. 46.
113
Love P A, E. P. M., Leach M O and Webb S (2003). "Polymer gel measurement of dose
homogeneity in the breast:comparing MLC intensity modulation with standard wedged
delivery." Phys. Med. Biol. 48.
Low D A, D. J. F., Venkatesan R, Mutic S, Markman J, Haacke E M and Purdy J A
(1999). "Evaluation of polymer gels and MRI as a 3-D dosimeter for intensity-modulated
radiation therapy." Med. Phys. 26.
M. M. T. Khan and A. E. Martell (1967). "Metal ion and metal chelate catalyzed
oxidation of ascorbic acid by molecular oxygen: I. Cupric and ferric ion catalyzed
oxidation." J. Am. Chem. Soc. 89.
Maryanski M J, G. J. C., Kennan R P and Schulz R J (1993). "NMR relaxation
enhancement in gels polymerized and cross-linked by ionizing radiation: a new approach
to 3D dosimetry by MRI." Magn. Reson. Imaging 11.
Maryanski M J, S. R. J., Ibbott G S, Gatenby J C, Xie J, Horton D and Gore J C (1994).
"Magnetic resonance imaging of radiation dose distributions using a polymer-gel
dosimeter." Phys. Med. Biol. 39.
Matt A. Bernstein, K. F. K., Xiaohong Joe Zhou "Handbook of MRI Pulse Sequence."
Mather M L, D. D. Y., Whitakker A K, Simon G P, Rutgers R and Baldock C (2002).
"Investigation of ultrasonic properties of PAG and MAGIC polymer gel dosimeters."
Phys. Med. Biol. 47.
Mather ML, Baldock C. Ultrasound tomography imaging of radiation dose distributions
in polymer gel dosimeters: preliminary study. Med. Phys 2003;30:2140–2148.
MSDS acrylamide, version 1.11 updated 02/01/2006, Sigma-Aldrich Canada, Ltd
M.Yoshioka, et al (2010), "Improved dose sensitivity of normoxic polyacrylamide gelatin
gel dosimeter with sucrose" J. Phys.: Conf. Ser. 250 012016 (4pp)
Novotny J Jr, D. P., Spevacek V, Tintera J, Novotny J, Cechak T and Liscak R (2002).
"Quality control of the stereotactic radiosurgery procedure with the polymer-gel
dosimetry." Radiother. Oncol. 63.
Oldham M, B. I., Lord C, Smith T A D, McJury M, Warrington A P, Leach M O and
Webb S (1998). "An investigation into the dosimetry of a nine-field tomotherapy
irradiation using BANG-gel dosimetry." Phys. Med. Biol. 43.
Olsson L E, W. B. A., Fransson A and Nordell B (1992). "Diffusion of ferric ions in
agarose dosimeter gels." Phys. Med. Biol. 37.
114
P. Francescon, S. Cora, C. Cavedon, P. Scalchi, S. Reccanello, and F. Colombo, (1998)
“Use of a new type of radiochromic film, a new parallel-plate micro-chamber, MOSFETs,
and TLD 800 microcubes in the dosimetry of small beams,” Medical Physics, 25, pp.
503-511
P. Metcalfe, T. K., A. Elliott, T. Wong, and P. Hoban (1993). "Dosimetry of 6-MV X-ray
beam penumbra." Phys. Med.Biol. 20: 1439.
Ramm U, W. U., Bock M, Kramer M, Bankamp A, Damrau M, Thilmann C, Bottcher H
D, Schad L R and Kraft G (2000). "Three-dimensional BANG gel dosimetry in
conformal carbon ion radiotherapy." Phys. Med. Biol. 45.
Robert J. Senden, Paul De Jean, Kim B. McAuley, and L. John Schreiner (2006) Polymer
gel dosimeters with reduced toxicity J. Phys.: Conf. Ser. 56 156
Stein, M. J. D. a. G. (1950). "Chemical effects of ionizing radiation in some gels." Nature
166: 1146.
Senden R J, D. J. P., McAuley K B and Schreiner L J (2006). "Polymer gel dosimeters
with reduced toxicity: apreliminary investigation of the NMR and optical dose–response
using different monomers." Phys. Med. Biol. 51.
Thwaites, C. M. a. D. I. (1999). "Assessment of new small-field detectors against
standard-field detectors for practical stereotactic beam data acquisition." Phys. Med. Biol.
44: 2143.
Trapp J V, M. G., De Deene Y and Baldock C (2002). "Attenuation of diagnostic energy
photons by polymer gel dosimeters." Phys. Med. Biol. 47.
W. L. McLaughlin, Chen Yun-Dong, C. G. Soares, A. Miller, G. Van Dyk, D. F. Lewis,
(1991)“Sensitometry of the response of a new radiochromic film dosimeter to gamma
radiation and electron beams,” Nuclear Instruments and Methods in Physics Research
Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 302, pp.
165-176
Watanabe Y, P. G. M. a. M. R. B. (2002). "Image distortion in MRI-based polymer gel
dosimetry of Gamma Knife stereotactic radiosurgery systems." Med. Phys. 29.
Wüthrich, K. (1986). "NMR of Proteins and Nucleic Acids." New York: John Wiley &
Sons.
115
Wuu C-S, S. P., Maryanski M J, Liu T, Borzillary S and Weinberger J (2003).
"Dosimetry study of Re-188 liquid balloon for intravascular brachytherapy using polymer
gel dosimeters and laser-beam optical-CT scanner." Med.Phys. 30.
X. A. Li, M. Soubra, J. Szanto, and L. H. Gerig, (1995) “Lateral electron equilibrium and
electron contamination in measurements of head-scatter factors using miniphantoms and
brass caps,” Med. Phys. 22, 1167–1170
X. R. Zhu, J. J. Allen, J. Shi, and W. E. Simon, (2000)“Total scatter factors and tissue
maximum ratios for small radiosurgery fields: Comparison of diode detectors, a parallelplate ion chamber, and radiographic film,” Medical Physics, vol. 27, pp. 472-477
Xuanfeng Ding, J. O., Ryan Best, Marcus Bennett, Inna McGowin, Jennifer Dorand,
Kerry Link and J Daniel Bourland (2010). "High resolution polymer gel dosimetry for
small beam irradiation using a 7T micro-MRI scanner." J. Phys.: Conf. Ser. 250.
Y De Deene, C., A Venning, K Vergote, MMather, B J Healy and C Baldock (2002). "A
basic study of some normoxic polymer gel dosimeters." Phys. Med. Biol. 47: 3441–3463.
Y. De Deene, N. R. a. C. D. "On the accuracy of monomer/polymer gel dosimetry in the
proximity of a high-dose-rate 192Ir source." Phys. Med. Biol. 46.
Zhu, T. C. (2010). "Small Field: dosimetry in electron disequilibrium region." Journal of
Physics: Conference Series 250.
116
Appendix:
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
Xuanfeng (Leo) Ding
336-782-3492
email:[email protected]
10030 Vanderbilt Circle, Rockville, 20850, MD
Education
Ph.D., Medical Physics, ABR Part I passed
Wake Forest University, Winston-Salem, NC,
2009-2012
M.S. Medical Physics
Wake Forest University, Winston-Salem, NC,
2006-2009
B.S. Physics,
Fudan University, Shanghai, China,
2002-2006
Experience
Clinical Medical Physicist trainee (20hrs/week)
Wake Forest University School of Medicine, Department of Radiation Oncology
Supervisors: J. Daniel Bourland, Ph.D. (advisor)











Under the supervision of certified medical physicists, function as a clinical medical physicist to
directly support the clinic and all quality assurance.
Perform Intensity Modulated Radiation Treatment (IMRT) and Volumetric Modulated Arc
Therapy (VMAT) quality assurance on each patient (10 patients /week). Measure precise radiation
dose for each plan using I'mRT matriXX and ArcCheck. Report to supervisor for approval.
Maintain the Varian 2100c, Elekta Axesse and Infinity by performing weekly/monthly/annually
quality assurance.
Perform annual TG-51.
Perform Dosimetrist’s duties designing treatment IMRT and VMAT plans including head and
neck, lung, brain, and pelvis patient in Pinnacle and Monaco treatment plan system.
Utilize diodes, TLDs, film, chromic film as a second check to machine output and radiation fields.
Perform Gamma Knife® PerfexionTM monthly/annual QA.
Assist 3T MRI unit for ACR MRI accreditation program
Assist accepting and commissioning new Linear accelerator (Elekta Axesse in 2009, Elekta
Infinity in 2011)
Teach and mentor new graduate students in weekly/monthly/IMRT/VMAT QA
Work with engineers in the diagnosis, repair, maintenance and approval of treatment machines for
clinical use.
Teaching Assistant
Wake Forest University, Department of Physics
Supervisors: Keith Bonin, Ph.D., Martin Guthold, Ph.D.




2008– 2012
PHY 110 Introduction to Physics Lab
PHY 113/114 General Physics Lab I, II
PHY 266 Intermediate Laboratory
PHY 692 Spec. Top. Adv. Imaging Sciences
133
2006- 2011
Knowledge and Skills
Imaging
 Magnetic Resonance Imaging (GE 1.5T & 3T & Bruker 7T Biospin & Siemens 1.5T )
 GE Pulse Sequence Programming and Design (EPIC) and Siemens Pulse Sequence Programming
(IDEA)
 CT/PET & 4DCT (GE Lightspeed Ultra 8-slice, Varian RPM Gating)
 Elekta Axesse Linac kV Cone Beam CT-XVI
External Beam
 Linear Accelerator Quality Assurance (Elekta Axesse, Elekta Synergy, Varian SC2100)
 IMRT Quality Assurance (PTW 2DArray, RadCalc, Wellhofer I'mRT Matrixx, ArcCheck)
 Stereotactic Radiosurgery (Novalis Brainlab, Elekta SBRT Body Frame, GammaKnife) & (Elekta
Perfexion GammaKnife)
Brachytherapy
 Nucletron HDR, PLATO Brachy Planning, MammoSite Prostate Brachytherapy

(LDR) Prostate Brachytherapy
Treatment Planning
 Pinnacle (ADAC -Philips)
 Perfexion (Elekta)
Development Languages
 Maple, Matlab, C language
 EPIC pulse programming (GE 1.5T/3T MRI Scanners), IDEA pulse programming (SIEMENS
1.5T MRI Scanner)
Operating Systems and software
 Mosaiq: Record and Verify
 Windows 3.X/9X/ME/XP/Vista/7, RedHat Linux, UNIX
Honors and Awards











2011 SEAAPM annual meeting poster award “Honorable Mention”
2011, 2010, 2009, Wake Forest Alumni Student Travel Award
2010 1st place winner of Wake Forest University Graduate School Research Day (1 out of 122)
2010 Southeastern Chapter American Association of Physicist in Medicine (SEAAPM) Travel
Award
2009 Runner-up winner of Wake Forest University Graduate School Research Day (3 out of 116)
2008 Sigma Pi Sigma Honorary Society
2002-2003 Fudan University People’s scholarship
2003,2004,2005,2006 Fundamental Science Scholarship
2001 3rd prize, 7th High School Physics Contest, Shanghai
2000 3rd prize, 6th High School Physics Contest, Shanghai
1994 3rd prize, Logo language programming Contest, Shanghai
Media Reports
2010, ScienceDaily, American Institute of Physics, Medscape Today, etc. Over hundreds media reports in
different languages
“Researchers Use Nanoparticles as Destructive Beacans to Zap Tumors”
 http://www.sciencedaily.com/releases/2010/07/100721172952.htm
 http://www.newswise.com/articles/view/566478?print-article
 http://www.medscape.com/viewarticle/725789
134
2009, PhysicsToday Featured New Picks “Cooking tumor will a little help from nanotubes”
 http://blogs.physicstoday.org/newspicks/2009/07/
 http://blogs.physicstoday.org/industry09/2009/07/carbon-nanotubes-help-cook-cancer.html
Professional Services:





Reviewer, Medical Physics
Reviewer, Journal of Applied Clinical Medical Physics
Reviewer, Magnetic Resonance Imaging
Athletic Chair, Graduate Students Association, Wake Forest University
Department Representative, Department of Physics, Wake Forest University
2011-present
2011-present
2011-present
2010-2011
2010-2011
Workshop & Symposium







SEAAPM symposium “The Practice of Quality Assurance in an Era of
Change”, Myrle Beach, SC, April 6-7, 2011
4th Annual North Carolina IMRT/IGRT Symposium, Greensboro, NC, Oct. 2, 2010
3D dose measurement and computational analysis, IC3DDose, Hilton Head Island, SC, Aug.
24-28, 2010
Advances in MRI: Applications in Diagnosis and Radiation Therapy, SEAAPM symposium,
Augusta, GA, Mar. 4, 2010
Nanomedicine Workshop, Wake Forest University, Winston-Salem, NC, Oct. 19, 2009
The Role of Imaging Informatics in Radiation Oncology, SEAAPM symposium, Chapel Hill,
NC, Mar. 13-14, 2009
Hyperthermia and NanoMaterials , Wake Forest University. Winston-Salem, NC, April 7-9,
2008
Professional Society and Network




American Association of Physicists in Medicine (AAPM)
South East Chapter American Association of Physicists in Medicine (SEAAMP)
North America Chinese Medical Physicists Association (NACMPA)
American Physics Society (APS)
Selected Presentations:
1.
2.
3.
4.
2012 AAPM annual meeting, Charollette, NC, (poster discussion) “Localization Error in
Gamma Radiosurgery with 3T MR Due to Fiducial Box Chemical Shift From Incorrect Liquid
Filling Materials”
2012 AAPM annual meeting, Charollette, NC, (poster discussion) “Small Field Output Factor
Measurement Using MAGIC Gel Dosimeter in 3T MRI”
2011 Joint AAPM/COMP annual meeting, Vancouver, BC, Canada (poster discussion) “3D
Small Beam Dosimetry using MAGIC Gel with a 7T micro MRI unit”
2011, SEAAPM Annual Meeting (poster presentation) “High resolution polymer gel dosimetry
for small beam irradiation”
135
5.
2010, 52nd Annual
Meeting
of
AAPM,
Philadelphia,
PA,
(oral
presentation@Nanotechnology sessions) “ MR Relaxation Properties for Fe-Containing
MWCNTs and Potential for Combined MR Imaging and Tumor Ablation Therapy”
6. 2010, The 6th International Conference on 3D Radiation Dosimetry (IC3DDose), Hilton
Head Island, SC (oral presentation) “High resolution polymer gel dosimetry for small beam
irradiation using a 7T micro-MRI scanner”
7. 2010, Wake Forest Graduate School Research Day, Winston-Salem, NC (1st place winner)
“Potential of MWCNTs for combined MR imaging and tumor ablation”
8. 2010, SEAAPM Annual Meeting, Augusta, GA, (invited talk) “Feasibility Study: Magnetic
Resonance Temperature Imaging Guided Laser-Induced Thermal Therapy with Multi-Walled
Carbon Nanotubes.”
9. 2009, 51st Annual Meeting of the AAPM, Anaheim, CA, (oral presentation @Innovations
and Frontiers in Medical Physics sessions) “Magnetic Resonance Temperature Imaging Guided
Laser-Induced Thermal Therapy with Multi-Walled Carbon Nanotubes.”
10. 2009, Wake Forest Graduate School Research Day, Winston-Salem, NC (running-up winner)
“MR Themal Imaging Guided Laser Therapy using Multi-Walled Carbon Nanotubes as heat
absorber”
Publications
1.
Xuanfeng Ding, J. Daniel Bourland How Wrong, if we use baby-oil in Elekta MRI Gamma
Kinfe Fiducial Box instead of Cupric Sulphate solution (in preparation)
2. Marcus Bennett, David Wiant, Jacob Gersh, Ryan Best, , Xuanfeng Ding, J. Daniel Bourland,
The Causes and Prevention of RF Induced Heating of Gamma Knife Radiosurgery Headframes
During MRI Treatment Planning Scans, Journal of Applied Clinical Medical Physics, 2011
(accepted)
3. Xuanfeng Ding, John Olsen, Rest Best, Marcus Bennett, Inna McGrowin, Jenny Dorand, Kerry
Link, J. Daniel Bourland, Small Beam Dosimetry Using MAGIC Gel with a 7T Micro-MRI
Scanner abstract), Medical Physics, August (2011)
4. Xuanfeng Ding, Ravi Singh, Andrew Burke, Heather Hatcher, John Olsen, Robert A. Kraft ,
David Carroll, Michael Schmid J. Daniel Bourland, Steven Arkman, Frank Torti, Suzy
Torti, Development of iron-containing multiwalled carbon nanotubes for MR-guided laserinduced thermotherapy. Nanomedicine (Lond) April 20th,(2011)
5. Jon Whitney, Heather C. Hatcher, Marissa Nichole Rylander, Saugata Sarkar, Andrew
Burke, Xuanfeng Ding, Suzy Torti, Computational Models and Digital Image Analysis of
Carbon Nanotube Medicated Laser Cancer Therapy, Society of Biomedical Conference (2011)
6. Xuanfeng Ding, J. Daniel Bourland, Ravi Singh, Andrew Burke, Heather Hatcher, John Olson,
David Carroll, Robert Kraft, Suzy Torti, Frank Torti, MR Relaxation Properties for FeContaining MWCNTs and Potential for Combined MR Imaging and Tumor Ablation
Therapy (abstract), Medical Physics, (2010), vol.37, p. 3438
7. Xuanfeng Ding, John Olsen, Rest Best, Marcus Bennett, Inna McGrowin, Jenny Dorand, Kerry
Link, J. Daniel Bourland, High resolution polymer gel dosimetry for small beam irradiation
using a 7T micro-MRI scanner, Journal of Physics conference series, (2010), vol 250, p 0120901212.
8. Marcus C. Bennett, David. Wiant, Jacob. Gersh, Wendy. Dolesh , Xuanfeng Ding , Ryan Best , J.
Daniel Bourland ,Causes and Prevention of MR-Induced Skin Heating for Patients with
Attached Head frames for Gamma Radiosurgery (abstract), Medical Physics, (2010), vol37, p.
3304
9. Andrew Burke, Xuanfeng Ding, Ravi Singh, Robert A. Kraft, Nicole Levi-Polyachenko, Marissa
Nichole Rylander, Chris Szot, Cara Buchanan, Jon Whitney, Jessica Fisher, Heather C. Hatcher,
Ralph D’Agostino Jr., Nancy D. Kock, P. M. Ajayan, David L. Carroll, Steven Akman, Frank M.
Torti, Suzy V. Torti, Long-term survival following a single treatment of kidney tumors with
136
multiwalled carbon nanotubes and near-infrared radiation, Proceedings of the National
Academy of Sciences (2009), vol. 106, no. 31, p. 12897-12902
10. Xuanfeng Ding, Robert Kraft, Andrew Bruke, David Carroll, Suzy Torti, J. Daniel
Bourland Magnetic Resonance Temperature Imaging Guided Laser-Induced Thermal Therapy
with Multi-Walled Carbon Nanotubes(abstract),Medical Physics (2009), vol. 36, issue 6, p. 2775
11. Marissa Nichole Rylander, Jessica Fisher, Cara Buchanan, Chris Szot, Jon Whitney, Robert
Kraft, Suzy Torti, Andrew Burke, Xuanfeng Ding, Controlling Heat Shock Protein Expression in
Laser
Cancer
Therapy
Using
Multi-Walled
Carbon
Multi-Walled
Carbon
Nanotubes(abstract), Lasers in Surgery and Medicine, (2009) vol. 41, issue S21 (p 5)
12. Li Zhang, Linfeng Gao; Shuilong Wu, Xuanfeng Ding; The Shielding Calculation at the Site of
the electron-beam Facility, Chinese Journal of Radiological Health, 2005, vol 14, issue 4
137